pdfcoffee.com kingx27s-indian-warfare-ilya-smirin-2016pdf-pdf-free

KING'S INDIAN WARFARE
By
ILYASMIRIN
Quality Chess
www.qualitychess.co. uk
First edition 20 1 6 by Quality Chess UK Ltd
Copyright © 20 1 6 Ilya Smirin
KING'S INDIAN WARFARE
All rights reserved. No part of this publication may be reproduced, stored
in a retrieval system or transmitted in any form or by any means, electronic,
electrostatic, magnetic tape, photocopying, recording or otherwise, without
prior permission of the publisher.
Paperback ISBN 978- 1 -78483-025-0
Hardcover ISBN 978- 1 -78483-026-7
All sales or enquiries should be directed to Quality Chess UK Ltd,
Central Chambers, Suite 247, 1 1 Bothwell Street Glasgow G2 6LY, United Kingdom
Phone +44 1 4 1 204 2073
e-mail: [email protected]
website: www. qualitychess.co.uk
Distributed in North America by National Book Network
Distributed in Rest of the World by Quality Chess UK Ltd through
Sunrise Handicrafts, ul. Poligonowa 35A, 20-8 1 7 Lublin, Poland
Typeset by Jacob Aagaard
Proofreading by Andrew Greet and John Shaw
Edited by Jacob Aagaard and Ian Kingston
Photo on page 5 kindly provided by Maya Gelfand from the Gelfand archives
Photos on page 7 and 2 1 3 kindly provided by David Llada
Photos on page 39 and 333 kindly provided by Harald Fietz
Photos on page 9, 58, 67, 1 07, 1 5 1 , 1 59, 1 69, 1 98, 233, 266,
273, 30 1 , kindly made available by Max Notkin from the 64 archives
Cover idea by Jacob Aagaard
Cover design by www.adamsondesign.com
Printed in Estonia by Tallinna Raamatutri.ikikoja LLC
Contents
Key to Symbols used
Foreword by Boris Gelfand
4
5
Preface
7
1
Always Dangerous
9
3
The King's Indian Bishop
2
4
5
6
7
8
9
10
11
Kramnik
Line Opening
Destruction of Pawn Structure
Knight Agility
Kamikaze Rooks
Fighting for the lntiative
Materialism
Endgames
My Two Favourite Losses
Name Index
Game Index
39
67
107
151
169
213
233
273
301
333
344
349
Key to symbols used
±
+
+-+
;;
+:!
m
??
!!
!?
?!
#
White is slightly better
Black is slightly better
White is better
Black is better
White has a decisive advantage
Black has a decisive advantage
equality
with compensation
with counterplay
unclear
a weak move
a blunder
a good move
an excellent move
a move worth considering
a move of doubtful value
mate
Foreword
I first played against Ilya in 1 980, when we were both in our early twenties. He was already a
strong player of course - as well as a big King's Indian fan. Since then, he has played this opening
successfully against some of the best players in the world.
I believe there are two types of people who play the King's Indian. One type bases their play on
knowledge and analysis; the other type plays according to feeling and understanding. Ilya Smirin
undoubtedly belongs to the latter camp. I don't think he ever studied any deep theory! He has
an excellent feeling for the middlegame, particularly for dynamic possibilities. When he gets a
chance to attack the king, there are very few who can stop him.
Ilya may not know as much opening theory as some players, but he knows a lot about the history
of the King's Indian. I know that Ilya has been greatly inspired by the games of Mikhail Tal,
Robert Fischer and especially Leonid Stein,
one of his favourite players. The way Ilya
handles the King's Indian has been influenced
by all of them. Stein in particular is one of
Ilya's heroes; the Soviet GM would often aim
for the most complicated position possible, in
order to obtain attacking and other dynamic
possibilities. Ilya possesses the same type of
flair for such positions - and nowhere is his
talent more obvious than in the King's Indian.
Ilya has played the King's Indian successfully
against some of the very best players in the
world. See, for instance, Chapter 2, which
contains Ilya's games against Vladimir
Kramnik, who is known as one of the strongest
players on the White side of the KID. I find
Ilya's success against Kramnik and other elite
players extremely impressive, as each game is
a huge challenge. When you play this opening
you take a big risk from the very beginning.
White gets space in the centre and a lot of
freedom for his pieces, while as Black you must
rely on your dynamic and tactical qualities.
6
Foreword by Boris Gelfand
Ilya is deeply confident in the potential of the black pieces in the King's Indian. Chapter 8
contains the game Gelfand - Smirin from the 2005 European Club Cup in Saint Vincent, where
Ilya played brilliantly and I was really lucky to save the game. Ilya's resourcefulness and feeling for
dynamics was really impressive in this game.
Everything in chess is a reflection of one's personality. Ilya is a big fan of theatre, as well as action
movies - and I think this is reflected in his handling of the King's Indian, which always leads to
a lot of drama and action on the board! I'm sure the readers will enjoy the many ferocious attacks
and dramatic battles contained in this book. Hopefully it will inspire some readers to start playing
the King's Indian - or return to it, if they've not played it for a while.
Reading this book gives you a different type of insight into the King's Indian. It's not a theoretical
book, where you get answers to your opening questions - and probably the King's Indian is not
meant to be played in such a clinical way. Ilya's games have great instructive value, to be sure - but
this is also a book for pleasure and for inspiration. For players who enjoy wild chess, who value
imagination, creativity and epic fights on the board, this book will be a treat.
Boris Gelfand
20 1 2 World Championship Finalist
Preface
In this book I would like to present my best King's Indian Defence games and in the process share
with the reader my views on this complex and double-edged opening.
The King's Indian Defence is one of the most fascinating openings in chess. It involves everything
I love about the game: risk-taking, attacking, exchanging weaknesses or material for dynamic
chances, clever tactics, surprising turnarounds and a deep sense of possibility.
Many outstanding chess players played the King's Indian Defence throughout their whole careers.
It is enough to mention just a few names (in chronological order) : David Bronstein, Efim Geller,
Mikhail Tal, Leonid Stein, Robert Fischer, Garry Kasparov. In modern days the King's Indian
Defence is the main (and successful) opening in the games of Teimour Radjabov, and it is also
played by Hikaru Nakamura, Alexander Grischuk and various other top players.
Not a King's Indian, but at least a current photo...
8
Preface
I have been a great lover of the King's Indian Defence from the very beginning of my chess life
- that's over 30 years now. It is my favourite opening, so when I chatted to Jacob Aagaard after
our game at the 20 1 1 European Team Championship, and he asked if I had ever thought about
writing a book, I felt attracted to share my love for and experience in handling this exciting
opening.
What I want to achieve with this book is to give the reader a course in playing the King's Indian,
which in turn will also be a course in playing dynamic chess. Every chapter will be structured in
the following way:
�
�
�
A few positions from the games in the chapter for the reader to solve
General introduction to the topic; for example, tactics typical of the King's Indian
Some of my own games that are relevant to the topic
The basic idea is to explain the topic in a simple (yet hopefully not simplistic) way and then
show how the concepts showed up in my games, which were played at a high grandmaster level.
I would implore the reader to be realistic when trying to solve the positions from the games.
My intention is to help you broaden your mind and improve your imagination.
I decided quite deliberately to have 49 main games in this book. I could easily have made it 50,
of course, but I like the number 49 for several reasons. 7x7 is one of them, but there is also the
feeling of incompletion. Or in other words: I am leaving room for more. And it is certainly my
goal to add to this collection of memorable King's Indian games in the years to come.
I hope this book will help the reader to play better chess, and inspire more players to take up this
fascinating opening.
Ilya Smirin,
Kfar Sava, August 20 1 6
Chapter 1
Always Dangerous
Since we were young, baby!
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
8
... 8
6
6
7
7
5
5
4
4
2
2
3
3
a
b
c
d
e
f
g
a
h
How can Black fight for the
initiative? (see page 20)
b
c
d
e
f
g
h
It is rare you can play a trap that
is also the strongest move!
(see page 26)
... 8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
Black to play and win
(see page 1 3)
a
b
c
d
e
f
g
a
h
White has j ust received a big
surprise blow. How should he
react? (see page 2 1 )
8
b
c
d
e
f
g
h
A nice little tactic
(see page 33)
8
7
7
5
5
6
6
4
4
2
2
3
3
a
b
c
d
e
f
An easy one
(see page 1 5)
g
h
a
b
c
d
e
f
g
h
Train your positional play.
What should White play?
(see page 25)
a
b
c
d
e
f
g
h
Another nice little tactic
(see page 36/37)
Chapter 1
-
11
Always D angerous
I have played the King's Indian Defence
throughout my entire career. It's a big enough
opening to have enough dimensions not to get
boring. I have played it positionally and I have
played it in the style of a stuntman jumping
out of an airplane without a parachute. And as
you will see in this book, even the endgames
can be quite complicated and require a lot
from the players.
The KID has been a big part of my career and
life, so when I was finally talked into writing
a book by my good friend Boris Avrukh, I
felt that the natural place to start was to write
about my favourite opening. In this book I
hope to give the reader a good understanding
of the potential offered by the King's Indian. I
have chosen to do this through my own games,
based on some overarching themes. The games
are, however, all extensively annotated. I do
not like the idea of pretending that a big fight
lasting for a number of hours can be explained
through a simple idea. Trust me: the players
know the ideas quite well, but there are many
of them and they are often contradictory.
The question is which idea or theme is more
important in a position and who will benefit
from this. In order to determine this, the
players use all the tools available to them
during the game: their tenacity, calculation,
intuition, creativity and so on.
Nimrod Veinberg - Ilya Smirin
Jerusalem 20 1 5
l .d4 d6 2.c4 e5 3.d5 �d7 4.e4 �c5 5.�c3
a5 6.�f3 �f6
We have reached a position that could have
arisen from the King's Indian, but through an
unconventional move order. Black's position is
a bit more flexible than usual.
7.V!lc2 g6 8.b3 .ig7 9J�b l?
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
White mixes up his ideas. He wants to play
a2-a3 and b3-b4 in order to drive away the
knight from c5 . But this allows a common
King's Indian trick.
9. �fxe4! 10.�xe4 �xe4
I won the game on move 25. White cannot
take the knight on account of 1 1 .'1Wxe4 ifS .
..
The main reason I like this opening is that it
is almost always dangerous - for both sides of
course. From the very beginning the board
is transformed into a treacherous minefield
in a way that you will very rarely find in the
Queen's Gambit Declined. It is a very creative
opening in which both players have chances to
invent something from the beginning.
An example of an early disaster for White is
this game from the European Championship.
This i s o f course a rather simple trick, and
certainly uncharacteristic of Veinberg's
usual level, but at the same time also a good
illustration of how perpetually dangerous the
King's Indian is, even for someone who is a
very strong player by most people's standards.
In the following game an IM is approaching
the first time control in an entirely lost
position against an experienced grandmaster.
12
King's Indian Warfare
Still, it would be too early to give up. The
King's Indian supplies you with continual
counter-chances, even when things have gone
wrong.
The reason is found in the following variation:
41 ... �c7 42.�fl ! �c2?! 43.�xc2 �h l t 44.�e2
�xf3t 45.�d2 and White is completely
winning.
Yannick Pelletier - Andreas Skytte Hagen
Legnica 20 1 3
8
7
6
5
a
4
3
2
1
a
b
c
d
e
f
g
h
This game has gone in the way White usually
wants in the King's Indian. Black has thrown
everything, including the kitchen sink, into
an attack on the kingside. He has sacrificed a
piece and pushed his pawn to g3. Nevertheless,
White has managed to contain the attack, keep
the extra piece, and gain a winning position.
But even under such dire circumstances,
there is still hope of victory for the King's
Indian hero. The young Danish player decides
to make more of his rook, which is not doing
much at the moment.
b
c
d
e
f
g
h
Now, imagine that the white rook is on
d 1 instead of e 1 , and Black has 45 ... �e3#!
(Of course White can deviate, but not without
concessions, which is the point.)
4I ..J�c7 42.a6?!
After 42.�fl �c2 43 ..id2 White would
still keep most of his advantage. But it is
also evident that Black has gained some
counterplay. Remember that when we joined
this game, he looked entirely down and out.
Being j ust lost is not that bad!
42 .. J�c2!
8
7
40 .. J�b7!
White's task in this type of position is one of
containment. This is never easy, as you have to
both anticipate the opponent's ideas and find a
good way to meet them, while still advancing
your own agenda.
6
4I J�dl?!
It turned out when we analysed the game
afterwards that 4 1 .�e 1 ! was more accurate.
1
5
4
3
2
a
b
c
d
e
f
g
h
Chapter 1
-
Always Dangerous
43.Wfxc2?
Pelletier does not believe in the sacrifice. I
do not think this was because he was unable to
calculate the consequences correctly. It was just
that on this day, at this moment, a combination
of bad form and hours of trying to keep the lid
on Pandora's Box had drained his powers, and
he missed something. This will be a common
theme in this book, in virtually every game. It
is so basic that I presume that the reader will
take note of it now and not forget it.
44...Wfhl t?
Hagen plays the most tempting move,
presumably missing the winning idea entirely:
44 . . . lt:Jd4! 45.�xd4 f3!!
8
7
6
5
4
3
43.id2 �a2 44.lt:Jc4 �xa6 45 .ia5
2
a
a
b
c
d
e
f
g
13
b
c
d
e
f
g
h
This is another theme you will find again
and again in this book, in every chapter: in
the King's Indian material is only a part of the
equation. The dynamic potential of the pieces
is at least as important.
For example, in this position Black has two
pawns for a rook, a knight and a bishop, but
these two pawns are so powerful that the fight
is decided in Black's favour.
h
White would have kept the advantage,
keeping the black rook out of the game. But
it is clear that Black has also improved his
chances, having eliminated the a-pawn.
45.cbe2 � d4t 46J�xd4 f3t
8
7
43 ... �xf3t 44.cbfl
6
8
5
7
4
6
3
5
2
4
1
3
a
2
1
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
47.cbd3??
A clear sign of bad form. After 47.<±>d2! g2
48.ie6t <±>h8 49.Wc8 it is White who wins.
14
King's Indian Warfare
47 ... £2 48.®c4 fl=�t 49J�d3 g2 50. i> b5
gl=�
My Games
Lembit Oll
-
Ilya Smirin
8
Rostov-on-Don 1 993
7
Lembit 011 started his chess career as a prodigy.
He was a very talented player with a sharp sense
for the initiative and very good calculating
abilities. Lembit had an excellent memory and
was a strong theoretician. Unfortunately, at
some point he began to have mental problems,
which led to his tragic death. Lembit passed
away very young, being just 33 years old. He
held 42nd place in the FIDE rating list at
that time. Incidentally, he was the fi r st USSR
national master I ever played against. We
drew our first encounter back in 1982. Here
is our game from the Bondarevsky Memorial,
which was held in the southern Russian city of
Rostov-on-Don.
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Four queens on the board: three against one.
Black is winning.
5 1 .�c4 �112 52.�h3 �bl 53.®c6 �xh3
54J�xh3 �c2 55J�g3t ®h8 56.®b5 �al
57.�a3 �d4 58.�c3
0-1
In this chapter we shall see eight of my games,
all very different in character. The common
theme (as indeed it is throughout the book)
is the constant opportunity for creativity and/
or an attack on the king. I hope to inspire the
reader to always keep one eye on the prize
and not allow himself to be constrained by
materialism. Chess is like life: it is not about
collecting material goods and counting them
at the end. It is about finding fulfilment and
joy through being the best that you can be
and making the biggest possible mark on
your surroundings. Or in short, to deliver
checkmate!
l .d4 �f6 2.c4 g6 3.�c3 �g7 4.e4 d6 5.�f3
0-0 6.�e2 e5 7.0-0 �c6 8.�e3
Usually White chooses this continuation as a
way to avoid the main variations after 8.d5 CiJe7.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
8 ... �e8
8. . . CiJg4 9 .ig5 f6 1 0 .ic l f5 11 .ig5 has been
a more popular line in the last few decades. Here
is a high-level example from 1 990: 1 1 ... if6
12.ixf6 CiJxf6 13.dxe5 dxe5 1 4.�xd8 �xd8
15
Chapter 1 - Always Dangerous
15 .ltJd5 lDxe4 16.ltJxc7 �b8 17.�fd 1 id?
18.id3 �bc8 19.ltJd5 ltJc5 20.ifl ie6 2 1.b4
ixd5 22.cxd5 ltJxb4 23.ltJxe5 ltJe4 24.ic4
ltJc3 25 .�d2 ltJbxd5 26.g3 <i>g7 27.<i>g2?
a
b
c
d
e
f
g
h
27 . . . ltJe3t 0- 1 Portisch - Kasparov, Linares
1990. Which 8th move to choose is basically a
matter of one's mood during the game.
9.dxe5
9.d5 ltJd4! is the idea.
way - he grabs some space and prepares to
activate the e2-bishop on b5.
10 ....ig4 I I ..ib5 V9c8
A reasonable alternative was 1 1. . . ltJd7
12.ixc6 bxc6 13.h3 ixf3 14. l!Mxf3 ltJ f8, with
a similar structure to that which will arise later
in the game. Black will transfer the knight to
e6 and d4, thus compensating for the doubled
c-pawns.
12.h3 .ih5
I did not like the position after 12 . . . �d8
13.ltJd5 ltJ xd5 14.exd5 ie6 15 .dxe6!? �xd 1
16.exf7t <i>xf7 17.ic4t <i>f8 18.�axd 1 , which
happened in one game. Despite the material
advantage (queen for a rook plus bishop) Black
has to be careful in my opinion - his king is
vulnerable and White's pieces are active and
well coordinated.
13 ..ixc6 bxc6
8
7
6
5
4
3
a
b
c
d
e
f
g
h
10.ltJxd4 exd4 11.ixd4 ltJxe4 12.ixg7 <i>xg7
13 .ltJ xe4 �xe4 14 .l!Mc2 �e8 Yz-Yz was played
in the well-known game Reshevsky - Najdorf,
Zurich 1953, in which 8. . . �e8 was seen for the
first time.
9... dxe5 10.c5
After 1O.l!Mxd8 ltJxd8 11.ltJ b5 ltJe6 12.ltJg5
�e7 the position is rather simple and roughly
equal, a draw being the most probable
outcome. 011 continues in the most principled
2
1
a
b
c
d
e
f
g
h
14..ig5
14.l!Me2 was played against me by Van der
Sterren in a rapid game a year earlier: 14 . . . h6
15.<i>h2 g5 16. l!Mc4 �b8 17.b3 a5 18.ltJd2 g4
19.f3!, with a definite positional advantage for
White: Black's bishops are not very active and
his pawn structure is not ideal either; (0- 1, 34)
Van der Sterren - Smirin, Tilburg 1992.
16
King's I ndian Warfare
Of course, 1 4.g4?? is impossible due to
1 4 . . . lt:Jxg4.
14 .. J�b8 1 5J�b l
A forced draw could be the result of the
"petite combinaison" after: 1 5.ixf6 ixf6
16.g4 �d8 1 7Y!ie2 ixg4 1 8.hxg4 '1Wxg4t
1 9.@h2 Wff4t 20.@g2
a
b
c
d
e
f
g
h
20 . . . �xb2! 2 1.'1Wxb2 Wfg4t
Perhaps 1 5 .b3!? was slightly more precise,
as the rook on a 1 would then be freed from
defending the b2-pawn.
1 5 ... Lf3 16.'i;Yxf3 tiJd7 17.b4
Or 1 7.ie3 lt:J f8 and the knight goes to e6.
17...a5
1 7 . . . lt:J f8 was also perfectly playable, but I
wanted to use the knight's current position on
d7 to exchange the potentially weak pawn on
a7.
18.b5?!
Lembit is not satisfied with 18.a3 axb4
1 9.axb4 lt:J f8=. True to his style, he is looking
for more tactical play.
18 ... tlJxc5?!
During the game I thought that 1 8. . . cxb5
1 9.c6 lt:Jc5 20.lt:Jd5 (White should probably
continue 20.ie3 lt:Je6 2 l .�fd 1 !? lt:Jd4 22.ixd4
exd4 23.lt:Jxb5 ie5 with a slightly worse but
surely defendable position - the bishop on
e5 is stronger than the white knight and the
c6-pawn is more a weakness than strength
here) would be dangerous for Black. Perhaps
neither of us noticed that after 20 . . . lt:Je6
2 l .lt:Je7t? �xe7 22.ixe7 '1We8 23.ia3 Wfxc6
Black would have a material advantage and the
better position.
19.hxc6 tlJe6 20.'i;Ye3!
This is better than 20.ie3 �xb 1 2 l .�xb 1
'1Wa6, when the pawn on c6 is doomed.
20 ... f6 2I ..ih6
Perhaps 2 1 .ih4 �xb 1 22.�xb 1 g5 23.ig3
'1Wa6+ 24.lt:Jd5 Wfxc6 25 .�c l was more
accurate, but it's not easy to voluntarily put
one's bishop in an unfortunate position (g3)
when one is not looking at a computer screen.
2 1 ..J�xbl 22J�xbl 'i;Ya6 23.tiJd5 'i;Yxc6
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
1
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
Chapter 1
-
Right now Black is o n the defensive, but
he has enough resources to neutralize the
opponent's threats. Meanwhile, being a pawn
up gave me some optimism . . .
24J�cl VNb7!
The best square for the queen.
Worse was 24 . . . Wfd6 25.ixg7 @xg7 26.Wfa7,
with the initiative for White.
25.�xg7 ®xg7 26.VNf3 gf8 27.VNg4 f5!
Not falling into the trap: 27 . . J�e8?? 28.tt:lxc7
tt:lxc7 29.Wid7t
28.VNg3 VNb2
30.ga4 f4 3 1 .VNg4?
The second and last mistake. Lembit did not
feel the danger to his king, which has been
abandoned by its defenders.
The only way to minimize the damage was
3 l .Wic3 Wfxc3 (now Black can't create a mating
net by 3 1 . . . Wfb 1 t 32.@h2 tt:\d4 33.Wfxc5 tt:le2
due to 34.f3) 32.tt:lxc3 �a8 33.tt:\d5 and his
active pieces gives White reasonable hope for a
draw in the ending.
3 I . .. tlJd4 32.gxa5 gf7!
After this prophylactic move White is
defenceless against the mating threats. No
wonder Oll overlooked this - Black's attack
appeared as if out of nowhere.
33.gxc5
No better is 33.Wid 1 f3 34.g3 tt:le2t 35 .@h2
tt:lc3 36.Wfe 1 tt:lxe4.
8
7
6
33 ...VNal t 34.<it>h2
5
4
8
3
7
2
6
1
5
a
b
c
d
e
f
g
h
Defending the e-pawn and attacking the
rook at the same time.
29.gc4?
Suddenly White oversteps the safety line.
The only way to equality was 29.Ek6 Wfb 1 t
30.@h2 Wfxe4 3 l .�xe6 Wfxd5 32.�xe5 Wfxa2
33.�e7t �f7 34.Wfxc7, when the weak king
leaves Black with not a single winning chance.
29 c5!
The extra passed pawn advances. Now the
scope of the white rook is very limited.
..•
17
Always Dangerous
4
3
2
1
a
b
c
d
e
f
g
h
34 ... h5!
After 35 .Wig5 tt:le2 White cannot avoid mate
on gl . The black queen in the corner also
protects the pawn on e5. Lembit resigned. I was
very pleased that such a brilliant player as Rafael
Vaganian, who also played in the tournament,
praised my play highly at the end of the game.
0-1
18
King's I ndian Warfare
Alexey Kuzmin - Ilya Smirin
Nabereznye Chelny 1988
Alexey Kuzmin is a well-known theoretician
who for many years has worked in Qatar as a
coach. He is also a long-time opening adviser
of Alexander Morozevich. But at the time this
game was played, Morozevich was still too
young . . .
Safer is 19 ... lL\c8 20.lLlec3 a5 with a complex
struggle.) 20.gxf5 gxf5 2 1.if2 @h8 22.ixh4
if6 23.if2!? (23.ixf6t E!:xf6) 23 ... E!:g8
24.E!:xg8t lLlxg8 25.'1Mfc3 lLlg5 26.ig2
f4 27.lLlg1 'We8 28.'Wa3 ie7 29.lLlc3?
(29.ifl is much stronger) 29 ... 'Wg6 30.@fl
lLl f6 3 1.'1Mfxa6
8
7
l .d4 llJf6 2.c4 g6 3.llJc3 Ag7 4.e4 d6 5.8
o-o 6.i.e3 llJ c6 7.YMd2 a6 s.llJge2 gbs
The most popular move in the position,
preparing . . . b7-b5.
6
I played 8. . . id7 in the game Chigladze Smirin, which you will also find in this book
(page 223) .
2
9.llJcl
9 .E!:c 1 is the popular move nowadays.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
a) 9 . . . b5? 1O.cxb5 axb5 11.lLlxb5 is the main
idea behind the prophylactic 9 .E!:c 1.
5
4
3
a
b
c
d
e
f
g
h
3 1. . . E!:g8! (suddenly Black gets a strong
attack against the opponent's king) 32.ih 1
ixh3t 33.lLlxh3 lLlxh3 34.'Wa7 id8 35.@e2
lLlh5?! (35 ... 'Wh6FF) 36.@d2 lLl g3 37.lLl e2?
(the only continuation was 37.ig2 lLlxe4t
38.lLlxe4 'Wxg2 39.@e2) 37 . . . lL\xh 1 38.E!:xh 1
'Wg2 39.E!:xh3t 'Wxh3 40.'Wa3 ie7 4 1.ib6
'Wfl 42.'Wb3 ih4 43.'1Mfd 1 'Wxf3 0-1 Baron ­
Smirin, Israeli Championship 20 16.
c) In this line both sides often use somewhat
mysterious knight manoeuvres, as in the
following game: 9 . . . e6 I O. lL\d l El:e8 11.g3 lLle7
12.ig2 b5 13.c5 b4 14.lLlf2 a5 15 .0-0 ia6
16.E!:fd 1 lLld7 17.f4 d5 18.e5 lLlf5 19.@h 1 E!:a8
b) Recently I played the following tense and
interesting encounter: 9 ... h5!? (directed against
g2-g4 and potentially against lLl g3) 10.h3
e5 11.d5 lLle7 12.g4 c6 13 .E!:g 1 b5!? 14.b3
bxc4 15 .bxc4 cxd5 16.cxd5 id7 17.lLlg3 h4
18.lL\ge2 ( 18. lL\f5!? gxf5 19.gxf5 lLlxf5 20.exf5
ixf5 with an unclear position) 18. . . lL\h7
19.lLld1 f5 ?! (A brave but risky decision.
a
b
c
d
e
f
g
h
19
Chapter 1 - Always Dangerous
20. lLJg1 lLJb8 2 1 .g4 lD xe3 22.Wfxe3 f5 23.gxf5
gxf5 24.Wfh3 lD c6 25 . .if3 �e7 26.�d2 Wff8
27.Wfh4 lLJd8 28. lLJfh3 And White won
soon; ( 1 -0, 37) Vitiugov - Ganguly, Khanty­
Mansiysk (ol) 20 1 0. The bishop on a6 ended
up being entirely sidelined.
9.h4
This continuation occurred in my short but
very interesting game with Lautier:
9 . . . h5 1 0. lLJc l e5 1 1 .d5 lLJd4 1 2.lDb3 c5
1 3 .dxc6 bxc6
1 3 . . . lLJxb3?!
Black gives up a whole rook in order to keep
the white king in the centre.
20 ..ixb8 lDxc4 2 1 .Wfc l .if5
And in this really wild and unclear position
we both preferred not to take any further risks
(the game was played in a very important match
in the European Club Cup) and we agreed to
a draw, V2-V2 Lautier - Smirin, Rethymnon
2003.
9 J"Lle8!?
9 . . . e5 1 0. lLJb3 (or 1 0.d5 lLJd4 1 1 .lD b3) ts
considered to be the main continuation.
•.
8
7
6
5
4
3
a
b
c
d
e
f
g
h
This is worse due to the intermediate 1 4.c7!
Wfxc7 1 5 .axb3.
1 4 . lLJxd4 exd4 1 5 ..ixd4 �e8
A well-known pawn sacrifice.
1 6 ..ie2 d5 1 7.e5 c5!?
Over the board improvisation.
1 8. .ixc5 lLJd7 1 9 . .id6
a
1 9 . . . lDxe5!
b
c
d
e
f
g
h
2
1
a
b
c
d
e
f
g
h
This is one of my original ideas in the KID.
By withdrawing this knight Black prepares to
create counterplay by means of . . . f7-f5 . I just
thought: "If White can place his knight on c l ,
why Black can't do a similar thing and place his
knight on e8?"
10.d5
In the next game Kuzmin improved with:
1 0.lDb3 f5 1 1 .exf5 .ixf5 1 2 . .ie2 lLJf6 1 3.0-0
e5 1 4.�ad 1 exd4 1 5 . lLJxd4 lLJxd4 1 6 ..ixd4
Wfd7 1 7.�fe 1 �fe8 1 8. .ia7 �bd8 1 9 ..if2 Wff7
20 ..ifl �xe 1 2 1 .�xe 1 �h8 22.b3 �e8 23.�d 1
.ie6 24.g3 �d8 25 . .ig2 h6 26 . .id4, with a
small but lasting edge; ( 1 -0, 48) Kuzmin Badea, Moscow 1 99 1 . I think this is the right
way for White to continue.
20
King's Indian Warfare
10 ... tlJe5 1 1 ..ie2 f5
Today I would prefer 11. . . cS 12.0-0 (or
12.a4 fS 13.f4 lt:Jg4!? 14.ixg4 fxg4 1S .lt:Jd3
id? 16.eS ifS and the two bishops allow
Black to look to the future with optimism)
12 . . . bS 13.cxbS axbS 14.lt:JxbS ia6 with a
good version of the Benko Gambit. A sample
variation is 1S.a4 lt:J c7 16.f4 ( 16.lt:Jxc7 ixe2
17. fffxe2 �xb2!) 16 . . . ixbS 17. axbS lt:J d7
18.lt:Jb3 lt:JxbS and Black is OK, as 19.lt:J aS?!
runs in 19 . . . tt:Jd4.
1 2.£4 tljf7
Also here 12 ... lt:Jg4 13.ixg4 fxg4 was
not bad, even though after 14.0-0 cS
( 14 ... bS 1S .cxbS axbS 16.lt:Jd3 b4 17.lt:Jd1 lt:Jf6
18 .lt:J 1f2±) 1S .dxc6 bxc6 16.lt:Jd3 lt:J f6 17.b3
the position arising is in my opinion easier to
handle with White - the bishop pair doesn't
fully compensate for the pawn weaknesses and
White's solid centre.
Still I would prefer here to take en passant:
1 6.dxc6 bxc6 17.lt:Jd4 id? 18.cS ( 18.fS cS
19.fxg6 hxg6 20.lt:Jf3 ifS with good piece-play
for Black) 18 ... dxcS 1 9.lt:Jb3 ifS 20. fixd8
�fxd8 2 1.ixcS , with a pleasant endgame, for
instance: 2 1. . . lt:JdS 22.ixa6 lt:Jxc3 23.bxc3
ixc3 24.�ac l if6 2S.�fe 1 eS 26.ic4 exf4
27.ie7 id4t 28.lt:Jxd4 �xd4 29.if6 �e4
30.a4 and the passed a-pawn plus the bishop
pair outweigh the missing pawn.
8
7
6
5
4
3
2
1
13.exf5 hf5 14.0-0 t2Jf6
a
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
The knight has fulfilled its role (it made
. . . f7 -fS possible) so it may return.
1 5.t2Jb3 c5 1 6.h3
The usual prophylaxis against a possible
... lt:Jg4.
b
c
d
e
f
g
h
16 ... tlJ e4?
A serious mistake. I made this move as a
matter of "common sense" - to exchange
knights and activate the g7-bishop. But I
missed a much more serious factor - the
activation of White's e2-bishop.
Instead, Black could have played the
energetic 16 . . . bS! 17.cxbS axbS 1 8 .ixbS
lt:JxdS 19. fixdS ixc3 20.bxc3 �xbS with an
excellent position - White has a couple of
pawn weaknesses (a2, c3) and he can't exploit
the weak dark squares around the opponent's
king because of the passive bishop on e3.
17.tlJxe4 .ixe4 18 ..ig4 .ifS
If 18 . . . bS 19.lt:JaS bxc4 20.lt:Jxc4 �b4
2 1.�acl �h8 22.b3 White would keep some
advantage due to his active pieces (the knight
on c4 and bishop on g4) and better pawn
structure.
21
Chapter 1 - Always Dangerous
19 ..ixf5 gxf5 20J3f3
Now White is better, mainly because of the
unfortunate position of the black knight. The
pawn on f5 is also a weakness in Black's camp.
The next few moves are pretty natural.
20 ... <bh8 2 1 .Wc2 b5 22 ..id2 Wd7 23 ..ic3
ggs 24.gel gbe8 25. <bh2
25. El:fe3 was called for; my prospects are not
rosy after that.
8
7
6
5
4
3
2
25 ... ltJd8
b
a
d
c
e
f
g
h
30 ... ltJ d4!
An unexpected blow. This was my idea when
I played 26 . . . e6. It should have led to a draw
by force.
8
7
6
5
31 .Wd3?
Tantamount to resignation. My opponent
was in a kind of shock and did not find the
only defence: 3 l .cxd4 El:e2 32.Wfg5 Wfxd4
4
3
2
a
b
c
d
e
f
g
h
26.ltJd2?
A careless move which lets Black obtain
counterplay with the help of tactics.
After the simple 26.ixg7t El:xg7 27.El:fe3
El:eg8 28.El:l e2 I would have to "wait and see,"
whereas White will gradually strengthen his
position. The weaknesses on e7 and f5 plus the
pressure along the e-file would give White a
clear edge in this line.
26 ... e6! 27.dxe6?!
As is often the case, one mistake is followed
by another. Kuzmin does not see the danger.
27.El:fe3 e5!? 28.a3 would keep part of his
advantage.
27... ltJxe6 28.Wxf5 .ixc3 29.bxc3 Wg7
30.ggl
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
33.El:d3! Wfxd3 34.Wff6t El:g7 3 5 .Wff8t with
perpetual check. To be honest, I also did not
see that.
3 1 . .. ltJe2?!
An unnecessary pirouette. Much stronger
was the straightforward 3 1 . . . lt:Jxf3t 32.Wfxf3
Wfe7, with a trivial win after the unavoidable
exchange of queens.
22
King's I ndian Warfare
32J�fl!?
32.g4 lt:Jxg 1 33.<tt> xg 1 El:e 1 t 34.tt:Jfl Wffe7
35.f5 would be more resilient, even though
Black should win anyway - the extra exchange
is worth something after all.
32 ... tlJxgl 33.@xgl V:Yf6 34.cxb5 axb5
35.@h2
35.g4 c4 36.Wif3 El:e 1 t 37.<tt> g2 was slightly
better. The most precise move here seems to
be 37 . . . Wif5!, with a decisive material and
positional advantage.
35 ... c4 36.V:Yf3 V:Ye6 37.tlJfl d5
37 . . . Wffe4!? would win on the spot.
38.V:Yh5
After the most stubborn 3 8. Wffd 1 Black
would still have a clear win: 38 . . . Wffb 6 39. El:f3
b4! 40.cxb4 (or 40. Wffxd5 b3 and b-pawn
queens) 40 . . . d4 and this duo of passers decides
the outcome.
Alexander Beliavsky- Ilya Smirin
Bel grade 1 998
Alexander Beliavsky was one of the strongest
players in the world between 1 980 and 1 990.
He's always been a tough opponent for me. I
lost our three first encounters - in 1 988, 1 989
and 1 990 (all those games were played in the
USSR Championships) - and every time he
beat me in style. Therefore this game may be
seen as some kind of revenge.
l .d4 tlJ£6 2.c4 g6 3.tiJc3 .ig7 4.e4 d6 5.h3
tiJbd7
The most popular choice here is the natural
move 5 . . . 0-0. In this game, however, I decided
to develop the knight to d7 first, preparing
. . . c7-c5.
a
a
b
c
d
e
f
g
h
38 ... b4
A similar idea in even more favourable
circumstances.
39.cxb4 c3 40.f5 V:Ye5t 4I .@hl d4 42.V:Yf3
V:Ye4
0-1
c
d
e
f
g
h
6..id3
6.ig5 0-0 may lead to a transposition of
moves. Alternatively, Black may proceed with
6 . . . h6. I usually try to avoid this move, which
creates some potential weaknesses on the
kingside (the pawns on h6 and g6 sometimes
become vulnerable) , but on the other hand
this is an easy way to get rid of the bishop on
g5 and break the pin along the h4-d8 diagonal.
Black's play looks logical in the following
encounter: 7.ie3 e5 8.d5 lt:Jc5 9.Wic2 a5
23
Chapter 1 - Always Dangerous
1 0.0-0-0 h5 l l .tDf3 ih6 1 2.ixh6 �xh6
1 3.ie2 mf8 1 4 .'!Md2 mg7 1 5 .'!Me3 '!Me7
and in this position without dark-squared
bishops Black should be OK; Kazhgaleyev Shchekachev, Livry Gargan 2005.
6...0-0 7.i.g5 c5 8.d5
8.lD f3 occurred in the game Avrukh Smirin, Israel 1 999. You will find this game
later, on page 27.
8 .. JtJe5 9.tiJf3
Alexander allows the exchange of his d3bishop, counting on his strong centre and
good development. He is right in my opinion.
9 .ie2 essentially turns out to be a loss of
time, which was exploited by Vlad Tkachiev
in the following game: 9 . . . b5! In the spirit
of the Benko Gambit. 1 0.cxb5 a6 l l .bxa6
WaS 1 2.id2 ixa6 1 3.'!Mc2 �fb8 1 4.�b l
ttJ fd7 1 5 .f4 ttJc4 1 6.ixc4 ixc4 1 7.b3 ixc3!
1 8.ixc3 '!Mxa2 1 9.�b2 '!Ma l t 20.mf2 '!Mfl t
2 I .me3 �a3! 22.bxc4 �xb2 23 .'!Mxb2 ltJb6!
0- 1 Pinter - Tkachiev, Porec 1 998.
. . . f7-f5, undermining the e-pawn and banking
on future counterplay along the f-file.
1 2.0-0
Of course, White could play 1 2.g4 ttJ f6
here, but in this case he would have to answer
some questions, the main one being: "Where
do I hide my king?"
1 2 ... h6
The immediate 1 2 . . . f5 is dubious due to
1 3.exf5 ixf5 1 4.'!Me2 '!Md7 1 5.'!Mxe7 ixh3
1 6.'!Mxd7 ixd7 1 7.ltJe4 and the pawn on d6
cannot be defended.
8
7
6
5
4
3
2
9 ... tlJxd3t 10.YMxd3 a6 1 1 .a4
a
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 I . .. tlJh5!?
This move is connected with a rather
original idea to carry out the advance
b
c
d
e
f
g
h
13.i.e3
In the first game in which I played l l ... ttJh5
my opponent put his bishop on d2: 1 3.id2
e6 ( 1 3 .. .f5?! can hardly be recommended here
- 1 4.�ae l fxe4 1 5 .'!Mxe4 and the pawns on e7
and g6 are hanging) 1 4.dxe6 ixe6 1 5 .ttJd5
(more ambitious was 1 5 .a5!? b6 1 6.axb6
'!Mxb6 1 7.�fd 1 and White's position looks
good)) 1 5 . . . �e8 1 6.�ad l ixd5 1 7.exd5 '!Md7
1 8.b3 b5 1 9.axb5 axbS 20.cxb5 �ab8 2 1 .b4
�xbS Yz-Yz A. Mikhalevski - Smirin, Rishon
Le Zion 1 997.
13 ... £5 14.exf5
1 4.ttJh4!? could have led to an interesting
tactical struggle: 1 4 .. .f4 1 5 .id2
24
King's Indian Warfare
1 6.ttJh4 is not dangerous due to 1 6 . . . W/e8.
1 6... ltJf4
It was not easy to play some "improving the
position" moves, which is why I decided to
force matters.
a
b
c
d
e
f
g
h
1 5 . . . �xc3! 1 6.�xc3 e5 1 7.dxe6 (or 1 7.ttJxg6
El:e8 and the knight on g6 will be lost for
insufficient compensation) 1 7 . . :�xh4 1 8. W/xd6
El:e8 1 9.Wie5 @h7 20.g4 with unclear play. Of
course, this is an example of a "silicon" line.
The following sample line shows that it was
probably the correct decision. 1 6 . . . @h7
1 7.�d2 E!:f7 1 8.ltJe4 Wff8 1 9. El:a3 ( 1 9.ttJ eg5t
hxg5 20.ttJxg5t @g8 2 1 .ttJxf7 W/xf7 22.Wfxe7
El:f8 23.Wixf7t E!:xf7 and this endgame is not
that bad for Black, I guess) 1 9 ... ttJf4 20.�xf4
El:xf4 2 1 .a5 El:b8 22.E!:b3 and I do not like
Black's position.
17.hf4 gxf4
14....ixf5 15.%\'e2 .id?
This prophylactic move prevents the possible
threats g2-g4 or ttJh4.
a
b
c
d
e
f
g
h
18.%\'xe?
Beliavsky chooses the concrete approach.
a
b
c
d
e
f
g
h
The strategic picture is more or less clear
now: White has a better pawn structure, good
pieces and pressure on the e-pawn. My trumps
are the two bishops (yes, again!) and some
dynamic possibilities connected, first of all,
with the f-file. Overall I would perhaps prefer
White's chances here, but Black has his own
reasons for optimism.
In the style of (let's say) Magnus Carlsen would
be 1 8. El:a3, keeping the pressure and making
a useful waiting move. In that case Black
may switch his attention to the queenside:
1 8. . . W/b6 1 9.a5 W/b4 with a tense struggle, bur
again White's chances are somewhat better.
1 8 ....ixh3 19.gxh3
Or 1 9.Wie3 Wif8.
25
Chapter 1 - Always Dangerous
19 .. J�xf3 20J�a3!
Big AI (as they used to call Alexander
Beliavsky in those days) finds a strong resource
- the rook is excellently placed on the third
rank for both attacking and defensive purposes.
22.�e4?
Alexander misses his chance.
Very soon after the game he pointed out that
he could get an overwhelming position after:
22.�xe5! dxe5 23.lt:Je4
20 ....ie5
20 . . .id4 would be a reasonable alternative
here. The following variation looks like a
very logical one: 2 1 .�e6t @h8 22.lt:Je4 �xa3
23.bxa3 '!Nf8! 24. \Wxg6 \Wf4 25.<tt> fl �f8
26. \Wg4 '!Nh2 27.�g2 \Wf4 with a draw - the
active black pieces leave the opponent no
chance to use his extra pawn.
2 I .'i;Ve6t
a
8
b
c
d
e
f
g
h
Now 23 . . . �xa3 loses on the spot after
24.Wf7t <tt> h s 25.lt:Jf6.
7
6
And after 23 ... �f4 24.�g3 '!Ng8 25.lt:Jf6t
�xf6 26. \Wxf6 '!Ng7 27.\Wxg?t (the simplest)
27 ... <tt> xg7 28.�b3 �b8 29.�b6 the rook
endgame is obviously hopeless for Black.
5
4
3
2
8
1
7
a
b
c
d
e
f
g
h
2 1 . .. ®h7?
A big tactical mistake. It could cost the game.
After the correct 2 1 . .. �£7 22.lt:Je4 (22.�xe5
dxe5 23.lt:Je4 We? 24. \Wxe? �xe7 and Black
is fine) 22 ... <tt> g7 23.@h 1 \Wh4 the position
would be dynamically balanced. For example:
24.tt:Jxd6 ixd6 25.\Wxd6 '!Nxc4 26.�ae3 �af8
27.�e6 '!Nd3 28.�xg6t '!Nxg6 29.�g1 '!Nxg1 t
30.<tt> xg1 �xf2, with a peaceful outcome soon.
21 ... @g?? is much weaker because of the strong
22.�xe5! dxe5 23.lt:Je4 with a huge initiative.
6
5
4
3
2
a
b
c
d
e
f
g
h
22 JU5
So life goes on!
••
Of course, 22 . . .�xa3 23.bxa3 looks pretty
dangerous because of the strong potential
26
King's Indian Warfare
pressure against the b 7-pawn, but the computer
is not convinced: 23 . . . �f8 24.�b 1 ?! (better is
24. ltJxc5 .ih2t! 25.<;t>g2 dxc5 26.�e7t <;t>gs
27.�xf8t �xf8 28.<;t>xh2 �xf2t 29.<;t>g3 �c2
30.�e4 <;t>f7 3 1 . <;t>f4 �c3 32. <;t>es <;t>e? with a
probable draw) 24 . . . b5 2 5 . axb5 �f4+
White should have played 26. ltJg3! �e7
27.�xe5! �xe5 28.�xd6 with good drawing
chances. In fact, this was the only continuation
here.
23J�b3
23.<;t>h 1 was my opponent's suggestion in
Informant. He evaluates the position as better
for him. However, after 23 . . . �h4 it's Black
who enjoys the advantage - all his pieces are
well coordinated and the king on h 1 proves to
be more vulnerable than its black counterpart,
for instance: 24.�b3 �af8! 25.�xb7t? <;t>hs
26.�b3 �xf2 and White is hopelessly lost.
23 ...'l:Yc7 24.a5 gd8?!
Too passive. The correct decision would be
24 . . .�ff8, preparing to get rid of the annoying
queen on e6 by means of 25 . . . �ae8. In this
case I would have nothing to complain about.
In my opinion it's White who has to take care
to keep the position balanced - after all, his
king feels no more comfortable than Black's
king, and my pieces are active enough.
25J�b6?
I fully agree with Beliavsky's suggestion
2 5. <;t> h 1 !?, even though I think that after
25 . . . �£f8! 26.�g1 �f7 Black does not stand
badly at all.
25 ... gd7!
An overture to one of the best traps I have
ever set in my entire career!
26.�xc5
Falling directly into the trap. My opponent
very quickly and confidently made this and the
next moves, perhaps feeling sure that he was
on his way to winning another game against
his client. To tell the truth, it was unreal to
play over the board.
a
b
c
d
e
f
g
h
26 .. J�g7!!
The rook hides in ambush.
Both of the alternatives were losing: 26 . . . �e7
27.�xe7t �xe7 28.�xb7 and 26 . . . �g5t
27.<;t>fl dxc5 28.�xe5 �xe5 29.�xg6t <;t>hs
30.�xh6t <;t>gs 3 1 .�g6t �g7 32.d6 �d7
33.�xg7t �xg7 34.�xg7t <;t>xg7 35 .d7.
27J�xe5 'l:Yxc5!
White was counting on 27 ... �xe5 28.�xd6
or 27 . . . dxe5 28.b4, but the move I made came
as a real shock for my formidable opponent.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
27
Chapter 1 - Always Dangerous
28J�e2
Alas, after 28.!hf5 gxf5t - check! - this was
the reason for my 26th move. 29. Wh l (29.Wfl
�xc4t 30.We l E!:g l t) 29 . . . �xf2 and White
gets checkmated. Beliavsky now gets into
time trouble and loses quickly, but actually his
position is already beyond salvation.
28 ... 'i«xc4 29.'l«e4
More stubborn was 29.E!:xd6, but after
29 . . . �c l t 30.Wg2 E!:g5t 3 l . Wf3 �c4! the
white king would not escape the mating net.
b
c
d
Israel 1 999
Usually I was successful in games with Boris
Avrukh, my long-time teammate in the
Israeli national squad. It seems that I was an
"
uncomrorta
c
bl,
e opponent for him. Boris is
not only a talented and strong grandmaster
but also a well-known theoretician, and he
has written a few highly popular books on
openings. He now lives in the USA.
l .d4 tLlf6 2.c4 g6 3.tLlc3 .a,g7 4.e4 d6 5.h3
tLl bd7 6..a.d3 0-0 7..a.gs c5
29 ... 'i«cl t 30J�el 'i«cS
a
Boris Avrukh - Ilya Smirin
e
f
g
h
The attack is irresistible - the rook on b6 has
turned out to be completely out of play and
cannot help his monarch.
3 1 .'i«e3 ge5! 32.'i«cl ggSt
White resigned due to 33.Wfl �xd5.
0-1
8.lLlf3
Borya deviates from my game with Beliavsky
and keeps the pawn tension in the centre.
Objectively speaking it's hardly the best choice,
but at least it leads to relatively fresh positions.
8.d5 lLle5 was played in Beliavsky - Smirin
above.
8 ... cxd4
This leads to a version of the Maroczy
Bind that is good for Black - in my opinion
in this particular position the white bishops
are slightly misplaced. They would be better
positioned on e2 and e3.
28
King's Indian Warfare
9.tiJxd4 YNa5 10.�d2
If 1 O.�d2 I was planning:
I saw that 1 0 . . . �b6 1 1 .�e3 ( I I .ttJb3 ttJ e5
1 2.�e2 �e6 is not advisable for White)
1 1 . . . �xb2 1 2.ttJa4 �b4t 1 3.�d2 �a3 1 4.�c l
would lead to a draw by repetition. I was in a
fighting mood and declined that.
ll.i.e2
8
7
6
a
b
c
d
e
f
g
h
l O . . . ttJxe4! 1 1 .ttJxe4 ( l l .�xe4 �xd4 1 2.�xd4
�xg5) 1 1 . . . �xd2t 1 2.<;t>xd2 �xd4 1 3.�xe7
�e8 1 4.�xd6 �xb2 1 5 .�ab 1 �g7 with slightly
better chances - the king is stuck in the centre
of the board and will cause White some
trouble even without queens. After all, there
are enough pieces on the board to attack or
disturb it.
1 O.ttJf3!? ttJc5 1 1 .0-0 perhaps deserved
attention. Here the following tactical variation
is possible: 1 1 . . .�e6 1 2.�b 1 ( 1 2.�c2!?)
1 2 . . .�a6! 1 3.b3 ttJxd3 1 4.�xd3
a
b
c
d
e
f
g
h
1 4 . . . d5! 1 5 .exd5 �f5 1 6.�d2 �xb 1 1 7.�xb 1 ,
with compensation for the exchange, but Black
certainly remains no worse.
l O ... tlJe5
5
4
3
2
1
b
a
d
c
f
e
g
h
ll. ..YNc5
1 1 . . . ttJxc4? does not work here because of
1 2.�xc4 �c5 1 3.�a4!.
1 1 . . .�b6 1 2.�e3 �xb2 1 3.ttJa4 �b4t 1 4.�d2
�a3
a
b
c
d
e
f
g
h
This would probably lead to a similar
draw, because the position resulting after
1 5 .ttJb5? ttJ d3t 1 6.<;t>fl ttJxf2 1 7.ttJxa3 ttJxd 1
1 8.�xd 1 ttJxe4 1 9.�e 1 �d7 20.ttJb5 �ac8
would be clearly in Black's favour - three
Chapter 1 - Always Dangerous
pawns for a knight plus much better piece
coordination compared to the opponent's
forces. I decided to chase the d4-knight from a
different square.
his king being under pressure: 2 l .�g l f5
22.g5 �b4t 23.�c3 �h4t 24.�d2 f4 and
it's very messy.
1 6.cxb5 �e4 1 7.0-0 �b7
1 2.llJb3 �c6
By attacking the c- and e-pawns I can try
to disturb my opponent a little. If he can
finish development and place his pieces in
the centre, White's space advantage will tell,
as is often the case in such structures. This
is why I did not like 1 2 . . :�c7 1 3.tLld5 �d8
1 4.�g5 �e8 1 5 .ttJd2, when White is about to
consolidate.
13.lDd5
Avrukh could have played 1 3.tLld4, after
which a draw by repetition would be very
probable. He also shows fighting spirit and
made the more aggressive move.
13 ... llJxd5 14.cxd5?
But this is a serious mistake - Boris j ust
completely overlooked my next move.
The knight should have been taken by the
e-pawn:
1 4.exd5 �a4 1 5.�c3!
During the game I was planning to reply
with the sharp:
1 5 . . . b5?!
In fact Black had two better moves at his
disposal:
1 5 . . . �d7 1 6.0-0 (bad is 1 6.f4? ttJxc4 1 7.�xg7
tLle3 1 8.�d3 �xf4 1 9.�d4 tLlxg2t 20.�d l
�f5 2 l .�d2 �g3) 1 6 . . . �ac8 1 7.ttJd2 �xd l
1 8.�axd l g5!? with an acceptable endgame.
Or the sharper: 1 5 ...�f5 1 6.g4 �d7 1 7.f4
(here 1 7.tLld2 �xd l t 1 8.�xd l g5!, securing
the outpost on e5 for the knight looks
attractive for Black) 1 7 ... lt:Jxc4 1 8.�xg7 tLle3
1 9.�d3 �xf4! 20.�xf8 �xf8 Now Black
is playing without a whole rook, but the
opponent faces unpleasant problems with
29
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
But it would be a dubious decision after:
1 8.f4 tLld7 1 9.�xg7 �xg7
And now, let's say:
20.lt:Jd4
20.�f3 �xf4 2 l .tLla5 is also not bad.
20 . . . �xd5 2 l .�f3 �e3t 22.�h l �xf3 23.�xf3
�e4 24.�d2 �d5 25 .a4
White would have the initiative.
14 ...�a4!
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Suddenly White has no convenient way to
protect the pawn on e4.
15.0-0!
30
King's Indian Warfare
Relatively best. White gives up a pawn, but
finishes his development and may even hope to
obtain some compensation by making use of
the unstable position of the black queen on e4.
1 5.f3 f5 with the idea 1 6.exf5? �h4t or
1 5 .�c2 f5 would be weaker, at least from a
practical point of view.
1 5 ...'1Nxe4 1 6 ..tc3
19 ... f4?!
Not the best decision. I wanted to create
the threat of . . . f4-f3, but I underestimated
Avrukh's next move. I had two better ways to
continue:
a) The sharp: 19 . . . g4!? 20.f4 I was afraid of
this move during the game, but in fact I have
a strong sacrifice at my disposal: 20 . . . gxh3!
2 1 .fxe5 dxe5
8
7
6
5
4
3
2
a
a
b
c
d
e
f
g
h
1 6... g5!
The purpose of this move is to vacate g6 for
the queen and simultaneously to prevent f2-f4,
thus securing the knight in its central position.
17J�el '1Ng6 1 8.tlJd4 f5 19.-tfl
Or 1 9.lt:Jf3 lt:Jxf3t 20.ixf3 g4.
b
c
d
e
f
g
h
Black has three pawns for a piece and a
strong initiative. It's not easy for White to
find a way to solve his problems, for instance:
22.lt:Je6 ixe6 23.dxe6 hxg2 24.ixg2 e4
25.ixg7 �xg7 26.�b3 f4 27.�h l e3 28J�g l
�h8 29.ih3 �e5 30.�c3 �xc3 3 1 .bxc3 f3
and the pawns have advanced too far - White
is helpless.
Still, after the best 22.lt:Jf3 e4 23.d6! exd6
24.lt:Jh4 �g3 25 .�b3t �h8 26.ixg7t �xg7
27.ic4! hxg2 28J��ad l the position is not at all
clear: temporarily Black has five (!) pawns for
the piece, but White's forces are active.
8
7
6
b) The simple 1 9 . . . id?, fi n ishing development.
This seems to be the most practical decision:
after the logical 20.lt:J f3 (20.�b3 g4)
20 ... lt:Jxf3t 2 1 .�xf3 g4 22.�d3 ixc3 23.�xc3
El:ac8 Black preserves his extra pawn and is also
active enough to claim an edge here.
5
4
3
2
1
a
b
c
d
e
f
g
h
31
Chapter 1 - Always Dangerous
20.i.e2!
A good prophylactic move, directed against
both . . .f4-f3 and . . . g5-g4.
20 .id7 2 1 .�8
The correct idea in general, intending to
exchange the strong knight on e5 .
•••
But probably slightly more accurate would
be 2 1 .'1Wb3 �ab8 22.lL:lf3 lL:lxf3t 23.�xf3,
and thanks to the pressure on e7 and Black's
somewhat spoiled pawn structure, White has
reasonable compensation for the pawn.
2 1 . .. �xf3t 22.i.xf3 i.xc3 23.bxc3
Now White's pawn structure is not ideal
either.
25 ...i.f5 26J�b4
If 26.�e4 �f7 27.c4 Black would exchange
bishops in much better circumstances:
27 . . . �xe4 28 .'1Wxe4 �c8 and the weakness c4
can be felt. Nevertheless, it deserved serious
attention in my opinion - White would still
have some compensation.
In contrast, 26.�b3 �ac8 27.'1Wxe7 '1Wxe7
28.�xe7 �f7 cannot be advised. The resulting
endgame is very unpleasant for White - he has
the worse pawn structure and passive pieces
after 29.�e2 mf8, with ideas of 30 . . . �e7 or
30 . . . �fc7.
26 .. J�ac8 27.'flxe7
23 ... 'flf6 24J�b l b6
8
7
6
5
4
3
a
2
a
b
c
d
e
f
g
h
25.'fle2?!
My opponent did not notice that after the
exchange of bishops the black king would
become more vulnerable, otherwise he would
have played: 25 .�g4! �xg4 (25 . . . �ad8 26.�b4
�xg4 [26 ... '1Wxc3?! 27.�be4] 27.'1Wxg4 �f7
28.�be4 does not change much) 26.'1Wxg4
�f7 27.�e4! �af8 (or 27 . . . '1Wxc3 28.�xf4!)
28.c4 mh8 (28 . . .f3 29.g3) 29.�be l and
White is perfectly OK - the extra pawn is not
significant.
b
c
d
e
f
g
h
27 .. J�xc3?!
An interesting moment, which is instructive
to some extent. After making some calculations
I played this tempting move. Alas, as is often
the case, my calculations were not free from
mistakes and objectively speaking this capture
misses the chance for a big advantage.
Instead Black should have chosen the prosaic
exchange of queens: 27 . . .'\Wxe? 28.�xe7 �f7
29.�xf7 mxf7 I do not think White can hold
this. After, say, 30.h4 h6 3 l .hxg5 hxg5 32.�b3
�d3 33.g3 mf6 the difference in activity of the
pieces is too great. As we know, piece activity is
one of the major factors in endgames.
32
King's Indian Warfare
28 ..ih5?
My opponent also did not calculate accurately.
Mter 28.'tWxa7 I was going to play the pretty
28 .. Jhf3! 29.gxf3 Wc3 - a slightly unusual
kind of fork. However, after 30.�eb 1 .ixb 1
3 l .�xb 1 Wxf3 32.'tWxb6 'tWxd5 (32 . . . 'tWxh3
33.Wxd6 �e8 34.�b8=) 33.�b5 'tWd 1 t 34.@h2
h6 35 .a4 f3 (35 . . . 'tWxa4 36.Wxd6 Wxb5
37.Wg6t @h8 38.'tWxh6t with a draw) 36.'tWc6
'tWfl 37.Wd5t �f7 38 .�b8t @g7 39.'tWd4t �f6
40.�b7t @g6 4l .'tWe4t @hS 42.Wg4t @g6
43. 'tWe4t the game would end with perpetual
check. Of course this is not a forced variation,
but the outcome looks j ustified.
After 3 l ..ie2 �e1 (3 l . .. �xf2? 32.@gl ) 32.�d4
�f7 33.�e8t @g7 34.�d 1 �xd1 35 .ixd 1 @f6
36.�d8 @eS Black would obtain a technically
won position.
The most complex variations would arise after:
3 1 .�c7
For the moment preventing the f8-rook
from joining the attack via c8.
3 l . . . �xf2 32.�xa7 �c8 33.@g1
33.�xb6 .ie4 and wins.
33 . . . �fc2
33 . . . �d2 34.@h2! and it's hard for Black to
make progress.
28 ...%Vxe7 29.gxe7
Avrukh was counting on his active rook, but
Black's trumps are more important.
29 . J�cl t 30. �h2
.
8
7
a
6
5
4
3
2
1
a
b
c
d
e
f
g
h
30 .. J�fl !
The key move - the rook not only attacks
the f-pawn, but also helps to set up mating
motifs along the first rank.
3 1 .gc4
These mating threats may be seen in the
following variation: 3 1 .�xa7 �c8 32.�xb6
�ee l 33.�b8t .ic8 34.g3 �xf2#
b
c
d
e
f
g
h
34.�a3
This seems to be the most resilient. Other
options are:
34.�xb6 .ie4 still wins.
34..ig4 �d2! (a nice move) 35 .�xb6
(35 . .ixf5 �c l t 36.@h2 �dd 1 with inevitable
checkmate; 35 . .if3 �c l t 36.@h2 �f2
37.�xb6 �ff1 38.�b8t .ic8-+) 3 5 . . . ixg4
36.hxg4 f3! 37.�c6 �xg2t 38.@fl �b8
39.�c l �xg4 40.�cc7 �h4 4l .�g7t @h8
42.�xg5 �h2 We see that this endgame is
full of tactics and mating motifs.
34. . . �c l t 3 5 .@f2 �8c2t 36 . .ie2 �d2!
37.�xb6 �cc2 38.�b8t @g7 39.�a7t @h6
40.�e8 g4! 4l .hxg4
Or 41 .a4 g3t 42. @e 1 ( 42. @f3 �d4 is another
mating net) 42 ... �xd5 with checkmate soon
after 43 .... �cl t.
33
Chapter 1 - Always Dangerous
41 . . . ixg4 42.�ae7 �xa2
Transposing into a won pawn ending.
3 I ..Jlxf2 32J�xa7
This allows a nice finish, but the position
was already beyond salvation.
After the best 32.a4 f3! 33.ixf3 id3 34.�d4
�8xf3 3 5 .mgl �fl t 36.mh2 mf8 37.gxf3
mxe7 38.�xd3 mf6 39.�e3 �a l the endgame
is lost.
Alexander Grischuk - Ilya Smirin
Beersheba 2005
The main reason why I included this game is
that the book lacked examples of "Benoni­
like" games. Besides, victory over a player
as strong as Alexander Grischuk is not an
everyday event.
I .d4 ttlf6 2.c4 g6 3.ttl c3 .ig7 4.e4 d6 5.f3
0-0 6. ttl ge2
8
7
6
5
4
3
V//
/.O/
'uu U
//0 ' " ",///////
2
a
b
c
d
e
f
g
h
32 ... f3! 33.Lf3 .id3 34J3c3 �8xf3 35. <bgl
�c2!
This slightly unusual trick forces immediate
resignation - after 35 . . . �c2 36.�xc2 �fl t
37.mh2 ixc2 Black remains with an extra
bishop.
0-1
The final game of this chapter was played at
the World Team Championship on home soil.
Sadly we were not able to take any medals on
that occasion.
a
b
c
d
e
f
g
The logical move. The main advantage is
that White can decide later whether to develop
the bishop on e3 or g5 .
6... ttlbd7
There is a wide choice here: 6 . . . lt:J c6,
6 ... c5 and 6 ... c6; and after 6 ... e5 7.ig5!?
White profits from this move order - even
Fischer scored only 1 Yz points out of 4 from
the resulting position. He was young though ­
those games were played in 1 958-59. This line
has also occurred once in my practice: 7 . . . lt:J c6
8.d5 lt:Je7 9.�d2 h5 1 o.lt:Jcl lt:Jh7 l l .ie3 f5
1 2.lt:Jd3 c6 1 3.0-0-0 ( 1 3.dxc6!? bxc6 1 4.�d l )
1 3 . . . cxd5 1 4. cxd5 id7 1 5 .mb l lt:Jf6, with a
complicated struggle; (Yz-Yz, 5 1 ) I. Sokolov Smirin, Groningen 1 993.
King's Indian Warfare
34
7 ..ig5 c5 8.d5
Now the game develops in the spirit of the
Benoni structure.
8
7
6
8 ... h6 9 ..ie3
5
8
4
7
3
6
2
5
1
b
a
4
d
c
f
e
g
h
13 ... fxe4!?
Probably the best way to handle the tension
in the centre.
3
2
1
9 h5
a
b
c
d
e
f
g
h
..•
If instead 1 3 . . . lt:J f6, White would get a
positional advantage after the simple: 1 4.exf5
ixf5 1 5 .h3! ( 1 5.f4? ttJeg4)
Directed against ttJ g3. An original idea, bur
I think that the more usual 9 . . . ttJ e5 1 0.ttJg3 e6
(or 1 o . . . h5 first) would be the sounder choice.
10.liJc1
I would prefer 1 O.lt:Jf4!? ttJ e5 1 1 .ie2 and it
is not so clear what Black is going to do next
( l l . . g5?! 1 2.ttJd3 ttJxc4 1 3.ixg5).
.
10 ...tiJh7
The standard manoeuvre in such positions;
Black prepares . . . f7-f5.
1 I ..ie2 tlJe5 1 2.0-0
If 1 2.f4 lt:Jg4 1 3.ixg4 ixg4 Black would be
OK - the bishop pair is an important factor
here.
12 ... £5 13.'1Nd2
If White tries for activity in the centre 1 3 .f4!? ttJ f7 1 4.e5 - Black would obtain good
counter-chances after 1 4 . . . dxe5 1 5 .ixc5 'lfff c7
1 6.ia3 a6!?, with an interesting struggle ahead.
a
b
c
d
e
f
g
h
1 5 . . . a6 ( 1 5 . . . e6?! is not good due to 1 6.f4 lt:J f7
1 7.dxe6 ixe6 1 8.id3) 1 6.a4 'lff! a 5 1 7.E!a3 Black has insufficient dynamic possibilities to
compensate for his structural weaknesses.
14.tlJxe4 .ifS 15.tiJfl
1 5 .ttJg3 was a worthy alternative. After
1 5 . . . e6!? ( 1 5 . . . id7 1 6.ih6!? and Black's king
will have to be taken care of after the exchange
of bishops) and, let's say, 1 6.E!e l h4 1 7.ttJxf5
exf5 1 8.f4 lt:Jg4 1 9.ixg4 fxg4, a double-edged
situation would arise.
35
Chapter 1 - Always Dangerous
I S YNb6!?
20.ttJxd2 ifS may have led to a slightly more
pleasant ending for White. Grischuk is playing
for complications instead.
••.
8
7
8
6
7
5
6
4
5
3
4
2
3
b
a
d
c
f
e
g
h
1 6.a4?!
A sloppy move.
2
1
b
a
Correct was 1 6.ttJb3 a5 1 7.El:ae l a4 ( 1 7 . . . Wb4
1 8.ttJcl) 1 8.ttJcl El:f7 and the "full-contact"
struggle is yet to start.
1 6 J�ac8?
I failed to exploit White's inaccuracy.
.•
Concrete play was demanded: 1 6 . . . Wa6!
1 7.El:a2 ttJxc4 1 8.We l b5 with the idea:
d
c
e
f
g
h
I? ...YNds?!
A retreat at the wrong moment. Black should
have stuck to his principles and played the
intended: 1 7 . . . Wb4 1 8.ttJa2! (I didn't notice
this in my previous calculations) 1 8 . . . ttJxc4
1 9.Wcl ixb2 20.Wxc4
8
7
6
5
8
4
7
3
6
2
5
4
a
3
2
a
b
c
d
e
f
g
h
1 9.axb5?! Wxa2 20.ttJxa2 ttJxe3 would lead to
an excellent position for Black.
17.a5
1 7.El:a2 ib l 1 8.El:a3 Wb4 1 9.ttJb3 Wxd2
b
c
d
e
f
g
h
20 . . . Wxa5! (Nor did I see this. I certainly did
not want to play a nearly hopeless position
after 20 . . .Wxc4 2 1 .ixc4 ixa l 22.El:xal .) 2 1 .g4
hxg4 22.fxg4 ixa l 23.El:xa l (23.gxf5 ? id4)
23 . . . id7 24.We4 <j{f7 with a wild position.
Black has a material edge, but his king is
unsafe and the white pieces may become very
active. Still I believe that Black should be more
or less OK.
36
King's Indian Warfare
1 8.�a2 e6 1 9.�c3
White has managed to consolidate and has a
clear positional edge.
Of course, Alexander did not fall into the
trap: 1 9.g4? hxg4 20.fxg4 id3! 2 1 .l2Jxd3
l2Jxc4 would lose on the spot.
19 ... exd5
I did not like the prospects after 1 9 . . ltJ f6
20.ig5!?.
.
8
a
7
b
c
d
e
f
g
h
24 ..if4?
Around here my opponent started to run
short of time. Perhaps he had not noticed
some of Black's tactical possibilities.
6
5
4
3
More prudent was 24.El:fe 1 bxa6 25.l2Jc3
(25 .El:xa6? l2Jxf3t! 26.ixf3 El:xb5) 25 . . . l2Jf7
26.ixa6 Wic7, with a tense struggle.
2
1
a
b
c
d
e
f
g
h
20.cxd5?
The question mark may seem too severe
an evaluation, but after this wrong capture I
could breathe normally again.
Much stronger was 20.l2Jxd5 l2J f6 2 1 .ig5 ,
exerting strong pressure on Black's position.
20 ... �£6 2 1 .h3
With the idea of proceeding with f3-f4 or
g2-g4, but later Sasha changed his mind.
The immediate 2 1 .a6 El:b8 22.El:a3!? was
more to the point.
2 1 ..J�e8 22.a6 gbs 23.�b5
It is not clear whether White needs to play
23.f4 l2J f7. In my opinion, before seeking
concrete play it was better to make a useful
move such as 23 .El:fe 1 .
However, the strongest continuation here
would probably be the computer's suggestion
24.l2Jd1 !?, protecting the bishop on e3. But
such a move is not easy to play over the board.
24 ... bxa6 25.gxa6
It's hard to say whether 25.l2Jc3 El:b4 26.ixa6
would be a better choice - Black would still
have a promising position in that case.
8
7
6
5
4
3
2
23 ...We7!?
Aiming at the white bishops on the e-file.
a
b
c
d
e
f
g
h
37
Chapter 1 - Always Dangerous
25 ... tiJx£3t!
With this simple tactic Black takes over the
initiative.
8
7
6
26..ix£3
Of course not 26.gxf3? VMxe2.
5
26.. J�xb5 27..ixd6
Or 27.!!xd6 ltJd7! 28.ltJd1 ltJe5, with an
obvious advantage for Black.
4
3
2
a
b
c
d
e
f
g
h
29.YMf4?
Probably the decisive error.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
28.�c6?
This is a desperate move. I remember being
very surprised by it during the game.
After the simple 28.!!a2 there is nothing
terrible for White yet - his position is rather
solid. The strongest continuation for Black in
that case seems to be 28 ... VMb6 29.if4 !!b4
30.ie3 ltJd7!? with pressure on the b2-pawn
and some initiative (3 1 .g4? ltJe5 is bad) .
28 ... �xb2
It's not Black's extra pawn that matters most
now, but the activity of his pieces.
The only correct path is 29.VMa5 !!b 1 30.VMxc5
ltJd7 3 1 .VMc4 VMb2!? (or 3 1 . .. !!ee 1 32.ltJ d 1 and
apparently White can hold) 32.!!c8 !!xfl t
33.mxfl VMa l t 34.id l !!xc8 35 .VMxc8t mh?
36.g4! hxg4 37.hxg4 ixg4 38.mg2, and due
to the insecure position of Black's king, White
has excellent drawing chances. To choose the
best path in such a sharp position with very
limited time left, one must mainly rely on
intuition. In this game it let Grischuk down.
29 ... �bl
8
7
6
5
4
3
2
1
a
30J�xc5?!
b
c
d
e
f
g
h
38
King's I ndian Warfare
This third mistake in a row leads to
immediate disaster.
8
7
But even after the relatively best 30.�c4 �b2!
3 l .�xb 1 �xb 1 t 32. mh2 ltJd7 33.ig3 ltJe5
34.�xc5 ltJxf3t 35.gxf3 �fl White would be
helpless.
Or 30.�xb l �xb l t 3 l .mh2 �e l ! 32.ixc5
ltJd7? 33.id4 �b4-+ .
30 J��xfl t 3 1 . <it>xfl VNa6t 32. <it>gl
6
5
4
3
2
1
••
a
b
c
d
e
f
g
h
32 j.h6!
White resigned because of 33.�g3 h4
and 33.�xh6 :gel t 34.@h2 �xd6t 35 .g3
�xc5 . To give up one piece to capture two is
reminiscent of a combination in checkers.
0-1
•••
Chapter 2
Kramnik
Kramnik at the Dresden Olympiad 2008
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
8
7
6
5
4
3
2
6.
a
b
c
d
e
f
g
2
a
h
Kramnik missed his chance.
Can you see it? (see page 5 1 )
... 8
6
6
3
d
e
f
g
h
8
b
c
d
e
f
g
a
h
Rook endings are difficult.
What should I have played?
(see page 56/57)
8
7
5
5
5
4
2
2
b
c
d
e
'---=""'--"";;.....J
f
g
6.
h
Can you find Kramnik's nice
combination?
(see page 45/46)
d
e
f
g
h
6
4
3
a
c
8
6
3
b
White has sacrificed a piece in
this line, but how can he prove
compensation? (see page 65)
7
4
h
3
7
6
g
2
a
A hard one. How can Black
energize his pieces? (see page 44)
f
4
2
c
e
5
4
b
d
7
5
a
c
Think prophylactically.
How should Black progress?
(see page 63)
... 8
7
b
3
2
a
b
c
d
e
f
g
h
White is much better.
How should he consolidate?
(see page 60)
a
b
c
d
e
f
g
h
The moment of triumph.
How should Black continue?
(see page 66)
Chapter 2 - Kramnik
41
I n the period 1 992-2002 I played a "mini­
match" of four games against one of the
strongest anti-King's Indian players of all time.
I am happy, both on my own behalf and on
behalf of the King's Indian, that I was able to
score 2 Yz-1 Yz against this outstanding player,
who has a tremendous lifetime score against
the King's Indian.
Kramnik was a talented j unior in the late
1 980s, but not visibly more so than many
other talented players from his generation, like
Konstantin Sakaev and Mikhail Ulibin. But in
the early 1 990s his strength increased rapidly.
It was seen as risky when Kasparov wanted
Kramnik's participation as first reserve in the
Russian team for the Manila Olympiad. But
with a score of 8 Yz/9 he easily took the gold
medal on his board and helped secure gold for
Russia.
Only a year later he entered the world's top
ten. The year after that he dealt a major blow
to the reputation of the King's Indian with this
complicated game.
a
b
c
d
e
f
g
h
7... tlJ bd7
I used to play this as well, although in most
cases in the last 20 years I have placed the
knight on a6, with or without . . . a7-a5.
8 ..ig5 h6 9 .th4 g5 IO ..ig3 tlJh5 l l .h4 g4
12.tlJh2 tlJxg3 13.fxg3 h5 14.0-0 f5
A typical dynamic reaction.
.
15.exf5 tlJc5 1 6.b4
Vladimir Kramnik - Garry Kasparov
Linares 1 994
I .tlJf3 tlJf6 2.c4
This was Vladimir Kramnik's move order
in the 1 990s, eventually leading Chess Stars
to publish a series of books with a Kramnik
repertoire, starting with this move order.
2 ...g6 3.tlJc3 .ig7 4.e4 d6 5.d4 0-0 6..ie2
e5 7.d5
This move defines the Petrosian System,
named after the ninth World Champion. This
was quite popular in the 1 980s, but less so
these days. Kramnik also played it against me
in 1 995, as we shall see below.
a
b
c
d
e
f
g
h
16 ... e4!?
Typical of the King's Indian: opening up the
diagonal for the King's Indian bishop at the
cost of a pawn.
42
King's I ndian Warfare
The position after 1 6 ... l2Ja6 1 7.�b 1 ixfS
1 8.id3± is simply pleasant for White - for
example in Handke - Socko, Groningen 1 998.
8
7
6
17J�cl ltJd3 1 8 ..txd3 exd3
5
8
4
7
3
6
2
5
1
a
4
3
2
a
b
c
d
e
f
g
h
19.£6!
Kramnik has a good sense of the nature of
these positions as well, even though he is more
comfortable playing White in them. After
1 9.Wxd3 Wf6+ Black is very harmoniously
placed and the bishops will become dominant
very quickly. White's extra pawn is irrelevant.
19 .. J:!:xf6 20.�xd3
This has been tested in some games since
1 994 and the general conclusion is that White
is a bit better, but Black's position is by no
means untenable.
b
c
d
e
f
g
h
22J:l:xf5! �xf5 23.tbxc7 �c8 24.ltJe6 �f6
25.tbfl
Kramnik pointed out in his 1 994 analysis
that 25.�fl is met strongly by 25 . . . �f8!!
26.l2Jxf8 Wd4t 27.Wxd4 ixd4t 28.<tt> h 1 �xf8,
where White has no winning chances despite
his two extra pawns.
But maybe 25.�e 1 !? was a worthwhile try.
25 ... �e5 26.�dl �f5 27.�xf5 �xf5 28.c5
.tf8 29.tbe3 �f6 30.tbc4?!
This seems a bit too creative. Maybe 30.�c 1
with even chances was better.
30 ... dxc5 31 .b5
20 ...�£8 2 I .ltJ b5!?
A very ambitious idea.
2 I . ...tf5
a
3 I . ...th6?!
b
c
d
e
f
g
h
43
Chapter 2 - Kramnik
Vladimir Kramnik - Garry Kasparov
This was heavily criticized 20 years ago, but
it is not that bad. Still, Kasparov seems to have
missed White's main threat.
Critical was 3 1 . .J3:e8 32.�e l �f5! 33.�fl �xd5
34.l2Jc7 �ed8 35.l2Jxd5 �xd5, when White has
decent drawing chances, but Black should be
happy with the way things are going.
Novgorod 1 997
I .tlJf3 tlJf6 2.c4 g6 3.tlJc3 .ig7 4.e4 d6 5.d4
0-0 6 ..ie2 e5 7.0-0 tlJc6 8.d5 tlJe7 9.b4
tlJhS IO.�el!
8
32J�el!;f; �e8 33.�e5 �e7 34.�xh5 �ef7
3S.@h2! .icl 36.�e5 �fl 37.�e4 �dl
38.�xg4t
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
This idea was introduced by Ivan Sokolov
and quickly became one of the big main lines
of the King's Indian.
a
b
c
d
e
f
g
h
38 ... @h7??
A simple blunder that loses the game
immediately and shows the pressure Kasparov
has been under for most of the game.
38 . . . @h8! 39.�e4 �xd5 40.l2Je5 looks
dangerous, but after 40 . . . �f2!, with
counterplay against g2, Black would have his
chances as well.
39.tlJe5 �e7 40.tlJf8t
1-0
A messy game that should by no means make us
lose heart, but this is exactly what Kasparov did
after losing his second full-length tournament
game in the King's Indian to Kramnik a few
years later.
lO ... tlJf4 I I ..ifl aS 1 2.bxa5 �xa5 13.tlJd2
c5 14.a4 �a6 1 5.�a3 g5!?
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
A novelty at the time. Black offers a pawn in
return for White making some concessions on
his kingside.
King's Indian Warfare
44
1 6.g3
Kramnik accepts the challenge.
I6 ... �h3t 17.hlt3 .ixh3 18.YMh5 YMd7
19.YMxg5
8
�xe4 28.�d3 �xd3 29Jhd3 �b8! and Black
has enough counterplay for the pawn.
24 . . . �h5
24 . . . �f7!oo was stronger.
25.ttJd2 ttJ d4 26.ttJxd4 exd4 27.�e7?
27.ttJ e4± was probably a bit better for White.
Instead he is blown off the board.
7
6
5
4
3
2
b
a
d
c
e
f
g
h
19 ... h6?
Later it turned out that this was the
moment where Black lost the opening battle.
Subsequent games mainly took the following
direction:
a
c
b
e
d
f
g
h
27 . . . b5!!
An amazing idea, activating the rook on a6.
28.cxb5 �a7!! 29.�el
29.�xa7 �e2 and White is mated.
1 9 .. .f5!? 20.exf5
The critical line is probably 20.f3 f4oo, as in
Balcazar Novoa - Craciunescu, email 2006.
20 . . . ttJ xf5 2 l .ttJf3 �e8 22.�d2 h6 23.ttJb5
�g6 24.�c2
24.ltJc7 is met with the active:
a
a
b
c
d
e
f
g
h
24 ... e4! 25.ltJxa6 bxa6 26J�xe4 ltJxg3 27.hxg3
b
c
d
e
f
g
h
29 . . . �xf2!
The attack does not slow down.
30.<j;>xf2 ifS 3 l .�e4 �xh2t 32.<j;>e l �h l t
33.ttJfl
33. <j;>e2 �e7! and White's position is
collapsing.
33 . . .ixe4 34.�h2 �xh2 35.ttJxh2
White has managed to avoid mate, but the
ending is not much fun either.
45
Chapter 2 - Kramnik
2 5 .f3 ih5� would give Black great counterplay.
The knight comes to e5 and the white pieces
will find it hard to become active.
8
7
6
25 ... �g6 26.ggl ha t
26 . . . <;t>h8 was tried in Baeckstroem Petraitis, corr. 2002.
5
4
3
2
a
b
c
d
e
f
g
h
35 . . . �b7 36.if4 c4 37.ixd6 c3 38.ltJg4 h5
39.4Jf2 ixd5 40.ltJd3 ic4 0- 1 Goncalves Sender, corr. 200 1 .
20.�e3 f5 2 1 .�e2 f4 22.� b5 <if?h7
22 . . . ttJg6 was met strongly by 23.<;t>h 1 !
<;t>h? 24.�g 1 ± in Szeberenyi - Zimmerman,
Budapest 1 999.
23.gxf4!
The opening of the kingside turns out to be
to White's advantage.
23 ... exf4 24.<if?hl .ig4
8
7
a
b
c
d
e
f
g
h
White is much better after 27.ltJc7! �b6
28.h3!±.
27.�xf3 �e5 28.�h5 �f7 29.�h3 �xc4?!
Black wins back his pawn, bur White
is allowed to activate all of his inactive
pieces and start a great attack. The position
was dismal anyway, but this cut the game
short.
30.gf3! .ie5
6
5
8
4
7
3
6
2
5
1
a
b
c
d
e
f
g
h
25.�f3!!
The star move of the game. Kramnik enters
into a pin voluntarily, knowing it to be
temporary.
4
3
2
1
a
b
c
d
e
f
g
h
46
King's Indian Warfare
3 1 .�c7!! gxa4
3 1 .. Y!Jxc7 32.�xh6t leads to mate.
32.Lf4!
1-0
Kasparov resigned on account of: 32 ... .ixf4
33.tDe6 :gg8 34.:gxg8 �xg8
8
7
6
My Games
Vladimir Kramnik - Ilya Smirin
Moscow 1 992
This game was played in the Alekhine Open.
In 1 992 Vladimir Kramnik was not yet such
a formidable player, but he was already one of
the strongest in the world. Playing him with
Black was a serious test.
l.d4 �f6 2.c4 c5 3.d5 g6 4.�c3 .ig7 5.e4
0-0 6.�f3 d6 7.h3
5
4
3
8
2
7
a
b
c
d
e
f
g
h
3 5.�f5t �h8 36.�f6t �h7 37.tt:Jf8t
6
5
4
3
2
1
a
b
c
d
e
f
g
h
7... e6
Actually this is a mix of the King's Indian
and Benoni openings. Still I think it's logical to
consider this opening as a King's Indian - the
pawn structure is similar and the bishop is on
g7 after all!
On a few occasions I have played a slightly
different version here: 7 . . . tt:Ja6 8.id3 tD c7
9.0-0 e5. It's not a big difference - Black is
still worse. The position is somewhat cramped
and White's forthcoming offensive on the
queenside may be very unpleasant. Yes, Black
has possibilities for active play on the opposite
wing, but they are not easy to carry out. His
practical chances should not be underestimated
though, and Black's position is solid.
47
Chapter 2 - Kramnik
s ..id3 tlJa6
I used to play this system sometimes.
The main (and better) continuation here
is 8 . . . exd5 9.cxd5 (or 9.exd5 �e8t 1 0.ie3)
9 . . . b5 1 0.ixb5 (or l O.ltJxb5) 1 0 . . . ltJxe4 with
a lot of theory to follow. I was never fond of
entering this theoretical discussion and usually
preferred the move in the game, even though it
is objectively rather dubious, as I said.
9 .ie3 was played against me by the very
talented and almost forgotten player Leonid
Zaid: 9 . . . ltJ c7 1 0.e5 (this leads to a sharp
battle) 1 0 . . . dxe5 1 1 .ixc5 :ge8 1 2.d6 ltJa6
1 3.ia3 ltJd7 1 4.0-0 f5 1 5.ie2 b6!? 1 6.ltJd5!
8
7
6 l r.JI\ �A'%
5
4
3
r�" " """
'"/
2
a
b
c
d
e
f
g
h
1 6 . . . ttJ dc5 1 7.ltJe7t �xe7 1 8.dxe7 �xe7 1 9.b4
ltJd7 20.c5 e4 2 1 .ltJd4 ltJe5 22.�c l ltJc7
23.ltJb5 ltJd5 with good compensation for the
exchange; (0- 1 , 39) Zaid - Smirin, Israel 1 99 1 .
9 ... h6 10 .ie3 tlJ c7 l l .Y:Yd2
A standard manoeuvre: White gains a tempo
by attacking the pawn on h6.
.
l l . .. <it>h7 12.0-0
a
b
c
d
e
f
g
h
1 5 .ic2! ltJf4 1 6.ia4 cxb4 1 7.ixd7 �xd7
1 8.axb4 f5 1 9.exf5 �xf5 20.ltJe4, with an
overwhelming positional advantage; ( 1-0, 42)
Beliavsky - Ilya Smirin, Moscow 1 988.
48
King's Indian Warfare
1 4.a3 aS etc. Today I would probably choose
this livelier structure, despite the knight on c7
being placed slightly passively - it would look
better on d7. But how to carry out . . . b7-b5
then?
13.�h2 llJh5 14.g3
Keeping the knight out of f4.
14 ... a6
This is not the time to start play on the
kingside: 1 4 . . . f5 ? 1 5 .exf5 gxfS
8
Prophylactic defence of the pawn on h3.
Meanwhile, Vladimir is planning to regroup
this knight to e2. Also, the pawn on f2 may
move now - either one or two squares - if an
appropriate moment arises. But this is quite a
distant prospect.
1 6 ....id7 17.aS b6
With the slim hope that White will not take
on b6 and this pawn will go further. Alas . . .
1 8.axb6 gxb6 19. llJ ge2
8
7
7
6
6
5
4
5
3
4
2
3
a
b
c
d
e
f
g
2
h
1 6.g4! e4 1 7.ltJxe4. Black has no choice but
to try to do something on the queenside - the
area of White's supremacy.
8
7
1
a
b
c
d
e
f
g
h
19 ... £5
It's not clear which side will benefit more
from this advance, but Black has to do
something.
The alternative 1 9 . . . 'W'c8 20.g4 tt:Jf6 2 1 .cj;>g2
hS 22.:gh 1 ! seems to be too passive.
20.exf5 gxf5 2 1 .gab1
White protects the b-pawn and implies that
b2-b4 will become possible.
6
5
4
Black would welcome 2 l .g4 e4 22.gxh5 exd3
23.'W'xd3 :gxb2, with real counterplay.
3
2
a
16.llJg1!?
b
c
d
e
f
g
h
2I. .. llJf6
2 1 . . . f5-f4 is impossible due to the pin, and
this advance cannot be prepared by means of
2 1 . .. cj{ h8? because of 22.ixh6.
49
Chapter 2 - Kramnik
I did not want to play 2 1 . .. e4 22 . .ic2 either it is not clear how to develop an initiative, and
meanwhile some weaknesses (such as the f4square becoming available to the white knight)
have been created.
22.f3
Just in case, Vladimir deprives Black of a
possible . . . e5-e4.
22 ... h5!?
Black made this move to fi n ally escape the
unpleasant pin along the b 1 -h7 diagonal.
23.�c2 @gs 24.� dl
This was the better option - Black's king
becomes more open, but his pieces get some
oxygen and the pawn on f4 offers hope of
doing something on the kingside. After let's
say 28.ltJec3 (28.b4 .if5 is unclear) 28 . . . .if5
29.E!g5 ltJh7 White probably has a positional
edge, but Black is not without counter-chances
and the situation remains tense.
25.b3 %YeS
All of Black's pieces are now well-placed,
with one notable exception - the knight on c7.
As Siegbert Tarrasch used to say: "If one piece
is bad the whole position is bad." This applies
to this game to a certain extent.
26J�gl!?
Another prophylactic move "just in case".
�.!1•r
� !�'
- :-�- -��
�.Lr•• � •
�-�"��
Y,� �� "��i'i!��
� �-,.�,�i
��-lLS� �--
6
5
4
JY£0
"
x
Now White is ready to start the queenside
attack after an eventual b2-b4. I decided to
stop it physically for a while.
24 .. J�b4?!
24 . . . h4!? 25.gxh4 f4 26.if2 'Wc8 27.E!g1 .ixh3
8
a
b
c
d
e
f
g
h
26... a5
Intending to play . . . a5-a4 to undermine the
white pawn chain on the queenside.
Intuitively I did not want to play 26 . . . e4 here.
The following sample variation shows that
White remains on top: 27.ltJf4 exf3 28.'\Wf2!
a5 29.�xf3 a4 30 ..id2 axb3 3 l ..ixb3 E!b7
32 ..ic2 E!xb 1 33 ..ixb 1 ltJe4 34 . .ixe4! fxe4
35 .�e2 and despite the simplification, Black
is worse - the pawns on h5 and e4 are about
to fall.
7
6
5
4
3
2
a
b
c
d
e
f
g
h
50
King's Indian Warfare
27 ..ih6!
White wants to exchange the g7 -bishop, the
main potential defender of Black's king.
We may conclude that Black has been
outplayed - he has weaknesses on both wings
without real counterplay.
27.'Dcl deserved some attention as well White is going to play 'D d3 or 'Da2 in order to
win the aS-pawn after the rook retreats from
b4. The following line looks logical and almost
forced: 27 . . . a4 28.'Da2 gb8 29.bxa4 gxb l
30.ixb l ixa4 3 1 .ixf5 'D fxd5 32.ie4 'D f6
and in the event of 33.�xd6 'Dxe4 34.fxe4
'D e6 35.'Dac3 White remains the stronger side;
but, as is often the case in the King's Indian,
Black has some compensation and reasonable
practical chances for the pawn.
30 ... e4
This kind of desperate move (the black king
now becomes really weak) could have led to
a loss by force. Still it was not a bad decision
- after all, I managed to draw the game! I
was eager to obtain counter-chances at any
price.
27 ...'1Mg6 28.bg7 '1Mxg7 29.ltH2
Mter the exchange of dark-squared bishops
the unfortunate placement of the knight on c7
is felt even more. Black has to hurry to really
include it in the battle. The question is: "Can
he?"
29 .. J!a8
From a practical point of view the following
pawn sacrifice deserved some consideration:
29 . . . �h8!? 30.'Dd3 gb7 3 1 .�xa5 gg8, trying
to fi s h in troubled waters. But objectively the
compensation is clearly insufficient. Moreover,
I did not yet feel that Black's position was so
dangerous strategically.
30J!al!
If Black had proceeded instead with
30 . . . a4, trying to open the queenside, then
after 3 1 .'Dd3 gbb8 32.bxa4 e4 (the only real
try) 33.fxe4 fxe4 34.'D df4 gb4 3S .ggb l ! (the
key move) 35 . . .gxc4 36.gb7
a
b
c
d
e
f
g
h
. . . the rook would penetrate with devastating
effect, for instance: 36 . . . 'Dg4t (36 . . . 'D fe8
37.ga2 ic8 38.gb6 would prolong the game,
but Black's pieces work as single units - there
is no coordination between them whatsoever.
His position will soon fall apart) 37.hxg4
�xa l 38.gxc7 �b2 39.'Dxh5 gxc2 40.�g5t,
with a complete rout.
31 .fxe4 l!Jxe4 32.lbxe4 fxe4 33.l!Jf4 '1Md4!?
This and the next move is the only option
to complicate White's task. Passive defence did
not offer any hope.
34.'1Me2 a4 35.'1Mxh5
a
b
c
d
e
f
g
h
51
Chapter 2 - Kramnik
35.lt:Jxh5 would lead to a winning position
too, but the move made by Kramnik is more
natural and no less strong.
40 'lMf8 41 .'lMxd7
Now 4 1 .Wg5t would lead to a draw threefold repetition.
35 ...axb3 36J�xa8t llJxa8 37.'lMg5t 'lMg7
41. bxc2
.•.
•.
8
8
7
7
6
6
5
5
4
4
3
3
2
2
b
a
d
c
f
e
g
h
38.'lMd8t
Vladimir did not find the forced win:
38.lt:Jg6! bxc2 (other moves don't save Black
either) 39.Wd8t i>h7 (39 . . . i>f7 40.lt:J h8t)
40.lt:Jf8t i>h6
8
a
b
c
d
e
f
g
h
42.'lMg4t
Finally missing the win. It was not trivial
though: White must first play 42.Ek l ! El:xc4
and only after that 43.We6t Wf7 44.Wxd6.
Black cannot hold this - j ust look at the knight
on a8!
42 'lMg7 43.'lMc8t 'lM£8 44.'lMe6t 'lMf7
45.'lMxe4
Now 45 .Wxd6 El:b l ! would lead to a draw ­
the c-pawn is too strong.
7
•..
6
5
4
3
8
2
7
a
b
c
d
e
f
g
h
4 1 .Wh4# Truthfully, I also failed to see it - it
was not so easy to spot in mild time trouble.
38 ...'lMf8 39.'lMg5t 'lMg7 40.'iMd8t?
Kramnik could still have won as shown
below, but instead he preferred to repeat the
position.
6
5
4
3
2
1
a
b
c
d
e
f
g
h
52
King's I ndian Warfare
45 ... �b6!
This knight was bad on c7 and miserable on
a8, but now it proudly joins the battle.
46JWxc2 �xc4 47.�c3 �e5
Suddenly all of Black's pieces become very
active.
48J�al �b7 49.�a3
a
b
c
d
e
f
g
h
Vladimir offered a draw, which I accepted.
Black is not worse, but he has no real grounds
to play for a win - his king is too exposed. The
following logical variation 49 . . . gb2t 50.�h 1
c4!? 5 1 .Wxd6 gb 1 t 52.gxb 1 Wxb 1 t 53. �g2
We4t 54.�h2 (the only move) 54 ... tDf3t
55.�h 1 leads to a draw by force.
1!2-lf2
Vladimir Kramnik - Ilya Smirin
Moscow 1 995
This game was played in a short (two games)
match with a rapid time control. The match
was a part of the excellently organized
Professional Chess Association (PCA) cycle
of tournaments which took place in cities like
New York, Moscow, Paris and London. The
venues for these events were also very good,
such as the Javits Convention Center in New
York. The PCA was founded in 1 993 and
ceased to exist in 1 996. Pity! Those events had
a 1 6-player knockout format, among them
1 0 invited personally and six qualifying from
Swiss-system tournaments. The invitees, such
as Kasparov, Anand, Kramnik and Ivanchuk,
were the strongest players in the world. The
qualification tournaments were also high­
level events. I managed to qualify four times
in a row, which I consider one of the main
achievements in my career. On the fifth
occasion I was invited directly to the final (in
New York in 1 995) . The World Champion,
Carry Kasparov, was the main force in creating
the PCA, and (alas) he was also responsible
for its death. Once again, it's a pity that it all
ended after three years, but at least those years
were very productive and left their mark on
modern chess history.
1 .�8 �f6 2.c4 g6 3.lLlc3 .ig7 4.e4 d6 5.d4
0-0 6.i.e2 e5 7 .d5
Again the Petrosian System.
7...a5 8.h3
A relatively rare continuation in this
particular position. Often the game transposes
to the 6.lD f3 0-0 7.h3 e5 8.d5 line.
8.ig5 would lead to the "pure" Petrosian
System.
8 ... �a6 9.i.g5
53
Chapter 2 - Kramnik
Black transfers this knight to c5 and
implements the . . . f7-f5 push after that. The
main negatives are a certain loss of time and
the non-ideal placement of the knight on a6.
Vassily lvanchuk chose a different route for
the knight in his encounter with Kramnik from
the same tournament: 1 o . . . lt:Jh5 1 1 .CtJd2 lt:J f4
1 2.ifl ctJc5 1 3.CtJb3 b6 1 4.g3 lt:Jh5 1 5 .ctJxc5
bxc5 1 6.'1Wd2 @h? 1 7.ie2 CtJ f6 1 8.0-0-0 '1We7
1 9.g4 CtJd7 20.h4 lt:J b6 2 1 .h5 g5 ; (Yz-Yz, 35)
Kramnik - lvanchuk, Moscow 1 995.
9... h6
Later in this position I preferred to avoid
or to postpone this move, which somewhat
weakens Black's kingside formation. A relatively
recent example is 9 . . . lt:Jc5 1 0.ctJd2 id? 1 1 .h4
h6 1 2.ie3 h5 1 3.f3 c6 ( 1 3 . . . a4!?) 1 4.CtJb3
cxd5 1 5 .ctJxc5 d4 1 6.lt:Jxd7 '1Wxd7 1 7.ctJa4
( 1 7.ixd4!? exd4 1 8.'1Wxd4±) 1 7 . . . dxe3 1 8 .ctJ b6
'1Wc7 1 9.ctJxa8 El:xa8 20.'1Wd3 ih6 2 1 .g3 lt:Jd7
22.Ei:d 1 Ei:a6 with good compensation for
exchange; (Yz-Yz, 4 1 ) Cramling - Smirin,
Stockholm 20 1 2.
IO ..ie3
l l .Y*fd2 @h? 1 2.g4 �deS 13.0-0-0 f5
A risky continuation. Now the black king
becomes more open, but there was nothing
better.
14.exf5 gxf5
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
I SJ!dgl !
A strong move. The target - the black king.
1 5 ...Y*fe7
It's better for Black to refrain from moving
the central pawns: 1 5 . . . e4 1 6.ctJe1 '1Wf6 1 7.g5
hxg5 1 8.ixg5 '1Wd4 1 9.'1Wf4 and the king
is under attack, or 1 5 .. .f4? 1 6.ixc5 ctJxc5
1 7.'1Wc2t @g8 1 8.lt:Je4 and Black is strategically
lost - his bishop on g7 is firmly blocked and
White controls the vital light squares on the
b 1 -h7 diagonal.
54
King's Indian Warfare
1 6.gxf5
Not the best choice, helping Black to activate
the c8-bishop. After 1 6.g5! f4 1 7.ixc5 'Dxc5
1 8 .id3t if5 ( 1 8 . . . e4 1 9.'Dxe4 'Dxe4 20.'\Wc2
if5 2 1 .'Dh4 �ae8 22.�e l ±) 1 9.ixf5t �xf5
20.gxh6 if6
d3 and threatening to play 2 l .ih5 (or ig4 in
some cases) deserved serious attention.
20 ...e4 2 1 .�el .ieS!
Of course 2 l . . . 'Da6? is out of the question.
The knight on b4 is a kamikaze at the moment.
22J�e3
22.axb4? axb4 in turn is also out of the
question.
8
7
6
5
a
b
c
d
e
f
g
h
2 1 .�g4!? <;t>h8 22.�c2 e4 23.<;t>b l White's
chances should be preferred - the opponent's
king feels unsafe.
16 ....bfs 17J�g3 �b4 1 8 ..ixc5
Of course not 1 8.�hgl ? 'Dxa2t.
8
7
6
5
4
3
2
b
c
d
e
f
g
3
2
a
1 8 ... dxc5 19J�hgl �H7
More precise was 1 9 . . . .if6 20.a3 e4, but this
was hard to determine during a rapid game.
a
4
h
20.a3
20.'Del !?, providing extra cover for c2 and
b
c
d
e
f
g
h
22 ...YMh4?
The wrong step.
I should have played 22 ... if4, resulting in a
sharp and completely unclear position with
both kings in potential danger. In the event of
the brave and greedy 23.axb4?! axb4 24.'D b 1
�a2 25.ig4 .ig6 26.'Dg2 ie5 27.�b3 b5!?00
White's extra piece doesn't immediately make
its presence felt.
23.�bl?
Too soft.
After the energetic 23 ..ig4! ixg4 (23 . . .if4
24.axb4 [now it's time!] 24 . . . axb4 25 .ixf5t
�xf5 26.'Dxe4) 24.�xg4 �xf2 25.'Dxe4 �xd2t
26.<;t>xd2 ixb2 27.axb4 axb4 28.'Df3 �a2
29.@d3 White would have a clear advantage
in the endgame - the compensation for the
piece is clearly insufficient.
Chapter 2
-
55
Kramnik
3 1 .tLlg4
23 ... tLla6
It's time to retreat!
8
24 ..ig4 .ig6
7
8
6
7
5
6
4
5
3
4
2
3
1
a
2
a
b
c
d
e
f
g
h
25.gg2
Vladimir decides to part with the exchange.
Indeed, he will have clear compensation,
mostly because of the poor coordination of
Black's pieces. In particular, the knight on a6 is
currently searching for something at the edge
of the board, far from its colleagues. But if it
manages to become useful without making
substantial concessions, the extra exchange will
start to become significant.
In my (and the computer's) opinion 25.lDdl
was a better way to give up the exchange after
25 . . . if4 or 25 . . . ih2!? 26J�g2 if4.
25 ....if4 26.tlJc2 gam 27.tiJdl .ixe3
28.tLlcxe3 b6
Not forgetting about the pawn on aS.
29.<tt> a2 tiJb8
The knight is coming closer!
30 ..ie6 ge7
Black has managed to stabilize the position
and with the extra exchange may look to the
future with hope.
b
c
d
e
f
g
h
3 1 . .. tiJd7
Mter the centralizing 3 l . . . tDc6! I would have
been on the verge of winning - the knight will
go to d4 (could it have dreamed about that just
seven or eight moves ago?) .
32.tiJc3 YNg5 33.'iNel tlJ f6 34.%Vgl tLlxg4
35.gxg4 YNf6 36.%Vg2?
Correct was the simple 36.ttJxe4 ixe4
37J:he4 gg7 38.Wb l �h8 39.Wc l and White
is OK - the bishop on e6 plus a pawn are not
inferior to the black rook.
8
7
6
5
4
3
2
a
b
c
d
36 ... gxe6?
This was unnecessary.
e
f
g
h
56
King's Indian Warfare
36 . . . �g7 37.lLlxe4 Wf3-+ would have led to
a winning position - there is no compensation
any more.
8
37.dxe6 Wxe6 38.'it>al
The smoke has cleared. Black has an extra
pawn and an obvious advantage.
7
38 ... e3?!
The most rational approach was 38 . . . �f3
39.�g3 �d3, with j ust a few technical
difficulties remaining, but adrenaline drove
me forward!
4
39.fxe3 .id3 40.Wd5!?
40.�g7t mhs 4 L�xc7 We5 42.�b7 �fl t
43.ma2 ixc4t 44.b3 Wxc3 45 .�b8t mh7
46.We4t mg7 47.Wg4t mh7 48.We4t with a
"simple" perpetual.
40 ... Wxd5 4l.cxd5 go 42.e4 gxh3 43.e5
ge3
Vladimir had little time left; I was better in
this aspect.
6
5
3
2
1
a
b
c
d
e
f
g
h
50.gxd6?
It seems that in time trouble Vladimir lost
his usually cool state of mind.
The other capture 50.�xh6 �d 1 (50 ... dxe5
5 l .�xb6 e4 52.mb3 should be drawish as well)
5 l .exd6 md7 etc. would lead to a simple draw
in the "three versus two" rook ending.
50 ... h5 5 l .gxb6 gxe5
The rook ending is on the board! All the black
pawns are separated, but the extra h-pawn is a
dangerous passer.
8
7
6
52. 'it>b3 i> f7 53. 'it>a4
5
4
3
2
a
b
c
d
e
f
g
h
44.d6!
Correct. This endgame activity should have
led to a draw.
44 ... cxd6 45.ltJd5 gel t 46.'it>a2 .ig6
46 . . . �xe5?? 47.lLlf6t mh8 48.�g8#
a
b
c
d
e
f
g
h
57
Chapter 2 - Kramnik
53 ... ®g7?
With just three pawns left, one should take
more care not to give up any of them easily.
Correct was 53 ... c4 54.Ek6 �g5 ! (I did not
see this move) and Black is close to winning
with the white king stuck on a4 - 5 5 .�c7t
(certainly not 5 5 .�xc4? �g4) 55 . . . \t>e6 56.Ek8
�d7 57.�h8 �c5 and so on.
8
7
6
5
4
3
2
54.®xa5 h4
1
a
8
b
c
d
e
f
g
h
62 ... ®g2?
I needed to cool down here (as I remember,
I still had more than one minute) , take a deep
breath and play 62 . . . �g l 63.�xh2t �xh2
64.�b5 �cl 65 .a5 \t>g3 66.a6 c4, when White
should resign.
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
55J�b7t?!
From now on the lack of time started to tell
more and more.
55 .�d6 h3 56.�d2 �h5 57.�h2 �f6 58.a4=
55 ... ®g6 56J�b6t
Time trouble. . . Why push the black king
toward supporting his passed pawn? 56.�d7
would draw easily.
56 ... ®g5 57J�b8 ®g4 58.a4 h3 59J�g8t?
This should have been the losing mistake.
59.�d8 h2 60.�d l �f3 6 l .�b5 was still
enough for a draw.
59 .. J3g5 60J3d8 h2 6I J3dl ®h3 62J�hl
63J3xh2t ®xh2 64. ®h5
Now the rook is placed slightly differently
(not stopping the a-pawn from behind) and it
means that the win has j ust slipped away.
64... ®g3 65.a5 c4t 66.®xc4 gxa5 67.b4
gh5 68.h5 ®f4 69.h6 ®e5 70.h7 ghs
71. ®c5 ®e6 72. ®c6
Very soon only two kings will remain on the
board, so we agreed a draw. An interesting game
with lots of ups and downs. Unfortunately, the
last mistake was mine.
1!2-If2
59
Chapter 2 - Kramnik
Vladimir Kramnik - Ilya Smirin
Belgrade 1 999
I would not have included this game if it
had not been one of my four games against
Kramnik. But it is, so here it is. Besides, his
play in the opening phase of the game may
serve as an example of neutralizing Black's
counterplay in the Bayonet.
I .llJf3 llJf6 2.c4 g6 3.lDc3 .ig7 4.e4 d6 5.d4
0-0 6..ie2 e5 7.0-0 lDc6 8.d5 lDe7 9.b4 aS
IO ..ia3 b6 l l .bxa5 lDh5
8
1 6.a4!
An impressive piece sacrifice; White has
three pawns for it, one of which - on c7 - is
incredibly annoying for Black.
1 6 . . . ia6 1 7.ltJb3 ixb5 1 8.cxb5 iWxc7
1 9.ltJxa5 iWxa5 20.g3 ltJxe2t 2 1 .iWxe2
White won very convincingly in Pelletier
- Nakamura, Skopje 20 1 5. Yannick Pelletier
revealed after the game that he had been
waiting for several years to implement this
idea.
1 2 ... £5 13 ..ih4!
This looks stronger than 1 3.ltJd2, as was
played by Bareev against me in the game on
page 205.
7
13 ...bxa5 14 ..ia3
6
5
4
3
2
1
b
a
d
c
e
f
g
h
12J3el
A serious blow to this line was delivered by
the following recent game:
1 2.ltJd2 tt:Jf4 1 3.axb6!
The novelty that possibly kills this particular
variation.
13 .. J�xa3 1 4.ltJb5 :gaS 1 5 .bxc7 iWd7
a
b
c
d
e
f
g
h
14... llJf6
After this game I switched to 1 4 ... tt:J f4, but
also without much success - see the annotations
to the game Bareev - Smirin page 205.
15 ..id3!
A strong positional move. Now the pawn
e4 is firmly protected in a convenient way and
Black has problems finding counterplay.
I was hoping for 1 5 .ltJd2 ih6!, with a position
to my taste.
a
b
c
d
e
f
g
h
King's Indian Warfare
60
1 5 ... fxe4 16.ll:)xe4 .ig4
I could not find anything better; nor do I see
anything now.
24.�fl !
This strong prophylactic move takes away
even a hint of counterplay and Black is doomed
to passive defence.
24 ...'1Me7 25.h3 .ixf3 26 ..ixf3 .i£8
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
This move defines White's strategic advantage
- he has more space, he has more active pieces
and he has a plan, connected with the usual
c4-c5. Black, alas, has none of that. He may
just hope that his position will be solid enough
to survive - surely not the dream of those who,
like me, are loyal servants of the King's Indian.
18 ... ll:)f5 19.'1Md3 �f7 20.�ab l .ih6 2 1 .c5
According to plan.
A slight hint of counterplay against the f2pawn.
22.cxd6 cxd6 23.�b6 �d8?!
Too passive. Also, the a-pawn is now left on
its own.
Correct was 23 . . . ixf3 24.�xf3 �d8 25.El:eb l
'Lld4 26.�d3 ifS and Black may count on a
successful defence - at least he has a central
knight on d4 and the d-pawn is protected for
now.
a
b
c
d
e
f
g
h
27.g3
Not the most accurate move.
After the precise 27.ie4! (this move is useful
in any case, and White can carry out the plan
with g3, h4 and so on later) 27 . . . �e8 28.El:fb l
�a4 29.El:c6 Black i s left with all his problems
while White controls the important b- and
c-files.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
61
Chapter 2 - Kramnik
29 .. Jk8?!
Much stronger was 29 .. J�b8 30.Ek 1 �d7
(attacking the pawn on h3!) 3 J.lit>g2 �a4 and
suddenly Black seems to be more or less OK.
For example: 32.h4 (32.Ek4 �b5)
31 .. J:k5
3 1 . .. ig7 32.ic3 El:c5 33.El:b 1 �f8 34.El:b2±.
32 ..ixd4 exd4 33.%Vxd4 �b5
8
7
6
5
4
3
2
a
b
c
d
e
f
g
1
h
32 . . . lLl b3! 33.ixg6 hxg6 34.�xg6t ig7
35 .El:b6 El:bf8 36.El:xb3 El:f6! 37.�c2 El:xf2t
38.�xf2 El:xf2t 39.<it>xf2 �d4t 40.<it>g2 �xd5t
and in the resulting position the queen is not
inferior to the pair of rooks.
8
7
a
b
c
d
e
f
g
h
34.�c6?
But this is the wrong move.
White should have exchanged his passive
rook on fl : 34.El:b 1 ! ig7 35 .�d2 and White
should win - the a-pawn will fall or Black's
king will be under attack.
34 ... �b4 35.�c4
35 .�d3 El:b2 would lead to a similar position
to that which arose in the game.
6
5
4
35 ....ig7 36.%Vd3 �b2 37.�c2 Wb6 38.�c8t
.ifS 39.�c2 .ig7
3
2
8
1
a
b
c
d
e
f
g
h
30.Wg2
My hopes were connected with 30.El:c 1
�xc 1 t 3 1 .ixc 1 El:xc 1 t 32.<it>g2 El:c2. Vladimir
avoids that, of course.
30 ...%Vd8 3I..ib2
White has regained control - there is not
much that Black can do.
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
62
King's I ndian Warfare
Black's heavy pieces are quite active - White
has serious problems converting his extra
pawn.
40 ..if3?!
I believe that the immediate 40.h4 would
offer better winning chances. 40 . . . .id4? would
be bad due to 4 l ..ixg6!.
40 ....id4 41 .h4 <i>g7 42J�d2
42.h5!? deserved attention, with the idea
42 . . . gxh5?! 43.�h l .ixf2? 44.�xh5 h6 45 .Wfd2.
42 .. J�xd2 43.Y;Vxd2 Y;Vb4 44.Y;Ve2 Y;Vb2
Black is chasing the white queen and
Kramnik has nothing better than to exchange
it.
8
7
6
5
4
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
5 l .�e2
Vladimir spent some time considering
the attempt 5 l ..ig4!? �xf2t 52.@g3 �xa2
53 . .ie6t @f8 54.�f7t @e8 5 5 .�xh7. In
the end he decided against it. After 55 . . . �al
(55 . . . a4? 56.h5 gxh5 57.g6 .id4 58.g7 .ie5t
59.@f3 .ixg7 60.�xg7 a3 6 l .�a7) 56 ..if7t
@f8 57 . .ixg6 a4 58.h5 �gl t 59.@f4 a3
60.�h8t @g7 6 l .�a8 �fl t 62.@e4 �e l t the
game would be drawn anyway.
51 ... �xe2 52 ..ixe2 <t!/g7
The draw is evident. The final moves were
made by White out of inertia.
3
2
1
a
b
c
d
e
f
g
h
It seems that in this ending with bishops of
opposite colours Black may hold - White can
hardly avoid the exchange of rooks due to the
pressure on f2.
46 ....id4 47.�e4 .ic5 48.g4 �b7 49.g5
49.�e2 .id4 50 . .ie4 �b2 does not change
anything.
53 ..ih5 h6 54..ie8 hxg5 55.hxg5 .ih4
56.<i>f3 .id2 57.<t!/g4 .iel 58.£4 .id2 59.<i>f3
.icl 60.<i>e4 .id2 61 .<i>f3 .icl 62.<t!/g4 .id2
63.£5 gxf5t 64.<i>xf5 .ih4 65.ci>e6 .ic5
66.<i>e7 .ib4 67.ci>d8 .ic5 68.ci>c7 .ih4
69.<i>h6 <tt> rs ?o ..ihs <i>g? ?L<i>hs <tt> rs
72 ..idl <i>g7 73 ..ic2 ci>f8 74. <i>a4 <i>g7
75.a3 .ic5
•!2- l/2
63
Chapter 2 - Kramnik
Vladimir Kramnik - Ilya Smirin
Russia - ROW (rapid) 2002
This game is memorable for me for two
reasons: it is my only win over a reigning World
Champion (albeit in a rapid game) and the
event in which it was played was a very special
one. The match Russia - Rest of the World was
the third such competition in the history of
chess. In the first two, in 1 970 and 1 984, the
Soviet Union team (which no longer existed in
2002 and was replaced by Russia) had beaten
the World team by 1 and 2 point margins
respectively. But this time the ROW team won
with a 52-48 score (the rapid time control
format allowed so many games to be played) .
Each team consisted of 1 0 players plus a few
reserves. It was an honour for me to be a part of
the World team. The Russian team line-up was
exceptionally strong: 3Ks (Karpov, Kasparov,
Kramnik) plus such players as Morozevich,
Grischuk and Svidler. Our team was also not
so bad - Anand, lvanchuk, Gelfand, Shirov etc.
I think it was the strongest team versus team
chess competition at least since 1 970 - the year
of the original "Match of the Century". The
Kremlin Palace of Congresses in Moscow was
the venue for the match.
The so-called Bayonet Attack - the most
straightforward way of developing a queenside
initiative.
9... �h5
Another popular line is 9 . . . a5, trying to play
on "White's" part of the board.
IOJ�el
This is Vladimir Kramnik's pet line. His
results in it are very impressive - for example,
I will mention below his win against Carry
Kasparov in his prime years.
IO ... a5
The second most popular move in this
position. The first is 1 o . . . f5, after which play
usually goes 1 1 .tLlg5 lLlf6 1 2.i.f3 (or 1 2.f3) .
l l .bxa5 f5
This is a rather risky move. I played it mainly
for practical reasons, hoping that my opponent
would not be too familiar with this line.
The main continuation is 1 1 .. Jha5 1 2.lLld2
lLl f4 1 3.i.fl c5 1 4.a4 (including Kramnik's
win over Kasparov after a transposition of
moves - see page 4 1 ) .
l .d4 �f6 2.c4 g6 3.�c3 .ig7 4.e4 d6 5.�f3
0-0 6 ..ie2 e5 7.0-0 �c6 8.d5 �e7 9.h4
a
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
think the resulting semi-closed posmon
slightly favours White, mainly because the
knight on f4 is somewhat misplaced. Often
it goes back to h5 and f6, losing time in the
process.
64
King's I ndian Warfare
1 2.tlJd2
1 2.c5!?
emotionally attached. As compensation he
gains the central pawn on e4.
1 2 ... ttJf6 13.c5
In 200 1 Radjabov played 1 3.f3 !haS
1 4.ttJb3 !!a8 1 5 .c5 against me. You will find
this game on page 242.
17.'l;Yxd2 fxe4 1 8.�b5
8
7
13 .. J�xa5 14.cxd6 cxd6 15.a4!?
1 5 .ltJ c4 !!a6
6
5
8
4
7
3
6
2
5
1
a
4
b
c
d
e
f
g
h
1 8 ...�£5
This move was asking to be played - Black
develops the bishop while protecting his extra
pawn - but now I consider it to be inaccurate.
3
2
1
a
b
c
d
e
f
g
h
This move serves two purposes: White gains
space on the left wing (the b5-square is now
under firm control) and at the same time
prepares to develop the bishop on a3. From
there it will exert strong pressure on the most
valuable pawn in Black's camp - the d-pawn.
1 5 ...�h6
By activating this bishop, Black sets a small
trap.
16.�a3
The trap was 1 6.ttJc4 i.xc l 1 7.ltJxa5 ? i.b2.
Of course, I did not expect Vladimir to fall
into it.
16 ...�xd2
Almost forced, but not at all a bad move.
It's true that Black parts with the dark­
squared bishop, to which I sometimes feel
Better was the prophylactic 1 8 . . . !!a8 and in the
event of 1 9.ltJxe4 ltJxe4 20.!!xe4 id7 2 l .i.fl
ixa4 22.i.xd6 'Mfxd6 23.!!axa4 !!xa4 24.!!xa4
\t>g7 the draw should not be far away.
19.h3 �aS 20.g4
White gets the pawn back, but his king's
shelter becomes slightly weaker.
Instead 20.!!ac 1 deserved serious attention.
White j ust improves his position, gradually
applying pressure on the opponent's position.
Indeed, it's not so easy to offer a good plan for
Black, whereas White has plenty of ways to
proceed: i.b4, 'Mfa2-a3 etc.
20 ... �c8 2 I .tlJxe4 tlJxe4 22.�xe4 �d7!
23.�fl
Interesting complications could arise
after: 23.i.xd6!? ixb5 24.ixe5 i.e8 25.ib2
(25 .'1Mfh6 !!f7) 25 . . . ttJxd5
65
Chapter 2 - Kramnik
compensation. Only the two bishops give him
some hope.
25J!a3 gcs 26.gc3 %Yb6 27 ..ig2 gxc3
28 ..ixc3 .ih3 29.gel .ic4 30 ..ia5 %Yb7
3 1 .gdl gf4 32 ..ic3 .ih3
8
7
a
b
c
d
e
f
g
h
26J�e6!! tt:Jf4!? (26 .. Jha4? 27Jha4 ixa4
28.'1!Md4; 26 . . . ic6 27.'1!Md4 tt:Jf6=) 27.'1!Mc3
tt:Jxh3t 28.@h2 El:xf2t 29.<tt> xh3 El:xb2 30.'1!Mxb2
Wfd3t 3 I .@h2, with an approximately equal
position.
6
5
4
3
2
23 ....ixa4
b
a
8
d
c
f
e
g
h
33 ..ixe5?
Until this moment Kramnik has defended
very well but, being short of time, he commits
a fatal blunder.
7
6
5
After the correct 33.El:b l ic4 (the point is
that after 33 . . .ixd5 34.ifl ! the b-pawn falls)
34.El:d 1 it would not be easy at all for Black to
increase his advantage. The best way seems to
be: 34 . . . h5!?
4
3
2
1
a
b
c
d
e
f
g
h
24 ..ib4?
Vladimir goes for too much. Perhaps he just
missed Black's answer.
After the obvious 24.ixd6 Wfxd6 25.El:axa4
El:xa4 26.El:xa4 lt:Jc6 27.ig2 tt:Jd4 both sides
have trumps and the position is roughly
balanced.
24 ... b5!
I definitely prefer Black's chances - the extra
pawn is secured and White has no obvious
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
35.gxh5 (in case of 35 .ixe5!? dxe5 36.d6 'I!Md7
37.dxe7 [also interesting is 37.'1!Me l !? hxg4
King's Indian Warfare
66
38.�xe5 �f7 39.dxe7 �xd l t 40.�h2 �xe7
4 1 .�xe7 �d4 42.hxg4 �xf2] 37 . . . �d4 38.�e l
�xd l 39.e8=�t �xe8 40.�xd l hxg4 4 l .�xg4
and in both variations White's chances for a
draw and Black's for a win are about equal)
8
35J�cl
Honestly, I was amazed by this move during
the game. It took me some time to realize that
after 35.dxe7 Black could play 35 . . . �d4 with
an immediate win. Would I have found this
move over the board? I hope so . . .
35 ....ic4 36.VMb4
36.dxe7 �xd2 37.e8=�t �f8
7
6
5
36 ... lb c8 37.VMc5
39.�dl
4
lbxd6 38.VMxe5
�£8
3
2
8
a
b
c
d
e
7
f
g
h
35 . . . lt:J f5! (I like this idea) 36.hxg6 lt:Jh4 Now
White in turn has an extra pawn, but he will
experience great difficulties trying to defend
against his opponent's very active pieces.
33 ... dxe5 34.d6 VMd7!
6
5
4
3
2
1
a
b
c
d
e
f
g
h
39 ... lbf7
Not an outstanding game, but it's a great
feeling to beat the champion of the world!
0-1
a
b
c
d
e
f
g
h
Chapter 3
The King's Indian Bishop
Black against a young Alexander Khalifman
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
T 8
7
6
5
4
3
2
2
a
b
c
d
e
f
g
a
h
8
T 8
T 8
6
6
6
4
4
4
2
2
2
3
a
b
c
d
e
f
g
T 8
6
6
5
b
c
d
e
f
g
a
h
7
T 8
7
5
5
4
4
2
2
2
3
c
d
e
f
g
b
c
d
e
f
g
h
3
3
b
h
6
4
a
g
A small positional exercise. How Playing the KID is about getting
to reorganize the pieces?
the most out of the pieces. How?
(see page 97)
(see page 83)
8
7
f
3
a
h
Can you find Ding Liren's
brilliant combination?
(see page 7 1 )
e
5
5
3
d
7
7
5
c
I believed this was winning for
Black. What did I miss?
(see page 90)
The knight is under attack.
Where to go? (see page 80)
7
b
h
Here is a nice novelty. Don't tell
anyone! (see page 73)
a
b
c
d
e
f
g
h
It is hard to improve the
pieces. Maybe it is time to do
something? (see page 84)
a
b
c
d
e
f
g
h
This is the stage for building up.
What is next?
(see page 1 03)
69
Chapter 3 - The King's I ndian Bishop
In this chapter we shall look at the potential
power of the King's Indian bishop on g7 (with
some occasional assistance from its comrade­
in-arms, the bishop on c8).
hold: 29.:gfc l :gfbg 30.:gc2 :gb4 and only now
3 1 .h4! gxh4 32.�h2, when after 32 . . . :gab8
33.b3 axb3 34.axb3 :gxb3 3 5.:gxb3 :gxb3
36.�h3 White has enough counterplay.
In the standard King's Indian formation the
bishop looks buried on g7, but this is just a
superficial appearance. Black will frequently
give up a pawn or even a piece to bring the
bishop into the game. The most thematic
idea is ... e5-e4, opening up the long diagonal.
This idea can seem very basic, but it is still
important at the highest level.
29 ... gxh4 30.<tt h2 �Ub8 31 .h3 axb3 32.axb3
gal 33.gfdl <tt f7 34.gd2 ga3 35.<tt h 3?!
White continues to be under strong pressure.
The best defence here seems to be 35 .:gd3
�g6 36.�h3 :ga2 37.:gd2, although Black
continues to be a bit better after 37 . . . :gxd2
38.ltJxd2 �f5!.
Wang Yue - Teimour Radjabov
Wij k aan Zee 2009
8
35...gaxb3 36.gxb3 gxb3 37. <ttxh4 <tt g6
38.gc2 gb l 39.tLlg3 h5
White is under pressure and on the last move
before the time control he errs.
8
7
7
6
6
5
5
4
4
3
3
2
2
1
a
b
c
d
e
f
g
h
25 .. .'\Wxb l 26J�axbl e4!
A typical sacrifice in the King's Indian. The
bishop is so strong on the long diagonal that its
release is easily worth the investment of a pawn.
27.tLlxe4 �d4t 28.<tt h l a4
Wang Yue no doubt felt under great pressure
at this point. The combination of the strong
bishop and the open b-file is very unpleasant.
Still, his next move is dubious.
29.h4?!
Black has plenty of compensation for his
pawn, but the computer is able to find a way to
a
b
c
d
e
f
g
h
40.f4? �f6t 4I.<tt h3 gb3! 42.<tt h2 h4
43.tLle2 <tt £5
The black king invades the white position
and wins the game.
44.ga2 gb4 45.ga8 gxc4 46.ge8 gb4
47.ge6 gb3 48.g4t hxg3t 49.tLlxg3t <tt g4
50.tLle2 gb2 5 I .<tt g2 �e5 52.<tt f.2 h£4
53.ge7 <tt £5 54.gf7t <tt e5 55. <tt f3 �d2
56.gxc7 <ttxd5 57.tLlg3 gh3t 58.<tt g2 .1f4
59.tLle2 �e5 6o.<tt n <tt e4 6I .gh7 ga t
62.<tt e i d5 63.<tt d2 d4 64.gh4t <tt d5
0-1
70
King's I ndian Warfare
I do not want to give the wrong impression.
You should not give up the pawn without
pondering the downsides. It does lose a pawn
after all. So, without being downbeat about
it, I want to show a classic game that on the
surface looks very impressive, but is a bit less
so when we get into the details.
Florin Gheorghiu - Garry Kasparov
� i!,aa•l!
:� �. �-!Z- ��
5
4
��zzr•l��
�.�.�
�.�.
l'�&i � �
2
I
�
�,�,
� &. � 0 �'�'
t� s:- "R� '�
�
3 �
� -z•• z� �-�
a
b
c
d
e
f
g
7
6
5
4
3
2
Thessaloniki Olympiad 1 988
6
8
h
The opening is over and the middlegame has
just begun. Kasparov shows that he wants a
fighting game by sacrificing a pawn in order to
bring his bishop to life.
1 5 ... e4!?
If you read the newspaper columns from
30 years ago, this decision was universally
commended. It looks so thematic and the game
was so impressive. But a deeper look suggests
that the decision was rather risky as well.
a
b
c
d
e
f
g
h
Black has long-term compensation for
the pawn. White might be able to fight for
an advantage if he played ltJe2-d4-e6 at this
point. But with timid play, he allowed the
World Champion to show the full potential of
his position.
2 1 .�bl YNf6 22.i.e2?! tlJg3 23.i.xg3
This leaves Black in control of the dark
squares for the rest of the game and could easily
be criticized, but luckily things are never that
simplistic in chess; White also neutralizes the
knight and reduces Black's attacking options.
And he does have an extra pawn to play with . . .
2 3. . .fxg3 24.i.f3 gac8 25.tlJe2 'lNg6 26J�cl
gxcl t 27.'lNxcl gc8 28.'lNe3 'lNf6 29.'lNd2
gc5
8
7
6
1 5 . . . f4! 1 6.ltJxh5 fxe3 1 7.lLlxf6t ixf6 1 8 .�xe3
ih4t 1 9.g3 ig5 was another way to bring the
bishop to life and was also strong.
1 6.tiJxh5 tiJxh5 17.fxe4 f4 18.i.f2 i.g4
19.h3 i.d7 20.0-0-0 i.e5
5
4
3
2
a
b
c
d
e
f
g
h
71
Chapter 3 - The King's Indian Bishop
Black has great compensation for the pawn.
The right choice here is 30.Ek l with chances
to defend, as the absence of knights would
make it harder for Black to find a way through.
Instead Gheorghiu plays too timidly again,
trying to reroute the knight to d3. 1his is swiftly
punished by a powerful bishop invasion.
Teimour Radjabov - Ding Liren
aan
c
d
Zee 20 1 5
8
7
6
30.tiJ cl? �f4! 31 .%Vb4 �b5 32.tiJb3
Black wins slowly after 32.a4 i.xc l ! 33.axb5
�xb5 and White has to choose between giving
up the queen and playing 34.e5 �xb4 3 5 .exf6
i.xb2, both leading to hopeless endgames.
4
32 ... �d3t 33.<bal gc2 34.gbl �e5
1
5
3
2
b
a
8
Wijk
f
e
g
h
33 ... c3!!
Oing Liren gives up a pawn in order to win
a tempo and open the c-line. All other moves
would allow White to get coordinated or
advance with his own threats.
7
6
5
4
34.�c4
The idea behind Oing Liren's combination
is revealed in the following line: 34.lt:Jxc3 �c6!
3 5.lt:Jd5
3
2
1
a
b
c
d
e
f
g
h
A great picture: the bishops are all-powerful.
35.tlJcl �xb2t 36.%Vxb2 YMxb2t
0-1
In the King's Indian we frequently get into
situations where both players are carrying
out their plans, and arriving first is the way
to reap the rewards. If you hesitate, as poor
Gheorghiu did above, you will end up second
best. In the following game we see a reversal
of the fortunes of Radjabov versus a Chinese
opponent. Once again the power of the
g7 -bishop is the dominant feature of the
position.
a
b
c
d
e
f
g
h
35 . . JWxd5!! 36.�xd5 �xb2t with mate in a few
moves.
34... cxd2 35.tiJf6t
35 .Wfxd2 �xb2t and Black wins.
72
King's Indian Warfare
35 ... \t>ffi 36.tlJh7t ®es 37.L£7t ®xf7
38.�xf5t ®gs 39.®c2
This loses quickly, but Black is also winning
in the following long line: 39.'\We6t <j;>xh7
40.'l.Wxe4t <j;>hs 4 1 .'\Wh4t <j;>gs 42.'\Wc4t <j;>fg
43.'\Wf4t <j;>e7 44.'\We4t ieS 45 .'\Wh4t <j;>d7
46.'\Wh?t <j;>e6 47.'\Wg6t if6 48.'\We4t <j;>d7
49.'\Wh?t <j;>es 50.'\Wg8t <j;>e7 5 1 .'\Wh?t <j;>fg
and White is running out of checks.
39 .. J�xb2t 40.\t>dl
8
7
6
5
4
3
2
Jan Gustafsson -Vladimir Kramnik
Dortmund 20 1 2
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
18 ... �xe3! 1 9.tlJxe3 tlJb4 20.�c4?
Gustafsson has been under a lot of pressure
since the very beginning of the game and here
he fails to find the accurate defence.
a
b
c
d
e
f
g
h
40 .. J�b l t! 4 1 .®xd2 �xa2t 42.®e3 �el t
43.®£4 �xflt
0-1
One of the ways to make more of the King's
Indian bishop is to neutralize the opponent's
dark-squared bishop. Kasparov managed to
do so above by taunting his opponent with a
cheeky knight. But Gheorghiu had no need to
take it, and maybe should have considered not
doing so. A more forceful way to ensure dark­
squared domination is to give up the exchange.
This is obviously a double-edged decision; a
rook is more powerful than a bishop in general.
But there are situations where the bishop,
because of how the rest of the pieces are placed
or interact, is very powerful and offers at least
sufficient compensation. The following game
is such a case.
After 20.<j;>f2! ttJxa2 (20 . . . id4? 2 1 .El:c4)
2 1 .El:c7! ttJc3 22.d6 lDxe2 23.<j;>xe2 ie6
24.E!:xb7 a2 25 .d7 El:d8 26.El:a7 ixb3 27.El:a3
ie6 28.El:dl <j;>fg 29.El:a7 the game would end
in a draw.
20 ... tlJxa2 2 1 .�a4 �xa4 22.bxa4 .1d4
23.\t>fl tlJb4!
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
73
Chapter 3 - The King's Indian Bishop
24J�cl a2 25J�xc8t i> g7 26J�cl tlJxd5
27J�dl tlJxe3 28J�xd4 al=VN 29. i>xe3 VNgl t
0-1
To me the most important thing about this
game is that Kramnik played the KID and
won. It shows that he probably does not think
that the King's Indian is a bad opening.
Before we move on to the more complex case
of my own games, I want to show a modern
masterpiece, in which the King's Indian bishop
is freed from its cage by violent means and
comes out to dominate the board.
Loek van Wely - Daniel Stellwagen
Amsterdam 2009
l .d4 tlJf6 2.c4 g6 3.tlJc3 i.g7 4.e4 d6 s.tlJf3
0-0 6.i.e2 e5 7.0-0 tlJc6 8.d5 tlJe7 9.b4
tlJhS 10.g3 f5 I I .tlJgS tlJf6 1 2.f3 f4 13.h5!?
fxg3 14.hxg3 h6 1 5.tlJe6 he6 16.dxe6 VNcS
17.tlJd5 VNxe6 1 8.tlJxc7 VNh3
z
.
z
J����z7 �i� • •
%-- -%� ____ ,;___ __ ,;_,,�..
� !� z. �z
"
�.
�-!�
3 . � - � di
S
__
/,
6
5
4
2 ��=�:-:��
� ��� M m
a
b
c
d
e
f
g
h
19JUl!
1 9. ctJ xa8 Wxg3 t has resulted in five draws
in my database. This is not really an opening
book, but I still find it worth mentioning
20.@h l lt:Jh5!!N.
a
b
c
d
e
f
g
h
I find it surprising that people who would
happily sacrifice a rook did not find this move.
Maybe the main problem is realizing that the
knight is heading for g3 not f4? 2 1 .Wc2 Wh3t
22.@gl ctJg3 gives Black a devastating attack.
The engine's main line goes like this: 23 . .ib2
�f4 24.�fe l .if6 25 . .id l Wh l t 26.@f2 ctJxe4t
27.�xe4 lt:J f5 28.�xf4 exf4, when Black is
completely winning after both 29.Wxf5 Wh4t!
and 29.@e2 .ixb2.
19... tlJxe4! 20.fxe4
20.�h2 Wd7 2 l .fxe4!N (2 l .ctJxa8 ctJxg300
was played in Van Wely - Degraeve, Mondariz
2000. The common wisdom was that Black
was doing very well.) 2 1 . .. Wxc7 22 . .ie3 would
give a strategically interesting position. The
engine thinks that White has the advantage,
while I would guess there are chances for both
sides.
20 .. J�xf'2 2 I . i> xf'2 �Ust 22. i> e3 VNxg3t
23. i> d2 �H2 24.tlJ e8
24.We l has also been played a few times.
The engine suggests 24 . . . h5!?N as the way
forward. For example: 25.@dl h4 26.c5 d5
27.ctJxd5 ctJxd5 28.exd5 h3 29.c6 Wh4 30.c7
Wa4t 3 l .@d2 Wa5t 32.@d l with a draw.
Luckily chess is so rich and humans so creative
that no one has followed this computer line to
the end yet.
74
King's Indian Warfare
26.c5?
This loses to a brilliant line-opening tactic
that brings out the King's Indian bishop.
26.@e l ?! is also unsatisfactory. After 26 . . . �g3!
27.@d2 .ig5t 28.@c2 .ixc l 29.�xc l �e3 we
have a complex endgame where Black's extra
pawn gives him all the chances.
a
b
c
d
e
f
g
h
The drawing line is rather difficult: 26.�e l !
�h2! 27.@di ! (27.c5 ? fails to 27. . .ttJc8!!+ with
the idea 28.ttJxc8 .ih4! 29.@d l �xe4 and
White's position collapses) 27 . . . �g2 28.@c2!
So far the players have repeated the moves
from Van Wely - Radjabov, Dresden (ol)
2008. In that game Black played 24 . . . h5 and
was worse. Stellwagen had a big improvement
ready.
24 .. .'�f3!!
The improvement. Objectively the position
is equal; in the real world, where engines
cannot assist you, White's task is very difficult.
25.�xd6 .if6!
This is the critical moment of the game.
Stellwagen has played his improvement and
Van Wely now had time to think about what
to do. But defending is harder than attacking,
and he did not manage to find the very difficult
sequence of moves needed to achieve a draw.
a
b
c
d
e
f
g
h
c
d
e
f
g
h
28 . . . ttJc8! (Black can also force a draw
immediately with 28 . . . �f3 29.@di �g2)
29.ttJxc8! �xe4t 30.@b3 �xe2 3 l .�c3
From here on, White has some equally good
alternatives, but also a lot of inferior ones.
3 l . . . �f2 32.ttJd6 �e2 33 . .ie3! e4 34 ..id4
.ixd4 35.�xd4 �f3t 36.@b4
a
a
b
b
c
d
e
f
g
h
75
Chapter 3 - The King's Indian Bishop
36 . . . :gd3! 37.�e5 �d2t
39.i>b4 �d2t with a draw.
38.i>c5
�e3t
8
7
6
5
4
a
3
c
b
d
e
f
g
h
With mate in two moves, usually starting
with . . . :gfl t.
2
1
a
b
c
d
e
f
g
h
26 ... ttld5!! 27.exd5 e4!
The bishop has finally come to life, with
devastating effect.
29.<tfid2 .ixal 30.e6?
30.ttJc4!+ would have kept the battle going.
30 .. JWO!
8
28.<tfiel
The only move if you are a human.
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
The dust clears and Black wins.
a
b
c
d
e
f
g
h
28 .. JWg2?
Stellwagen in turn misses a great sequence:
28 . . . .ic3t 29 . .id2 �g3!! 30 . .ixc3 e3!
3I.ttlxe4 �xe4 32.e7 �f4t 33.<tff e l .ie3t
34..id2 �g3 35 ..ixe3 g a t 36.<tfid2 gxe3
37.d6 �e3t 38.<tfiel gel 39.�xel �xelt
40.<tfif2 �eSt 4I .<tfif3 <tff f7 42..ifl �f5t
43.<tbg3 �e6! 44. <tfif2 h5 45.a4 <tfif6 46 ..ig2
�e8 47.<tff g3 g5
0-1
76
King's Indian Warfare
My Games
In the first game I managed to make the most
of my bishop by opening the centre in classical
style.
following interesting encounter: 8 . . . cS 9.'1Mfd2
tt:J b4 1 0.ig2 hS 1 l .ixf6 exf6 1 2.0-0-0 hxg4
1 3.hxg4 ixg4 1 4.a3 ttJc6 1 S.'Wf4 'WeB 1 6.ttJdS
Giorgi Kacheishvili - Ilya Smirin
Minneapolis 2005
This game was played in the last round of the
Minneapolis Open tournament. Both players
needed a win to guarantee a decent place.
I .d4 llJf6 2.c4 g6 3.llJc3 .1g7 4.e4 d6 s.llJf3
0-0 6.h3 llJ a6
Lately I have usually played 6 . . . eS 7.dS aS
here. Which line to choose is, as so often,
a matter of taste. It seems that my taste has
changed a little.
7..1g5 YMes
a
b
c
d
e
f
g
h
1 6 . . . 'We6 1 7.tLlc7 'Wxc4t 1 8.@b 1 fS 1 9.ifl
'Wa4 20.ibS 'WaS 2 l .ttJdS 'WxbS 22.'1Mfxd6
f6 23.'We6t �f7 24.lt:Jh4 'We2 2S .�de 1 fxe4
26.'Wd6 �d8 27.'1Mfh2 '1Mfd3t 28.@a1 �xdS
29.tLlxg6 �hS 0-1 Macieja - Smirin, Maalot­
Tarshiha 2008.
But the most logical continuation here seems
to be the natural developing move 8.id3. At
least in the following game I faced problems:
8 ... eS 9.0-0 ttJhS 1 0.ic2 fS 1 l .exfS ixfS
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
8 ..1e2!?
An apparently modest but quite logical
move. Why not? Mter all, the pieces have not
come into direct confrontation yet.
The alternative 8.g4, playing "wide" and
trying to grab the maximum amount of
space, is much more popular. I managed to
beat the Polish Grandmaster Macieja in the
a
b
c
d
e
f
g
h
1 2.�c l (after the correct 1 2.ixfS gxfS 1 3.ttJdS
Black would remain worse - the knight on a6
is not a piece to be proud of, as the following
sample variation shows: 1 3 . . . @h8 1 4.�e 1 e4
1 S .ttJ h4 'Wf7 1 6.ttJe3 f4 1 7.ttJefS±) 1 2 . . . '1Mff7
1 3.ie4? ixe4 1 4.ttJxe4 exd4 1 S .ttJxd4 �ae8;
(0- 1 , 36) Wendt - Smirin, Ohrid 2009.
77
Chapter 3 - The King's I ndian Bishop
8... e5 9.d5 ttlh5 10.g3
It's essential to take the f4-square away from
the knight on h5.
10... £5
Perhaps 1 O .. .f6 1 1 .id2 f5 , chasing the bishop
from g5 to d2, deserved some attention here. It
has been played in several games - here is one
of them: 1 2.ctJh4 tLlf6 1 3.exf5 gxf5 1 4.'!Mc2 e4
( 1 4 . . . ttJb4 1 5 .'!Mb3 a5 1 6.a3 ctJ a6 1 7.'!Mc2 ctJc5
1 8 .ie3 would transpose to the game) 1 5 .ie3
ctJd7 1 6.ctJg2! ctJe5 1 7.ctJf4 ctJc5 1 8.0-0-0;
(Yz-Yz, 35) Mchedlishvili - Avrukh, Beersheba
2005. The position is complicated, but White's
piece setup looks somewhat "prettier".
l l .exf5 gxf5
13 ... ttlb4
Another logical (and perhaps better) option
would be 1 3 . . . c6! ?, preparing to open the c-file
in the event of long castling. Mter that the
game might continue 1 4.g4!? ( 1 4.ctJxf5 ixf5
1 5 .'!Mxf5 ctJ xd5 is unclear) 1 4 . . . cxd5 1 5 .cxd5
f4 with an interesting position, full of potential
dynamics.
14.%Vb3 a5
This is a useful manoeuvre to improve Black's
position on the queenside.
1 5.a3 ttl a6 16.%Vc2
8
7
8
6
6
4
4
2
5
7
3
5
3
2
a
a
b
c
d
e
f
g
h
Black's position looks quite attractive due to
the active pawn duo on e5 and f5 . However,
these pawns are also a cause for concern:
the one on f5 may easily become a target.
Moreover, the advance of either pawn may
create vulnerable squares for the white knights:
d4 and f4 in the case of ... e5-e4 and e4 in the
case of .. .f5-f4. In other words, there is no
reason for Black to be too optimistic here.
12.ttlh4 ttlf6 13.%Vc2
Eyeing the pawn on f5 .
b
c
d
e
f
g
h
16 ... ttl c5
The most natural move.
After 1 6 . . . a4 1 7.0-0-0 ( 1 7.'!Mxa4?! '!Mxa4
1 8.ctJxa4 ctJe4) 1 7 . . . ctJc5 1 8.ie3 ctJ b3t
1 9.i>b 1 id? 20.ctJb5 '!Mc8 (20 ... Ek8!?)
2 1 .E!:hgl i>h8 22.f3 White's chances should be
preferred - it's very hard for Black to organize
decent counterplay, and his position lacks
flexibility; ( 1 -0, 35) Mchedlishvili - Hillarp
Persson, Gothenburg 2005.
17 ..ie3!
Of course not 1 7.ctJxf5 ? ixf5 1 8.'!Mxf5
ctJ fe4-+.
17... b6
78
King's Indian Warfare
Black wants to recapture with the b-pawn in
the case of a possible i.xc5.
17 ... a4 1 8.i.xc5! dxc5 1 9.'Dxf5 i.xf5 20.Wfxf5
ttJxd5 2 1 .Wie4 tDxc3 22.bxc3 would lead to a
very unpleasant position for Black, especially
since it was a must-win situation for both
players.
The careless 1 7 . . . i.d7 would allow my
opponent to reveal the following strong idea:
1 8.g4! fXg4 1 9 .hxg4 lDxg4 20.i.xc5 dxc5
2 l .i.xg4 i.xg4 22.f3 i.d7 23.0-0-0, when
White has good compensation for the pawn,
and will develop an initiative along the g- and
h-files, even though after 23 .. J!a6 the position
is not yet so clear.
1 7 ... e4 or 1 7 . . . tDce4 were also possible, but I
did not particularly like either move.
to be honest, after 20.El:hgl c±>h8 2 l .f3!?, with
the same g4 idea, I would still prefer White
here.
20.l!J b5
Giorgi correctly played this natural and
strong move. The knight occupies an active
position, at the same time "just in case"
closing the b-file. It looks highly improbable
that Black will develop a deadly attack along
this file, but actually it happened later in the
game! Of course, 20.ttJb5 is not to be blamed
for that . . .
8
7
6
5
4
1 8.0-0-0 a4 19.®b l
3
2
8
1
7
b
a
6
d
c
f
e
g
h
20 .. J3b7?!
This looks a bit ugly and indeed it is! I did
not want to allow:
20 . . . Wid8
5
4
3
2
a
b
c
d
e
f
g
h
19 .. J3b8?!
This artificial move appears to be a serious
inaccuracy.
The correct decision would be 1 9 . . . tD b3,
avoiding a possible capture on c5 and
neutralizing to some extent White's idea to
play g4, since Black can then answer . . .f4. But
a
2 1 .lDa7!?
b
c
d
e
f
g
h
79
Chapter 3 - The King's Indian Bishop
After 2 l .g4 lLlxd5 the knight on h4 is
hanging.
2 1 .tLlg2!?, preparing g4, deserves serious
attention here; White's prospects look good.
But I should probably not be too worried
about the text move. After:
2 1 . . .id7 22.tLlc6
As often in the KID, Black's resources should
not be underestimated: 22.ixc5? bxc5
23.lLlxf5 lLle4! 24.tLlc6 ixf5 25 .id3 lLlc3t
26.�xc3 e4 and Black completely takes over
the initiative.
22 . . . ixc6 23.dxc6 lLlfe4 24.if3 �f6
Black's position would be acceptable - the
white knight is rather misplaced on h4 under
these circumstances.
21 .g4!
This is the key move of White's strategy.
21. .. £4
It would be unwise to open files leading
to Black's king: 2 1 . .. fxg4? 22.hxg4 lLlxg4
23.ixg4 ixg4 24J3:dgl �d7 25.f3 and Black
loses material.
23 ... e4!?
It's easy to understand why the natural but
passive move 23 . . . tLld7 didn't appeal to me.
After 24.id3! (24.tLl f5 lLl b6 25.lLlxg7 c;t>xg7
26.id3 �f7 does not look bad for Black)
24 . . . �h5 25.lLlf5 lLl b6 26.lLle7t c;t>hs
a
b
c
d
e
f
g
h
27.ie4! (in the event of 27J3:hgl e4! 28.ixe4
lLlxc4! 29.�xc4 �e8 things are not so clear
- this line has some similarities with what
actually happened in the game) 27 . . . �xg5
28.lLlxc8 �xc8 29.�dgl
22 ..ixc5 bxc5
8
7
6
5
4
a
b
c
d
e
f
g
h
White would get a firm grip on the resulting
position. I was going to obtain counterplay at
any price (in this case - a piece) .
3
2
a
b
c
d
e
f
g
h
23.g5!
It seems that White rules, but I did not want
to accept that.
24J�del?
Just one mistake - and the game is effectively
gone! Giorgi should not have trusted me principled and correct was: 24.gxf6 ixf6
25 .�dgl t c;t>h8 26.�g4! This move was
80
King's I ndian Warfare
probably missed by both of us, but honestly I
was not in the mood to search for it too much.
26 . . . Wfe7 (or 26 . . . Wie5 27.f3!±) 27.!hf4 ixh4
28.!hf8t Wfxf8 29.ig4! Black would remain
with too many weaknesses to count seriously
on a positive outcome.
30 . . . �b3! 3 l .<±:>a2 ixc3 32.bxc3 �xb 1 33.�xb l
�xb l 34.<±:>xb l Wfxh4 35 .Wfxe4 Wfxf2 And the
black king will easily escape the checks.
27...VMxg5 28 ..ig4! !Ubs 29 ..ixd7?!
This allows an elegant fi n ish.
More stubborn was: 29.�b l ixg4 30.hxg4
ixc3 3 l .Wixc3 �b3 (weaker is 3 l . . . Wixd5t
32.<±:>al �b3 33.ctJf5! �xc3 34.ctJe7t <±:>g7
35.ctJxd5) 32.Wic4
24... tlJxd5!!
This sacrifice completely turns the tables.
The bishop on g7 becomes a monster, and
an assault begins along the b-file (see the
comment after move 20) .
25.cxd5 .id? 26.tlJc3 VMeS
Now the pawn on b2 looks really frightened.
27.<i>a2
27.<±:>al loses after a pretty forced line:
27 . . JHb8 28.�b l Wlxg5 29.ig4 ixg4 30.hxg4
a
b
c
d
e
f
g
h
32 ... Wie5 33.Wic2 Wfxd5 34.<±:>al Wfe5 35.<±:>a2
Wfe6! 36.<±:>al c4 with an irresistible attack:
37.ctJf5 c3 38.ctJd4 cxb2t 39.Wixb2 �xb2
40.ctJxe6 �xb l t 4 l .�xb l �xb l t 42.<±:>xb l e3
43.fxe3 f3 Still, I would have needed to find
all this over the board. Now Black's task is
simpler.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
81
Chapter 3 - The King's Indian Bishop
29. .J�xb2t! 30.Wxb2 gxb2t 3I .@xb2
YMxd5!
Despite having almost the whole set of pieces
for the queen, White is absolutely helpless
against Black's threats. White's numerous but
uncoordinated forces are unable to help His
Majesty.
Denis Khismatullin - Ilya Smirin
Poikovsky 20 1 5
I .d4 �f6 2.c4 g6 3.�c3 .ig7 4.e4 d6 5 ..ie2
0-0 6 ..ig5 � a6 7.YMd2 e5 8.d5 c6
8
32.@cl
Or 32.�xa4 �d2t.
7
32 ....ixc3 33.ghgl t @f8 34.ge2 YMb3
5
6
4
8
3
7
2
6
5
b
a
4
3
2
1
a
b
c
d
e
f
g
h
A picturesque position. The queen is ready to
pick up her last prey - the pawn on a3 - with
checkmate to follow soon. White resigned.
0-1
d
c
e
f
g
h
9.f3
Sometimes White tries to do without this
move and plays 9.�d3, protecting the e-pawn
in advance and vacating e2 to develop the
g 1 -knight. In a rapid game with Avrukh in
1 996 I quickly obtained a good position by
simple and energetic moves: 9 . . . tt.J c5 1 0.�c2
cxd5 1 1 .cxd5 a5 1 2.tt.Jge2 �d7 1 3.0-0 b5
1 4.f3 b4 1 5 . tt.J d 1 �b6 1 6.�e3 �fc8 1 7.lt.Jf2
�b5 1 8.�fd l �a6 1 9.lt.Jcl lt.Ja4
In the next game we shall see a thematic,
though not very common, bishop manoeuvre,
followed up with a great dynamic takeover of
the dark squares.
a
b
c
d
e
f
g
h
Black soon had a winning position - which
nonetheless I lost; ( 1 -0, 42) Avrukh - Smirin,
Tel Aviv 1 996.
82
King's I ndian Warfare
9 ... cxd5
I played 9 .. .'�a5 against Leitao -you will find
this game elsewhere in the book (page 1 28) ,
where I also look a bit on 9 . . . cxd5 . This time I
went for the more common continuation.
the disadvantages of this idea: 1 1 . . . h6 1 2 . .ie3
( 1 2 . .ixh6? 'Dxe4 1 3.'Dxe4 �h4t is a wellknown motif) 1 2 . . . h5 1 3.h3 What about the
knight on gl ? 1 3 ... 'Dc5 1 4.0-0-0
1 O.cxd5 .id?
8
7
6
5
4
a
3
2
1
a
b
c
d
e
f
g
h
I I ..idl !?
The standard manoeuvre in this type of
position. The idea is to set up the pieces in
a natural and harmonious way - the knight
on gl will be developed to e2 and the bishop
usually goes to c2 shortly after that. This is
more often carried out before the exchange
of c-pawns. This small difference allows Black
to gain some activity on the queenside and to
dispute White's usual supremacy there.
l l ..id3?!, with the same idea, is worse because
of l l . .. 'Dc5 1 2 . .ic2 a5 1 3.'Dge2 ( 1 3.a4 �b6)
1 3 ... b5 and Black is fine, as he was in the game
Avrukh - Smirin mentioned above (actually
after 1 4.0-0 the position from that game
would arise) .
White may choose another plan starting
with the popular l l .g4 - to grab space on
the kingside. The king usually hides on
the queenside after castling long. In my
opinion the following game demonstrates
b
c
d
e
f
g
h
1 4 . . . �b8!? 1 5 .@b l El:c8 1 6.El:h2 b5 1 7.El:c l a5
1 8 . .id3 El:a7 1 9.'Dge2 El:ac7 20.'Dd l a4 2 l .'Df2
b4 22 . .ixc5 dxc5 23 . .ic4 .ib5 24.'Dd3 .ixc4
25.El:xc4 �b5 26.b3
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
26 . . . ttJe8! 27.El:h 1 'Dd6 28.El:hcl hxg4 29.hxg4
axb3 30.axb3 'Dxc4 3 1 .El:xc4 �a6
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
83
Chapter 3 - The King's Indian Bishop
32.lDec l �f6 33.�e3 �h4 34.g5 �h3 35.lDf2
�fl 36.lDg4 �a8 37.�f2 �h l 38.@c2 �al
39.@d2 �b l 40.ltJe3 if8
8
7
6
13.axb4 llJxb4 14.llJge2
The main battlefield is the queenside. Black's
position is a bit cramped, so it's important
for him to place his pieces well to match the
opponent's activity. After some thought, I
found a good way to set up my forces.
8
5
7
4
3
6
2
5
a
b
c
d
e
f
g
h
4 1 .�g3 �a7 42.�xe5 �xf3 43.d6 �f2t
44.@d3 �b2 45.lDc2 �d7 0- 1 Alterman - Xu,
Cap d'Agde 1 994.
1 1 ... b5
Certainly not l l ... lDc5? 1 2.b4.
4
3
2
a
12.a3 b4
During the game I thought this move was
almost mandatory.
Now I believe that the following is a worthy
alternative: 1 2 . . . lDc7 1 3.lDge2 (if 1 3.b4 a5
1 4.�b l axb4 1 5 .axb4 �c8 1 6.lDge2
b
c
d
e
f
g
h
g
h
14...'/Nbs
Everyone onto the queenside!
1 5.0-0 gcs 16 ..ie3
8
7
6
5
8
4
7
3
6
5
2
4
1
3
2
a
a
b
c
d
e
f
g
h
1 6 ... ltJa8! This knight is going to c4. 1 7.0-0
lDb6 Black is OK.) 1 3 . . . a5 As played in the
game Ubilava - Grigoriants, Linares 2003.
b
c
d
e
f
16 ....ie8!
A very important move. The bishop frees d7
for the knight, and at the same time the pawn
on f7 will be defended. Why might it need
protection? Because White will start kingside
84
King's Indian Warfare
activity by means of f3-f4 and a white rook will
operate on the f-file.
17.�a4
White is doing some regrouping as well.
Something like 2 1 .El:fc l lDxa4 22.bxa4 ib6
could have been played. The position would
be roughly balanced after that - Black's pieces
are well placed and the opponent's space
advantage is not significant.
8
17... �d7 1 8.�ec3
7
8
6
7
5
6
4
5
3
4
2
3
2
a
a
b
c
d
e
f
g
h
18 ....if6! 19 ..ie2 .ids
Almost all the black pieces are on the 8th
rank, but they function and interact well
enough - an interesting positional observation.
Seemingly Denis was too optimistic here, but
it was easy to overestimate the meaning of
White's space advantage during the game.
20.b3 � b6
It looks as if Black is just going to exchange a
pair of pieces - a usual thing to do in a somewhat
cramped position. Khismatullin naturally wants
to avoid this, but falls into the trap.
2 1 .�b2?
This manoeuvre was prepared by the previous
move. It's interesting that Houdini initially
approves this very logical continuation - the
knight is on its way to c4 to exert pressure on
the queenside.
2 1 .f4!? lD xa4 22.bxa4 ib6 23.!hcl exf4
24Jhf4 ixe3t 25 .'.Wxe3 'Wic7 should be OK
for Black.
b
c
d
e
f
g
h
2 1 ..J�xc3!
A simple but nice tactical blow. Suddenly the
potential energy of Black's pieces is released.
It's interesting that if the white king were on
h 1 he would have nothing to worry about, and
might claim a positional advantage. But in that
case he would lack the time for the b2-b3 and
ttJ b2 setup. The tragedy of a single tempo?
22.%Vxc3 �6xd5 23.exd5 �xd5 24.%Vd2
�xe3
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
Chapter 3
-
85
The King's Indian Bishop
25JUel ?!
A generous move - who cares about the little
pawn on b3? Still, it was worth selling for a
higher price.
25 .8:fcl "1Wxb3?! 26.lLJc4 (not 26.8:c3? "1Wb4)
26 . . . lLJxc4 27.ixc4 "1Wb6t 28.@hl and White's
activity does not give the opponent real hopes
of converting the material advantage.
Better is 25 ... ib6 26.@h l id4, striving for
activity and centralization.
25 ...VHxb3 26.i.d3 i.b6
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
Black's advantage is now overwhelming.
27.�hl �k8 28.�dl!?
The most resilient. The strong knight on e3
should be exchanged or at least disturbed.
a
b
c
d
e
f
g
h
After 27.8:ab l Black may take the pawn
under better circumstances. (27.8:a3 lD f5 28.b4
"1Wd8 would leave Black with both material and
positional superiority) 27 . . . "1Wxb3 (also strong
enough is 27 ... "1Wd8 28.lLJc4 lLJ f5 29.id3 d5
30.ixf5 dxc4 3 1 .ie4 c3) 28.lLJd l (or 28.lLJc4
lLJxc4 29.ixc4 "1We3 30."1Wxe3 ixe3) 28 . . . "1Wa3
29.lLJxe3 "1Wxe3
28 ... �xdl
28 ... 8:c2!? 29.ixc2 lDxc2 30.lLJc3 lLJxe l
deserved attention as well, but I decided in
favour of the simple exchange of knights.
29J�axdl i.d4
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
And the bishop plus three pawns are too
many for the rook, even though White retains
fighting chances.
a
b
c
d
e
f
g
h
30.'\We2?!
30.8:b 1 was demanded and would indeed
have been the relatively best decision.
86
King's I ndian Warfare
30 ....ia4
30 . . . a5 3 l .�b l We6 was not bad either.
3 1 J�bl YMc3 32.YMfl
35J3xb3 aS 36.£4 e4 37J�bbl
37.�d l Wf2 38.Wxf2 ixf2 39.g3 would
be more resilient. We both were in mild time
trouble.
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
b
a
d
c
f
e
g
h
32 ...'iMd2! 33.�c4
If 33.�bc l �xc l 34.�xc l a5 35 .�c8t <;t>g?
36.ic4 Black would win rather simply after
36 ... Wc3 37.ie6 ic5 38.Ek7 ie8 and so on.
33 .. J3c5
33 . . .ic2, with the idea: 34.�bc l ib2
a
b
c
d
e
f
g
h
37 ... e3?!
Another inaccuracy in a technically won
position.
Stronger was 37 . . . d5 38.�bd l Wb4 39.f5 ic3
40.�e3 a4 4 1 .Wa6 Wc4 still with a fairly easy
win.
38J�edl YMc3 39J�d3 YMc4 40.gbdl?
40.Wd l �d5 4 l .�c l Wb4 42.�c2 would
make Black work hard to achieve victory - the
rook on d5 is quite awkwardly placed.
8
7
6
a
b
c
d
e
f
g
h
Would have won quickly.
34.�b3 hb3?!
Now the simplest way would be: 34 . . . ib5
35 .�ed l WaS 36.We l Wxe l t 37.�xe l a5
38.ia2 a4 39.�ec l id3 40.�xc5 dxc5 etc.
4 l .�b8t <;t>g7 42.�a8 c4 43.�xa4 c3
5
4
3
2
1
a
b
c
d
e
f
g
h
87
Chapter 3 - The King's Indian Bishop
40 .. J�c8!
Fortunately I found this strong move - the
last one before the time control. Now the win
is easy.
41.f5
Or 4 1 .�xd4 e2 42.�xc4 exfl =�t 43.�xfl
�xc4.
4I. ...ic5 42.VNf3 ges 43.f6
Losing at once, but 43.�c6 �e4 44.�xe4
�xe4 45.�e 1 a4 would also be hopeless.
Christopher Lutz
b
c
d
e
f
g
h
43 . .VNxd3
White resigned in view of 44.�xd3 e2.
0-1
.
Just as Kasparov's bishops dominated
Gheorghiu, I was able to dominate a top
German grandmaster with my own two
bishops in the following game.
Ilya Smirin
Gronin gen 1 993
l .d4 ttlf6 2.c4 g6 3.ttlc3 .ig7 4.e4 d6 5.ttlf3
0-0 6..ie2 e5 7.0-0 ttlc6 8.d5 ttl e7 9.b4 a5
IO.bxa5 gxa5
As in the game Postny - Smirin elsewhere
in this book (page 1 46) , I preferred this
immediate capture instead the popular 1 0 . . . c5.
b
a
a
-
d
c
f
e
g
h
l l .ttlel
In my opinion the moves 1 1 .lLld2 or 1 1 .a4, as
was played by Postny in the above-mentioned
game, look more logical than moving the
knight from f3 to d3 via e 1 . After 1 1 .a4 I had
played another interesting encounter in this
line: 1 1 . . . c5 1 2.�b 1 mh8 1 3 . .id2 �a6 1 4.�c l
lLl fg8 1 5 .g3 h6 (I did not like 1 5 . . .f5?! because
of 1 6.lLlg5) 1 6.lLlh4 g5!?
a
b
c
d
e
f
g
h
88
King's I ndian Warfare
1 7.ctJg2 ( 1 7.ctJf5!? ctJxf5 1 8.exf5 ixf5 1 9.�xb7
.ic8 20.�b8 f5 with double-edged play) 1 7 . . . f5
1 8 .f4 gxf4 1 9 .gxf4 fxe4 20.ctJxe4 Ct:J f5 2 l ..ic3
Ct:J d4 22 ..ixd4 exd4 23.�c2 ctJe7 24.ctJg3 ctJ f5
25.ctJxf5 .ixf5 26 . .id3 �d7 27.�fe l ih3
28.<;t>h l �g4 29.�e6
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
29 . . . �xa4 30.�g6 ixg2t 3 l .�xg2 �xg2t
32.<;t>xg2 �xf4 33.�xb7 �a2t 34.<;t>gl �al t
3 S .<;t>g2 Yz-Yz Bykhovsky - Smirin, Haifa
2008. In my opinion this game was played
quite well by both players.
1 I . .. ltJd7 12.ltJd3 f5 13 ..id2
Mter 1 3.f3 Black may continue 1 3 . . . c5!?
anyway.
13 ... c5!?
idea c4-c5, his queenside is solid enough and
has no visible weaknesses, and the pawn on d6
will be comfortably protected by a rook on a6.
14.exf5
Lutz prefers this exchange over the possible
1 4.f3. Indeed Black seems to be fine after
the probable 1 4 ... �a6 1 5 .�c2 Ct:Jf6 1 6.�ab l
f4 ( 1 6 . . . h5!? first deserves attention) and it's
advisable for White to play 1 7.g4 - otherwise
he may easily find himself under attack
without real counterplay on the queenside.
It's important that in the event of 1 4.f4 �a6!?
Black is well prepared for the clash in the
centre.
14 ... gxf5
1 4 . . . Ct:Jxf5 1 5 .ctJe4 �a6 was quite possible - I
believe Black is OK. But unlike in the much
later encounter with Postny I decided to take
on f5 with the pawn.
1 5.f4 ltJg6
I did not want to release the tension:
1 5 . . . e4 1 6.ctJcl (of course not 1 6.ctJ f2?? e3 or
1 6.ctJel ??) Even though objectively it was not
a bad decision at all; but I was eager to play a
more lively position.
16.%Vc2 �a6 17.�hl exf4 18.ltJxf4 ltJxf4
19.hf4 ltJe5
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
I like such pawn structures - Black stops the
a
b
c
d
e
f
g
h
89
Chapter 3 - The King's Indian Bishop
The strong knight on e5 offers Black some
dynamic possibilities - the weak pawn on f5 is
not actually that weak.
20J�abl
The prophylactic 20.g3 would be a subtle
choice. Black might sacrifice a pawn in this
case as well: 20 ... tLlg6 2 l .id2 f4!? 22.gxf4
(bad is 22.ixf4? :ga3! - a rather unexpected
rook incursion - 23.id2 ih3 and Black's
initiative is too strong) 22 ... 'W'h4 23.id3 ih3
24.ixg6 hxg6 25 .'W'xg6 ixfl 26.:gxf1 and
neither side has much to complain about - the
position is in dynamic balance.
20.a4!? deserved attention as well - as we saw
from the previous line it's useful to restrict the
potential activity of the rook on a6.
20 ... �g6 2 I ..id2 '1Mh4
The queen gets closer to the white king,
causing him at least psychological discomfort.
By the way . . . ie5 is a serious threat.
The only move, combining activity with
salvation of the queen.
24.�fl
Certainly not 24.gxf4? ifS.
24...'1Md7?!
The queen goes a little bit too far - it would
be better to stop on f5:
24 . . .'W'f5
Now White has to act precisely to avoid
trouble:
25.tLle4!
25 .id3 'W'h5 cannot be recommended.
25 . . . :ga3
The queen may return: 25 . . . 'W'h3 26.ixf4
(26.tLlf2 'W'd7 leads to a draw of course)
26 . . .if5 27.id3 lLle5 28.tLlf2 (28.ixe5?
ixe4t)
8
7
6
5
22.g3 '1Mh3 23.�dl !?
4
8
3
7
2
6
a
5
4
3
2
a
b
c
d
e
f
g
h
White also has some threats - for instance,
this knight is going to f2, trapping the intruder
queen.
23... f4
b
c
d
e
f
g
h
28 ... lLlxd3! 29.lLlxh3 ie4t 30.�g2 Forced.
(bad is 30.<±>gl id4t 3 1 .tLl f2 lLlxf2)
30 . . . ixg2t 3 l .<±>xg2 :gxa2t 32.<±>h l lLle5
And the game should end peacefully.
26.:gb3
26.ixf4? :gc3!
26 . . . :gxa2!? 27.tLlf6t ixf6 28.'W'xa2 'W'e4t
29.if3 'W'xc4 30.:gc l 'W'd4 3 l .ixf4 lLlxf4
32.gxf4 �xf4 33.:gf1 'W'h4
And despite Black's slight material advantage
(two pawns for the exchange) the white
heavy pieces are active and the position is
balanced.
90
King's Indian Warfare
29.ixf4 !ha2 30.ifl ig4 Black would obtain
a nearly-decisive advantage. Alas, my mind
was occupied with the other idea.
8
7
6
26..ixf4
The right decision.
26.gxf4? ifS 27.id3 !:!xa2! Again the blow
comes from the queenside. 28."1Wxa2 ixe4t
29.ixe4 1Wxe4t 30.mgl id4t 3 1 .Ei:f2 lt:Jh4
White loses on the spot.
5
4
3
2
1
b
a
d
c
f
e
g
h
25.tt)e4?
This centralization is a serious error.
Mter 2S .ig4 "1Wa4 (the best reply) 26."1Wxa4
!:!xa4 27.ixc8 !:!xc8 28.lt:Je4 fxg3 29.lt:Jxd6
g2t (29 ... Ei:f8 30.hxg3 !:!xfl t 3 1 .Ei:xfl !:!xa2)
30.mxg2 !:!ca8 3 I .ie3 ieS ! Black's activity
leaves the opponent with no real winning
chances.
25 ...�e7?
Returning the favour. If I had looked in
the opposite direction I might have found:
2S ..."\Wa4! This unexpected move does not
seem to be logically connected with the
previous play, but it is fully j ustified tactically.
When after: 26."1Wxa4 !:!xa4 27.lt:Jxd6 (27.gxf4
ifS 28.id3 !:!xa2+)
8
7
26 ... tt)xf4 27.gxf4
Instead 27.Ei:xf4 allows a familiar tactic, but
it's level if White is careful: 27 .. .!hf4 28.gxf4
!:!xa2!
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
The line I had spotted was: 29.1Wxa2? "1Wxe4t
30.mgl id4t 3 I .mfl "�Wh l # Did we not just
see that?
But the engine points out a miraculous escape
route: 29.lt:Jf6t!! ixf6 30.Ei:gl t mf8 3 1 ."1Wxa2
"1We4t 32J�g2 Now 32 . . .ih3? 33."1Wa8t is
winning for White, so Black should settle for
32 ... id4 when White can hold after 33.1Wb3.
6
27....if5
The pawn was sacrificed to obtain
this position. Black obviously has some
compensation (two bishops, pressure on a2
along the a-file) , and objectively it's probably
about level, but it's anything but clear.
5
4
3
2
a
b
c
d
e
f
g
h
� h 3! 28.c=.x
27 . . . ;;,
� f."<±/. Re1 ative1y b est. 28 . . . "'l..J
f':\ X f."<±/.
28 ..id3 gras 29.tt)g3! .ih3!
91
Chapter 3 - The King's Indian Bishop
Now both recaptures lead to equality:
8
35.<±>xg2 Vfffxf4 36.�a8t <±>h7=
7
6
35 .�xg2 mf8 36.�xb6 Vfffc l t 37.�gl Vfffxc4=
5
Instead after the forcing line: 30.ixh7t <±>h8
3 1 .ttJ f5
4
3
2
b
a
d
c
e
f
g
h
White has to make a tough choice to
withstand the opponent's threats.
30JUel
The position is still balanced after this move,
but perhaps now White has the tougher task
holding that balance.
After the following alternative it's still
objectively equal, but Black has to find the
more difficult moves:
30.�f2
The most principled and direct reply is:
30 . . . �xa2
30 ... h5!? is wild but also seems roughly
equal.
3 l .'�xa2 �xa2 32.�xa2 Vfie3 33.ie4 b6!
Ensuring the b-pawn does not drop with
check, but not an easy move for a human
to fi n d.
34.ig2 ixg2t
a
c
d
e
f
g
h
d
e
f
g
h
30 .. JWh4
Now White has to be exceptionally careful,
as otherwise Black's initiative will explode into
life.
31 J�e2?
b
c
3 l . . . Vfif7! 32.�gl �xa2 33.�xg7 Vfffxg7
34.ttJxg7 �xc2 35 .ixc2 <±>xg7 36.�xb7t mf6
37.�b l �a3 38.<±>gl �f3 the game should also
end in a draw.
a
a
b
b
c
d
e
f
g
h
92
King's Indian Warfare
After this mistake, White 1s definitely in
trouble.
After the cold-blooded 3 l .a4!? �xa4 32.�e8t
�xe8 33.Wfxa4 �f8 34.�xb7 ic8 (34 .. Jhf4?
35 .Wfa7; 34 ... Wfxf4? 35.�xg7t) 3 5 .�b l Wfxf4
36.�fl Wfe3 37.�xf8t �xf8 38.Wfd l ie5
39. �g2 the position would still be about
equal.
3 l .ie2 was also a good option. For example:
3 l . .. Wfxf4 (3 l . .. �xa2 32.Wfe4 and White is
OK) 32.�xb7 �xa2 33.�xg7t �xg7 34.Wfxa2
Wfxg3! 35.Wfb2t Wfe5 36.Wfxe5t dxe5 37.ifl
with a drawn rook ending on the way.
3l ...�g4! 32J�xb7?
In time pressure the German grandmaster
completely loses control.
After 32.lt:Jf5 if3t 33 .�g2 Wfg4 34.ie4
Wfxg2t 35.Wfxg2 ixg2t 36.ixg2 he would
retain some practical chances. Now the game
is over.
32 ....if3t 33.�gl �d4t
The decisive invasion by the bishop.
34. �fl �xf4 35J3fl �xa2
8
7
6
5
4
3
2
a
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
Black's triumphant position.
36.�b8t �xb8
White resigned due to 37.Wfxa2 Wfc l# (for
instance) .
0-1
Chapter 3 - The King's Indian Bishop
The following game again shows the pair of
bishops dominating. The centre is blown open,
and when White misplays the complications,
we get to see the full potential of the bishops
in action.
Boris Alterman
-
93
8
7
6
5
4
Ilya Smirin
3
Israel 1 99 1
2
Boris Alterman was one o f the best Israeli
players in the 1 990s. We played a lot of games
with each other, among them a few KID
encounters. This game was the beginning of
our KID dialogue.
l.d4 tlJf6 2.c4 g6 3.tlJc3 .tg7 4.e4 d6 5.f3
The Samisch System was Aherman's usual
choice against the KID.
5...0-0 6 ..te3 ttJ bd7
I played this move quite often until Alexey
Dreev beat me in 2002.
7.�d2 c5 8.tlJge2 a6
a
b
c
d
e
f
g
h
1 9.b4! cxb4 20.Wi'xb4 Wi'b8 2 1 .0-0 if8
22.Wi'b2±; ( 1 -0, 43) Dreev - Smirin,
Dos Hermanas 200 1 . In my opinion this
plan, releasing the tension in the centre, is
unpleasant for Black - he will find it hard to
create counterplay. White meanwhile enjoys
better piece-placement.
9 ... �a5 IO.ci>b l b5
Opposite-side casding should suit Black
well - he has already started his play against
the white king. That promises a sharp game.
l l .g4!?
A logical idea - White fights for control over
the central d5-square (after a potential g4-g5) .
A t the same time h e i s going to launch a pawn
attack against the black king.
8
7
6
5
A queen sacrifice occurred in the following
famous game of the young Mikhail Tal:
1 1 .lt:Jd5
4
3
2
a
b
c
d
e
f
g
h
9.0-0-0
Dreev used a different plan in the above­
mentioned game: 9.dxc5 dxc5 1 0J�d 1 WaS?!
(more precise here is 10 ... Wi'c7) 1 1 .tt:J c 1 Wi'c7
12.lt:Jd3 b6 1 3.if4 e5 1 4.ig5 �e8 1 5 .ie2
tt:Jf8 1 6.lt:Jd5 lt:Jxd5 1 7.cxd5 lt:Jd7 1 8.�c l Wi'd6
a
b
c
d
e
f
g
h
94
King's I ndian Warfare
1 1 . . . ttJxd5! 1 2.�xa5 ttJxe3 1 3.Ek 1 ttJ xc4
Tal mentioned in his book that before the
game he made a bet with his opponent - the
Bulgarian player Bobotsov - who was sure
that the resulting position after the sacrifice
is clearly in White's favour and even laughed
about it. Tal was of the opposite opinion and
proved to be right.
1 4.!hc4 bxc4 1 5 . ttJ c l E!:b8 1 6.�xc4 ttJb6
1 7.�b3 �xd4 1 8.�d2 �g7 1 9.ttJe2 c4 20.�c2
1 3.ttJd5 ttJ fxd5 1 4.exd5 ttJ a4 the white
king would be facing unpleasant pressure)
1 2 . . . ttJfxd5! 1 3.�xa5 ttJxe3 would lead to
interesting complications with an unusual
material balance, similar to the game
Boborsov - Tal.
8
7
6
5
8
7
4
6
3
5
2
4
3
2
a
a
b
c
d
e
f
g
h
20 . . . c3!
His attack became irresistible and the game
ended very soon:
2 1 .�d3 cxb2 22.ttJd4 �d7 23.Ei:d 1 E!:fc8
24.�b3 ttJa4 25 .�xa4 �xa4 26.ttJ b3 E!:c3
27.�xa6 �xb3 28.axb3 E!:bc8 29.�a3 Ei:cl t
30.Ei:xc l E!:xc l t
0- 1 Boborsov - Tal, Varna 1 958.
Needless to say, such a brilliant game by one
of my all-rime chess heroes inspired me a lot.
Actually this old game was probably the main
reason I started to play the 6 . . . ttJ bd7 line.
l l ... e6!?
Taking the d5-square under control.
I did not like the look of 1 1 ... b4 1 2.ttJd5 ttJxd5
1 3.exd5 or l l . . . bxc4?! 1 2.h4 with a rapidly­
developing initiative for White.
Bur 1 1 . .. ttJb6 1 2.ttJd5 (after 1 2.ttJg3 b4
b
c
d
e
f
g
h
12.tlJg3
After 1 2.g5 ttJe8 ( 1 2 . . . ttJh5? 1 3.ttJg3±;
In the case of the temporary piece sacrifice
1 2 . . . b4 1 3.gxf6 ttJxf6 1 4.dxc5 dxc5 1 5 .e5!
bxc3 1 6.�xc3 �xc3 1 7.ttJxc3 ttJ d7 1 8.f4 �b7
1 9.Ei:gl White would obtain a clear advantage
in the resulting endgame - the opponent's
pieces are rather passive and the weakness
on c5 is obvious.) 1 3.ttJg3 cxd4 ( 1 3 ... bxc4!?)
1 4.�xd4 �xd4 1 5.�xd4 b4 1 6.ttJce2 �xg5 it's
not clear ifWhire has enough for the pawn.
1 2 ...cxd4 13.hd4 b4
1 3 . . . ttJe5!? 1 4.�xe5 dxe5 1 5 .cxb5 axb5
1 6.�xb5 �a6 1 7.�xa6 E!:xa6 deserved serious
attention - on b 1 the king is under pressure
along the a- and b-files, the bishop on g7 will
soon join the play via f8, and the knight on f6
is ready to relocate via e8-d6 at the appropriate
moment. All this, in my opinion, is fully worth
just one pawn.
14.g5!?
In his turn attacking the knight.
95
Chapter 3 - The King's Indian Bishop
1 4.l[ke2 CDe5 1 5 .g5! ( 1 5.CDg1 does not look
good) 1 5 ... CD fd7 ( 1 5 . . . CDxf3? loses material
after 1 6.iWe3 CDxg5 1 7.h4 e5 1 8.ib6 CDg4
19.iWg 1 iWa4 20.b3 CD f3 2 l .bxa4 CDxg 1
22.CDxg l ) 1 6.CDg1 CD c 5 1 7.f4 CDc6 1 8.ixg7
mxg7 1 9.CDf3 �d8
15 .gxf6 t!Jxf6
For the moment Black is a piece down, but
my adversary cannot save both of his attacked
minor pieces.
1 6.tlJce2
This is of course much stronger than:
1 6.ie3? bxc3 1 7."1Wxc3 iWxc3 1 8.bxc3 �d8+
16 ... exd4 17.tlJxd4 YMb6
8
7
6
5
a
b
c
d
e
f
g
h
This would lead to a tense position where
Black should be okay. For instance, 20.h4 (or
20.b3 e5 2 l .f5 iWa3 and then 22.iWb2 aS or
22.h4? CDxb3) 20 . . . b3 2 1 .a3 iWxd2 with an
acceptable ending.
4
3
2
a
b
c
d
e
f
g
h
The position has stabilized for a while. White
has some superiority in the centre, but my
bishop on g7 is left without its counterpart. The
bishop's activity may become a major factor,
so Black should open the position somehow
on the queenside and/or in the centre. White,
in contrast, would prefer to keep those sectors
of the board as closed as possible and try to
attack the black king. Whoever executes his
plan better should be on top.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
14... e5!
Adding to the number of hanging pieces.
Bad would be 14 . . . bxc3? 1 5 .ixc3 CDxe4
16.CDxe4 ixc3 1 7."1Wxc3 iWxc3 1 8.CDxc3 ib7
19 .ie2 CD e5 20 .�xd6 CD xf3 2 l .ixf3 ixf3
22.�fl , and in this endgame White's advantage
is nearly decisive.
18.�e2?
This obvious developing move proves to be a
loss of valuable time.
Alterman should have immediately started his
kingside play with:
1 8.h4 �d8!?
Now the attempt to stop this pawn
mechanically by means of 1 8 . . . h5?! is very
96
King's Indian Warfare
dangerous due to: 1 9J!gl �d8 20.tDgf5!±
gxf5 ? 2 1 .�xg7t �xg7 22.Wg5t �h7
23.Wxf6
1 9.h5 d5!
And compared with what could have
happened in the game, White has clearly the
better version here:
20.hxg6 hxg6 2 1 .e5
8
7
6
5
4
3
2
8
7
6
b
a
5
d
c
f
e
g
h
19 . a5?
Now it's my turn to play slow chess.
4
..
3
Correct was the energetic 1 9 ... d5! at once.
White has a wide choice after that:
2
a
b
c
d
e
f
g
h
2 1 . . . dxc4
If 2 1 ... tDe8 22.c5! Wxc5 23.f4 and Black's
position is not rosy at all.
22 . .ixc4
Now bad is: 22 . . . tDd5 23.Wg5 �e8 24.�he l ±
Therefore the only way to continue involves
a piece sacrifice:
22 . . . �a7! 23.exf6 .ixf6
The next sequence of moves looks to be
pretty forced:
24.tt:Jge2 .ifSt 25.�al �ad? 26.Wh6 .ig7
27.Wh2 �f8!
The computer's move.
28 . .id3 .ixd4 29.tt:Jxd4
29 ..ixf5 .ie5!
29 ... '1lMxd4 30 . .ixf5 Wxd l t 3 1 .�xd l �xd l t
32 . .ib l �fl
And it all ends in a draw after:
33.Wh8t �e7 34.Wh4t �e8 35 .We4t �f8
36.Wxb4t �g8 37.'1lMe4 �dd l 38.b4 f5=
Quite a line!
1 8 .. J�d8 19.h4
20.cxd5? tt:Jxd5 2 1 .exd5 �xd5-+
20.tt:Jb3 aS! 2 1 .e5 (2 1 .cxd5 a4 22.tt:Jc1 a3
23.b3 tt:Jxd5 24.exd5 Wf6-+) 2 1 . . . tt:Jd7
22.Wxd5 .ib7 23.Wd6 .ixe5 24.Wxb6 tt:Jxb6+
20.e5 dxc4 2 1 ..ixc4
. . . tDc3t) 22.'1lMe l .ib7+
tDd5
(threatening
20.tt:Jc2 �b8 2 1 .b3 (2 1 .cxd5 b3 22.axb3 tt:Jxd5
23.exd5 Wxb3 24.tt:Jd4 Wxd5-+) 2 1 . . ..ib7
22.cxd5
a
b
c
d
e
f
g
h
22 ....ixd5!+ White cannot play 23.exd5 as the
f6-knight would soon land on c3.
97
Chapter 3 - The King's Indian Bishop
2 l .h6!? deserved serious attention. The main
purpose of this move is not obvious - after
2 1 . .. .ih8 White would have the option to put
his knight on g7(!) to block the dangerous
bishop with its body. 22.exd5 ib7 (the
piece sacrifice does not work here: 22 ... ttJxd5
23.cxd5 �xd5 24.'\We3
20.c5 looks like the best choice.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
20 . . . '\Wc?! Aiming at the g3-knight (wrong
would be 20 ... '\Wxc5? 2 l .e5 ttJ d7 22.f4) .
2 1 .�hgl �e8 (2 l . .. dxe4 22.ttJb5 iWd7 23.iWg5
Wff e7 24.ttJd6 with an unclear position)
22.ttJgf5 ixf5 23.ttJxf5 l.Wxc5 24.'\Wg5 \We?
Black remains better, although the situation is
still tense.
In all these variations the bishop on g7 plays a
major role.
20.h5
8
7
a
b
c
d
e
f
g
h
24 . . . ie6 [24 . . . ib7 25.ttJ df5 iWf6 26.ttJ g7! - see
previous comment] 25.ttJ df5 ixf5t 26.ttJxf5
l.Wxe3 27.ttJxe3 �e5 28.�d3 �ae8 29.�hd l
cj;>fg 30.�d8! �xe3 3 l .ib5 and White wins all thanks to 2 l .h6) 23.ttJb5 ic6 1he position
is hard to evaluate even with the computer's
help. My feeling is that Black has enough play
for the pawn.
2 1 . ..hxg6
6
8
5
4
7
2
5
6
3
4
a
b
c
d
e
f
g
h
20 ... d5!
Better late than never. This is basically the
only way to obtain serious counterplay.
21.hxg6
3
2
1
a
22.cxd5?!
b
c
d
e
f
g
h
98
King's Indian Warfare
As a result of this move I was forced to
sacrifice the knight for two central pawns,
but actually I was happy about this. Alterman
could have avoided that by sacrificing a
piece himself: 22.c5! Wff c7 (perhaps better
is 22 . . . Wffxc5 , but after 23.e5 tt:Je8 24.f4 my
pieces are very passive, for instance: 24 . . . Wib6
25.if3 lLJc7 26.Wih2 @f8 27.Wff h7 and I would
dislike this situation very much) 23.tt:Jgf5! gxf5
24.'1Ne3
Or 24.�h4 Wffc7! as the knight on g3 is
hanging. 25 .Wie3 id7! is logical, leading to a
very interesting and unusual position (worse
is 25 ... ib7 26.�gl Wid7 27.lLJb3 �c8 28.�c l ;
of course 2 5. . .Wffxg3 26.Wffe 8t if8 27.�h8t
is impossible) . 26.Wff e4 Wff c 5 27.tt:Jb3 �xd l t
28.ixd l Wla7
8
7
6
5
4
3
2
a
a
b
c
d
e
f
g
h
24.e5! Very nice. 24 ... Wffxe5 25.tt:Jc6
Wfc7 26.tt:Jxd8 Wffxd8 27.�dg l With an
overwhelming attack.
22.e5 ?! is less strong: 22 ... dxc4 23.ixc4 �a7!?
(23 ... tt:Jd5 24.Wig5!) 24.exf6 ixf6 25.tt:Jge2
if5t 26.@a l �ad7 27.Wff h6 ig7 28.Wih2 a4!?
This would lead to an unclear position with
great compensation for the piece.
22 ... llJxd5 23.exd5 gxd5
8
b
c
d
e
f
g
h
Black would get only one pawn for the
piece, but the bishop pair is incredibly strong
and White's forces are not coordinated well - I
would like my position here from a practical
point of view.
24 ....ie6!
The only move, but a very strong one. Black
appears to be fine in all variations.
Bad were the following: 24 . . . �xd4?? 25.Wffe 8t
if8 26.�h8t or 24 . . .ib7? 25.tt:Jdf5!+- or
24 . . . id7? 25 .Wie4 Wid6 26.ic4+-.
25.llJc2
A reasonable decision.
7
6
25.tt:Jdf5 ixf5t 26.tt:Jxf5 Wffxe3 27.tt:Jxe3 �e5
28.�d7 �xe3 29.ic4 �f8 would have led to
an endgame with an extra pawn and winning
chances for Black (but not 29 . . . �xf3?? 30.id5) .
5
4
3
2
1
a
b
c
d
e
f
g
Good enough was 25.Wff e4 ixd4 26.ic4
�ad8 27.ixd5 ixd5 28.Wie2 (28.Wff h 4? ixf3
29.Wih7t @f8 30.�hfl ig7!+) 28 . . . a4 with
h
99
Chapter 3 - The King's I ndian Bishop
a complicated but objectively about equal
position after the precise 29.�d2.
25 ...Y;Yc7
8
Weaker is 27 ... �d6 28.lLle4 ixa2t?! 29.<i>al
�dd8 30.f5.
28.<i>xb2 "1Wc3t 29.<i>c l ig4 30.lLle4 '!Mal t
3 l .<i>d2 "1Wb2 32.�cl �cd8 33.�b l �xd3t
34."1Wxd3 �xd3t 3 5 .<i>xd3 ie2t 36.<i>xe2
"1Wxc2t 37.<i>e3 "1Wxa2
7
6
5
4
3
2
1
b
a
d
c
g
f
e
a
h
As in some previous variations, we see that
Black has excellent compensation for the piece
here - the two bishops are very strong, White's
forces are not placed well, and the king may
easily find itself under a strong attack.
26J�xd5
It was a hard task for my opponent to make
a choice here. He could have played:
26.f4!? �c8 27.id3
27.�cl ? �c5 28.id3 iWb6!-+ and White has
no defence against 29 . . . �xc2.
b
c
d
e
27 . . . ixb2!?
Giving up another piece.
f
g
h
c
d
e
f
g
h
As is often the case, the forcing line leads
to a balanced endgame. But it was surely
impossible to find all this during the game.
Boris decided j ust to eliminate a pair of
rooks - an understandable choice under the
circumstances.
26 ...i.xd5
a
a
b
b
c
d
e
f
g
h
27J�gl?
But he errs here and immediately gets a lost
position - the price of a move in situations like
this is very high.
1 00
King's Indian Warfare
Not much better was: 27.lt:Je4? ixa2t
(27 . . . �e5 28.�d4 �xd4 29.lt:Jxd4 ixd4
30.El:d l El:d8 3 1 .El:xd4 ixa2t 32.c;t>xa2 El:xd4
33.c;t>b3 would leave White with good drawing
chances) 28.c;t>xa2 �xc2 29.�d2 b3t 30.c;t>a3
8
7
6
5
4
3
2
a
b
c
d
b
a
r�/-�.re�.ttirl:<i1lfr1
e
f
g
h
27.El:d l was the only correct continuation.
27 ... El:c8 (27 ... ixa2t? 28.c;t>xa2 �xc2 29.�b3
does not work) Alterman would have had to
find the paradoxical:
c
d
h
29 ...�c5
By attacking the rook I win a very important
tempo to play ... a5-a4, after which White is
doomed despite having an extra knight.
30J�dl
No better would be:
30.�d l b3t 3 1 .c;t>b1 a4
a
b
g
f
e
27...ha2t 28.<j;lxa2 �xc2-+ 29.�b3
Or 29.�c l �a4t 30.c;t> b l b3.
30 . . . El:b8! 3 1 .�xc2 bxc2 32.b3 ih6 33.c;t>b2
c l =�t 34.El:xc l ixcl t 3 5.c;t>xc l El:xb3 36.c;t>c2
El:b8 And I would rate Black's winning chances
here very highly.
a
d
c
e
f
g
h
28.lt:Ja l ! (28.El:d2 a4 29.b3 axb3 30.axb3
�a5 and Black's assault looks very dangerous)
This seems to lead to a dynamic balance after
28 . . . ie6 29.lt:Je4 a4 30.lt:Jd6 El:d8 3 1 .lt:Je4. No
wonder he did not find all this, and missed a
decisive blow.
b
c
d
e
f
g
h
32.�c l
Or 32.lt:Je4 �e5 33.lt:Jc3 a3 34.�xb3 axb2
35 .�xb2 (35.El:d l �e3!) 35 . . . �f5t 36.c;t>c l
�c5 .
32 ... �xc l t 33.c;t>xc l
33.El:xc l a3! Even the exchange o f queens
does not stop the attack. 34.bxa3 El:xa3
35.El:c8t c;t>h7 36.c;t>cl El:al t 37.c;t>d2 ih6t
101
Chapter 3 - The King's I ndian Bishop
38 .�c3 (38.�d3 b2) 3 8 . . . Ek l t
33 . . .Ek8t 34.�d2 ixb2
In the last game of this chapter we shall see
an early exchange of the King's Indian bishop.
Even though Black gives up a pawn on the way,
it is still a worthwhile investment, as White
suffers from a weakness on the dark squares for
the rest of the game.
30 ... a4 31 .�d3 h3t 32.®bl �e5!
8
7
Igor Naumkin - Ilya Smirin
6
Ischia 1995
5
4
3
2
a
b
c
d
e
f
g
h
Regaining the piece while continuing the
assault. Here White could easily resign, but the
next few moves were made.
33.gd2 �xg3 34.�d5 �c8 35.i.c4
This game was played on the beautiful Italian
island of Ischia, near Naples. The tournament's
atmosphere (the sea, Italian cuisine, etc.)
and environment were very inspiring, and
probably played a great part in my creative
mood during the tournament. In this game
with Grandmaster Igor Naumkin I managed
to find over the board one of my best opening
ideas i n the KID.
l.d4 lljf6 2.c4 g6 3.llJc3 .ig7 4.e4 d6 s.llJ£3
0-0 6.i.e2 llJa6 7.0-0 e5
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
a
b
c
d
e
f
g
h
35... �xc4
a
b
c
d
e
f
g
h
8.d5
White transposes into the Petrosian System.
Simple, but nice.
36.�xc4 �el t 37.�cl �xclt 38.®xcl i.h6
39.®dl i.xd2 40.®xd2 g5
0-1
8 .ie3 and 8 .El:e l , trying to keep the pawn
centre intact for a while, are the most popular
continuations here.
1 02
King's I ndian Warfare
Now the main drawback of releasing the
tension in the centre with 8 .d5 is evident - the
black knight occupies an excellent outpost on
c5. Black should not overestimate this factor
though. I have played a few interesting games
in this variation, and in this book you will find
my encounters with Zilberman in this line.
1 o .ig5 h6 ll .ie3 b6
.
.
Adding support to the knight on c5.
exchange on g4 is not advisable for White he would swap his bishop on e3 for the black
knight, and this of course would be in Black's
favour.
30 years ago I played the more natural
continuation 1 2 . . .lt:Jg4 against Ivanchuk. It
was an interesting struggle: 1 3 .ixg4 ixg4
1 4.a3 id7 1 5 .b4 lt:Ja6 1 6Jh b l f5 1 7.f3 f4
1 8 .if2 g5
12.tlJd2
8
7
6
5
4
3
a
2
1
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
1 9 .bxa5 bxa5 20.c5 g4 2 l .c6 ic8 22.fxg4
ixg4 23.lt:Jf3 if6 24.�e2 mh8 25.mh l �g8
26.�b5 lt:Jc5 27.ixc5 dxc5 28.lt:Ja4 �e7
29.lt:Jxc5 �g6 30.�d l �ag8 3 1 .d6 �g7
12 .ig4!?
...
An interesting, well-known idea. Usually the
exchange of the light-squared bishops favours
White in structures like this. There are two
main reasons:
(a) Black remains with the rather passive
bishop on g7.
(b) After . . . f7-f5 White will react with f3, and
after a probable . . . f5-f4 the bishop will go to
f2 and it will be very hard for Black to create
an attack on the kingside without the light­
squared bishop, which usually supports the
. . . g6-g5-g4 push and is also ready to sacrifice
itself around the h3 area.
But in this particular position the immediate
a
b
c
d
e
f
g
h
32.d7? (32.dxc7 �xc7 33.lt:Jd7±) 32 . . . ixf3
33.�xf3 �xg2 34.�bb l �g5 3 5 .�h3
ie7 36.�bcl �g4 37.lt:Jb7 �g6 38 .�f3
�h5 39.�c3 �g6 40.�cc l �h5 4 1 .�d3
Y2-Y2 Ivanchuk - Smirin, Pinsk 1 986.
1 03
Chapter 3 - The King's Indian Bishop
14 ... llJh5
13.f3
I have also faced 1 3 . h3 id? here.
1 3 .'tMl'd 1 ! ? was played against me by the very
talented Grandmaster Igor lvanov: 1 3 . . .
h 5 ( 1 3 . . . .id7! ?) 1 4. h3 ( 1 4.ig5 !?) 1 4 . . . id7
1 5 .'tMl'c2 lt:Jh7 1 6. ctJ b3 ctJ a6 1 7.'\Wd2 V!ff e7
1 8 .�ae 1 .if6 1 9 . ctJ c l ig5 20 .ctJd3 .ixe3
21 .'tMl'xe3
This move is the main reason behind
1 2 . . . .ig4: after 1 3.f3 the d 1 -h5 diagonal is not
controlled by the bishop on e2 anymore and
the knight goes to h S , preparing . . . f7-fS and
having the square f4 in mind.
1 5JUel
Vacating f1 for the bishop in case the knight
j umps to f4.
8
7
6
5
4
3
a
b
c
d
e
f
g
h
2 1 . ..'tMl'g5 22 .Vfixg5 ctJxg5 23 . .id 1 f5 24.f3 ctJ c5
25.ctJxc5 bxc5 ; 0- 1 (45) I. lvanov - Smirin,
P hiladelphia 1 990.
2
1
a
b
c
d
e
f
g
h
1 5 ...�f6!
13 ...�d7 14.b3
In my opinion, stronger was: 1 4.�fe 1 !? ctJ h 5
1 5 .ifl fS 1 6.b3 lt:Jf6 1 7.a3
I am proud of fi n ding this move, and the
idea connected with it, over the board.
1 5 .. .f5 was the usual continuation here.
16.�xh6
Accepting the pawn sacrifice fully passes the
ini dative to Black.
Bad was 1 6. a3 igS 1 7.if2? ixd2 1 8.V!ff xd2
ctJ xb3 .
a
b
c
d
e
f
g
h
This looks like a better version of White's
plan; (0- 1 , 47) Savchenko - Smirin,
St Petersburg 20 1 2.
I think 1 6.�ab 1 .igS 1 7 . .if2 fS 1 8.a3 was the
best continuation, with a complex struggle.
16...�g5! 17.�xg5
Immediately losing was: 1 7 . .ixf8 ? .ie3t
1 s . m h 1 os .mn Vff h4 1 9.ct:Jd 1 ct:Jg3t)
1 04
K i ng's I ndian Warfare
In the case of 1 8 .ifl �g7!, threatening . . . !!h8,
the picture would be different but still very
unpleasant for White.
18 ...£5 19.exf5
This exchange opens the g-file dangerously,
but 1 9.l!Mc 1 ttJ f4 20.g3 ttJ h3t 2l.<�g2 f4+ was
even worse - the attack would roll freely and
White would soon be doomed.
b
a
c
d
e
f
g
19 ...gxf5 20.a3
h
1 8 . . . ttJg3t 1 9 . hxg3 Wxf8 20.ttJfl Wh6t
2 1 . ltJh2 �g7 with inevitable mate - this is the
tactical j ustifi cation of my 1 4th move.
17.. JWxg5
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
20 ... a4!
a
b
c
d
e
f
g
h
The compensation for the pawn is more than
evident now - all Black's forces are active, and
after . . . f7-f5 White's right flank will be under
attack. Black may make use of the missing
h6-pawn - if the rook comes to h8 the pawn
on h2 will be under fire. In contrast, White's
pieces are quite passive. He will find it
particularly hard to give his rooks decent jobs.
l S.tlJfl
Naumkin decides to cover the h2-pawn in
advance, but now the g-pawn becomes an
object of attack.
An important finesse; Black also improves
his position on the queenside.
The preliminary 20 . . . ttJ f4 2 l .tDg3 and only
now 2 l . .. a4 was also quite strong.
21 .b4
2 l .bxa4 would not look pretty, but at least it
would not allow the immediate transfer of the
knight to d4 via b3 .
2 1 . .. tlJb3
Now Black's advantage is almost decisive.
22J�a2 gf7 23.g3 @hs! 24 ..idl ggs
Now all my pieces have joined the assault;
White cannot withstand this.
25.VNhl
1 05
Chapter 3 - The K i ng's I ndian B ishop
29.tLle2 'l;Yxc4 30.gxf4?
8
The simple 30.�xb3 would reduce my
advantage considerably.
7
6
30 .. J!xg2t 31 .®xg2
5
4
8
3
7
2
6
1
5
a
b
c
d
e
f
g
h
25 ...'l;Yh4?!
Starting here, I several times missed an
opportunity to finish the game quickly. I
should have kept my excellent knight from
being exchanged for the passive bishop on d 1 .
25 . . . lDd4 26J!f2 (26.�d3 �h4 27.�g2
lD f4-+) 26 . . . �h4 27.�g2 �fg7 and White's
position is falling apart: he cannot prevent
28 . . . lDf4, with decisive threats.
4
3
2
a
b
c
d
e
f
g
h
3 l . .. exf4?
26.ixb3 axb3 27 J!g2!
Now it is my turn again to make a serious
mistake. 3 l . . . lDxf4t 32.lD xf4 exf4 33 .�xb3
(33.�e4 �xd5-+) 33 . . . .ih3t would win by
force.
I did not notice this defence, counting on
27.�xb3? f4 28.ltJe4 fxg3 29. hxg3 lDxg3-+ .
32J!e4
Or 32.�xb3 �g7t 33.\t>h l �xb3! 34.�xb3
.ib5 3 5 .�b2 �e7 and one of the white knights
is gone.
27...'l;Yd4t 28J!e3
32 ...'i;Yxd5
Despite the mutual
position remains lost.
mistakes,
White's
33.<i>fl Y;Ygs
More resolute was 33 . . . c5! 34 .tDc3 (34.\t>el
c4) 34 ...�g5 3 5 .�xb3 �g7 36.\t>el ttJ f6
37.�e2 �h4t 38.\t>d2 .ih3 .
34.®el tLl f6?!
a
b
c
d
e
f
g
h
28 ... f4?!
Stronger was 28 . . . lD f4 29.�d2 �xc4.
Another inaccuracy. 34 . . . d5 3 5 .�xb3 �f5
36.�d4 �e5! would break all resistance on the
spot.
106
King's Indian Warfare
38 .. J�e7! 39.'1Nxb3t <i>g7
8
The position has simplified and my opponent
even has an extra pawn at the moment. But
despite numerous lapses, Black's advantage is
still large . . .
7
6
5
4oJ�c4 YNgi t 4t .<i>d2 YNxh2 42.'1Nd3
4
3
8
2
7
1
a
b
c
d
e
f
g
h
35.'1Nb2?
I missed this move, but much stronger was
35 .�xb3! when Black must settle for a clear
advantage after: 35 . . . d5 36.!hf4 �e7 37.�xf6
�xf6 38 .�xd5+
6
5
4
3
2
1
But not 3 5 .�xf4? �e7-+.
a
b
c
d
e
f
g
h
42 ... c5! 43Jk l <i>f7 44.bxc5 bxc5
35 ... <i>g8?
And here the immediate 35 . . .ib5 ! was a
winner.
36J�xf4 .ibS 37.tlJfg3 1xe2 38.tlJxe2
38 .�f5 fails to 38 ... �e3 39.t2Jxe2 �e7!
40.�xf6 �g l t 4 I .md2 �xe2t and the queen
is lost.
Although there is not much material left on
the board, Black's advantage is still decisive:
the white king is stuck in the centre, all my
pieces are more active than White's, and finally
the mighty pawn duo on d6 and c5 is much
more powerful than the isolated pawns on a3
and f3. No wonder the end is near.
45.a4
8
Or 45.�el d5 46.md l �f2 leaves White
helpless against the advancing pawns. For
example, 47.l!gl �xe2 48 .�g6t me6-+ and
White has already run out of good checks.
7
6
5
4S .. J�e5! 46.'1Nb5 YN£2 47J�c3 d5
4
3
2
1
a
b
c
d
e
f
g
h
The pawns have started to roll and Naumkin
decides to call it a day due to 47 . . . d5 48.�xc5
�xe2t 49.�xe2 �xc5 . I made quite a few
errors in the second part of the game, but
despite that, this remains one of my most
beloved KID encounters.
0-1
Chapter 4
Line Opening
In the King's Indian it is important to look
beyond apparent obstacles to success...
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering
these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
� 8
7
6
5
4
3
2
2
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
How can Black continue the
attack? (see page 1 38)
This is a bit easier. A piece down,
the attack is still very strong.
(see page 1 22)
8
7
� 8
� 8
6
6
7
6
5
7
5
4
5
4
3
4
3
2
3
2
a
b
c
d
e
f
g
h
2
a
Deep analysis. Black has a nice
difficult combination.
(see page 1 1 6)
b
c
d
e
f
g
h
a
d
e
f
g
� 8
6
6
4
4
2
2
One of the most imaginative
combinations in KID history
(see page 1 1 7)
f
g
h
initiative? (see page 1 42)
� 8
h
e
initiative? (see page 1 26)
7
5
3
c
d
How can Black fight for the
5
b
c
How can Black fight for the
7
a
b
3
a
b
c
d
e
f
g
h
Let your intuition guide you!
(see page 1 37)
a
b
c
d
e
f
g
h
How can Black fight for the
initiative? (see page 1 48)
1 09
Chapter 4 - Line Opening
One of the most important themes in the
King's Indian is line opening. From the
beginning of the game Black allows White to
take more space. The trade-off is that White is
advancing on the queenside and Black on the
kingside, where the king is. But this does not
fully compensate for the lack of space. Black
needs to open lines for his pieces, often with a
sense of urgency, in order to get the most out
of the position's potential.
This is probably the biggest topic in the
King's Indian and a natural follow-up to our
discussion of the Ki ng's Indian bishops, where
the main characteristic was the great potential
this bishop holds. In this chapter we shall
investigate the potential of all of the pieces.
As this topic is rather important to the
understanding of the King's Indian, the
introduction is longer than in all the other
chapters and somewhat generic. To some this
will probably mean seeing examples or ideas
they already know intimately, while for others
this will be essential background knowledge
to understand the more complex games that
follow.
We start with the most generic of all K ing's
Indian games, played in a famous Candidates
Tournament on which two great books have
been written, by Bronstein and Najdorf. I have
read the book by Bronstein maybe 20-25 times
and it has helped me a lot.
Mark Taimanov- Miguel Najdorf
Zurich 1953
l.d4 �f6 2.c4 g6 3.tlJc3 i.g7 4.e4 d6 s.ttJf3
0-0 6.i.e2 e5 7.0-0 tlJc6 8.d5 tlJe7 9.tlJe l
�d7 IO.i.e3 f5 l l .f3 f4 12.i.f2 g5 13.tiJd3
This idea, forcing through c4-c5, is no
longer popular.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
This looks a bit slow. As I understand the
King's Indian , Black should already be O K
now.
16 ...i.ffi
This is the modern system, invented
by Gligoric earlier in the same year in a
tournament at Mar del Plata. This name has
now been extended to cover all of the lines
after 8 . . . tt:Je7, as in the recent two-volume
work on this position by Vassilios Kotronias.
17.cxd6 cxd6 1 8.�d2 g4 1 9.�fcl
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
An important moment in the history of the
King's Indian Defence. Najdorf comes up with
a wonderful pawn sacrifice that since then has
become as much a staple as bread and butter.
1 10
King's Indian Warfare
19 ... g3! 20.hxg3 fxg3 2 I ..ixg3 tlJh5 22 ..ih2
.ie7
8
ixb5 3 1 .�a3 �g7 32.g4 ixd2 33 .�xd2 �xd2
34.�xd2 lDxf3 35 .�f2 ttJxh2 36.�xf8t �xf8
37.�xh2 ttJf6 and Black is close to winning.
28 ... b5 29.a4 a6 30.axb5 axb5 3 l .gc7 gg7
32.tlJ b3?!
7
4
The fi n al mistake, though it should be said
that White is under a lot of pressure anyway.
3 2 .ie2 ttJ gf4 3 3. g4 ttJ xd3 34 .ixd3 ttJ f6 also
gives Black a winning attack.
3
32 ... tlJh4 33.gc2 .ih3
6
5
33 . . . �xf3 was a bit more obvious, but we are
in the territory of "anything wins" .
2
1
a
b
c
d
e
f
g
h
Black's counterplay on the dark squares is
hopefully obvious to all. A few moves ago it
was hard to see how the black pieces were going
to get into the game. But through line opening
and a small material investment, we now have
four active pieces with great potential.
The rest of the game is well played by Najdorf
Today we might see a more dynamic approach,
but we should not fault the players for playing
as if they did not know Tal's games, as Tal was
still playing j unior tournaments in Riga at this
point!
23.tlJ bl
It was stronger t o play 23 . tD d 1 i g 5 24.tDe3,
although Black has a good game after 24 ... �f8 .
23....id7 24.�el .ig5 25.tlJd2 .ie3t 26.�hl
�g5 27 ..ifl gaf8
Black has the advantage. All his pieces are
playing their part.
34.�e2
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
34 ... tlJxg2! 35 ..ixg2 .ixg2t 36.�xg2 �h4
37.�xg7t �xg7 38.gg2t �h8 39.tlJel tlJf4
40.gg3 .ifl 41 .gg4 �h3 42. tlJ d2 h5 43.gg5
0-1
Grandmasters today have intimate knowledge
of how important it is to fight for the initiative
when facing the King's Indian, but we still see
wipe-outs like this. Some of them are below.
28.gdl
28 .�a 1 b5 29.a4 was analysed by Kasparov in
Part IV of his Great Predecessors series. He gave
a good line in reply, but he did not consider
the most direct approach: 29 . . . ttJ h4! 30.axb5
The following example from the Soviet
Women's Championship shows another typical
line opening: the check on g3 with a knight to
open up the h-file to the king.
111
Chapter 4 - Line Opening
Tatiana Shumiakina - Narine Karakashian
Volzhsky 1989
This is of course the point. The queen is
heading for h4 with great urgency.
2 I .lbxe3 fxg3
8
With the bishop gone, we need someone else
to assist with the attack on the dark squares.
7
6
22.i>gl
22.ttJg4 �h4t 23.�gl ttJ e3 and mate.
5
4
22 ... �xe3 23.i.xe3 Wfh4 24J�el Wfh2t
25.i>fl
3
2
8
1
a
b
c
d
e
f
g
h
Black has achieved a lot, getting her bishop
to e3, outside the pawn chain. But if White
has time to play lt:J c4, it will all be for nothing.
The bishop is exchanged and White is ready
to attack on the queenside. Black has to react
energetically.
7
6
5
4
3
2
1
18 ... lbh5! 19.lbc4
This allows Black to unfold her attack.
1 9 .�e l is met with 1 9 . . . ttJ f5!!, when White
has to play 20J�f2, which is obviously not
good, in order to avoid 20.exf5 ttJ g3t 2 l .hxg3
�g5 with mate.
19 ... lbg3t 20.hxg3 lbf5!!
a
b
c
d
e
f
g
h
25 ...i.h3!
The final piece comes into the attack, making
up for the fact that I have chosen not to include
any . . . .ixh3 examples in this chapter, fi n ding it
j ust a bit too generic.
26.gxh3 g2t
8
I have to say that I enjoy 26 . . .�h l t 27 . .ig l
�xh3# more. It is purer somehow. But mate is
mate and it wins the game.
7
6
0-1
5
As I said, there is no pushing the pawn to g3
with . . ..ixh3 to follow in this chapter, but I
have something almost as good for you.
In the next example, we have a non-generic
break, but otherwise a very typical reaction.
I want the reader to understand that it is
4
3
2
a
b
c
d
e
f
g
h
King's Indian Warfare
1 12
important both to know the typical ideas and
to get a sense of the dynamic attitude necessary
to play the King's Indian well.
20 . . . e4! 2 1 .lt:Jxe4 ifS , when all his pieces
are playing. For example: 22 .0-0 lt:J a4
23 .Ei:b3 ixe4 24.fxe4 1Mfe5 2 5 .ixe3 Ei:xfl t
26.ixfl lt:J c3 with a big advantage.
Igor Bjelobrk - Alexander Grischuk
20 ... i.f5 21 .tlJe4 YNa4!
Troms0 (1.1) 20 13
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
a
b
d
c
e
f
g
h
White has played a little slowly and is in
need of "just one more move" to have a good
position. If Black were to act timidly here,
White would have time to make this extra
move, but against an experienced King's Indian
maestro like Grischuk, there is no such luck.
19 ... e3!! 20.i.xe3
20.f3 would have allowed Black to play the
thematic:
a
b
c
d
e
f
g
h
22.YMd3
22.1Mfxa4 lt:Jxa4 and the knight is heading for
c3 in many lines. Besides the loss of at least a
pawn, the white position is full of holes. For
example: 23.f3 ixe4 24.fxe4 lt:Jc3 2 5 .Ei:c1 lt:Jxe4
22 ...YNa2 23.i.xb6
2 3 .ic 1 ixe4 24 .1Mfxe4 lt:Ja4 is also decisive.
For example: 2 5 .id2 lt:J c3! 26.ixc3 Ei:f4
27.1Mfd3 e4 and White's position collapses.
23 .. J�a3 24.i.e3 gxd3 25.i.xd3 i.xe4
26.he4 i.h6 27.0-0 i.xe3 28.fxe3 gxfl t
0-1
8
7
6
We shall continue the journey from the
classic ideas into the theme of dynamic line
opening. The following example is particularly
brilliant in the sidelines, where we can see the
importance of making space for the pieces.
5
4
3
2
a
b
c
d
e
f
g
h
1 13
Chapter 4 - Line Opening
Giovanni Vescovi - Teimour Radjabov
Bursa 2010
8
7
6
Making space for the second rook and
threatening . . . e4-e3 at the same time. 27 .ifl
E!:ag 8 28.E!: e 1 :8:8g6 29.E!:d2 tD d3 and White
needs to do something desperate, and give up
something he does not want to give, in order
to avoid being mated.
24 .. J�g8 25 ..ifl ttl d3
The black pieces are flowing in effortlessly.
5
4
26.ttlb5
3
26.\We3 also does not offer a serious defence.
The simplest line is probably 26 . . .\Wxe3t
27.El:xe3 id4 and the attack continues
unhampered, despite the exchange of queens.
For example: 28.lD d 1 lDxb2 29.lDxb2 ixe3t
30.<i>h2 id4 3 1 .E!:b 1 E!:g6 and White cannot
avoid further material losses.
2
a
b
d
c
e
f
g
h
Black has given up a piece and caused real
damage, but needs to find a way to bring the
remaining pieces into the attack.
23.. <i>f7!
.
Creating a path for both of the rooks to
enter the attack.
26....ih6!
26 . . . lD e5 ! ? is another typically atypical
King's Indian tactic that also gives Black a
winning attack.
27.�e3
24JUel
The important line to foresee is 24.\We3
id 4!!. A key tempo-winning move. This time
aro und, the King's Indian bishop is rather
useless, but it can be given away to win time
and clear the g-file for the rooks. 25 .\Wxd4 E!:g8
26J�f2 E!:g3!
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
27 ... ttlxf4!
A decisive tactic.
a
b
c
d
e
f
g
h
28J�xh3 ttlxh3t 29.<i>h2 .ixd2 30.<i>xh3
�g5 31.�dl gagS 32.�xd2 ghst
1 14
King's I ndian Warfare
33.liJh4 gg4 34.�h2 ggxh4t 35.�gl ghl t
White resigned. After 36.@f2 �5h2t 37.ig2
f4 the game is over.
28.Wc5 if8! and White suffers devastating
material losses.
0-1
In some of the examples above White has been
a bit overmatched. In the next examples, two
of the best players of our time are blown away
by - you guessed it - dynamic line opening.
Levon Aronian- Magnus Carlsen
Nice (blindfold) 2010
8
a
7
b
d
c
e
f
g
h
22 ... e4!!
6
The great point of the sacrifice. White has to
take the pawn, which gives Black time to bring
in the King's Indian bishop.
5
4
3
23.liJxe4 �d4t 24.gf2
2
24 .CLJf2 Wf6 is entirely decisive. The white
pieces are all tied up.
a
b
c
d
e
f
g
h
Magnus Carlsen is best known for his
endgame skills, but make no mistake about it,
Carlsen can play all types of positions. Here
he shows his skills as an attacking player in a
blindfold rapid game, a few months after he
took the Number 1 spot in the rating list.
24 ... gxf2
Black could have accelerated his attack a bit
with 24 . . . ixf2t! 2 5 .CLJ xf2 Wf6 26.Wfl �ae8
27.�e l CLJe3 28.ixe3 �xe3, where White is
defenceless.
18 ... fxe4 19.fxe4 cxd5 20.exd5 llJexd5!
For a King's Indian player it is easy to give
away a piece to open up lines for the rest of
the pieces.
21 .cxd5 liJxd5 22.�gl
22.@h3 CLJ f4t 23.ixf4 �xf4 gives Black a
winning attack. A line that a human might
calculate would go something like this: 24.Wd3
Wg5 25.�c7 Wh6t 26.@g3 e4 27.Wxd6 �f6
a
b
c
d
e
f
g
h
For example: 29.CLJ d l �g3t! 30.hxg3 Wd4t
3 1 . CLJ f2 We3 ! and White can do nothing.
Chapter 4 - Line Opening
25.tiJxf2?!
Aronian could have made things more
difficult with 25 .ig5 ! , when Black plays:
25 . . . if6 26.ixf6t (26.ctJ xf6?! loses to the
beautiful 26 . . . ctJ e3!) 26 . . .!hf6 27.�d4
115
8
7
6
5
4
3
2
a
a
b
d
c
e
f
g
h
27 . . . ctJ f4!! 28 .ifl ixe4 29.g5 ctJ e6+ and Black
will fight for a win with an extra pawn.
25 ... %Vh4 26.%Vel �Us 27 ..if3 �xf3 28.%Ve4
�xflt
0-1
Nakamura has won with Black in the King's
Indian against many of the top players. You
can look up his wins against Anand, Gelfand
and Kramnik, should you feel so inclined. For
our topic, his most recent example against his
new compatriot and Top- 1 0 colleague will do
just fine.
Wesley So
-
b
c
d
e
f
g
h
Black has managed to regroup in the
traditional manner without White managing
to create serious threats on the queenside. This
allows Black to crash through.
24...h4!
Preparing . . . h4-h3 at the right moment.
25.fxg4
So decides to take up the challenge.
The engine suggests 25 .ib4 is a better move,
bur there is something artificial about it. Black
does not have to play 25 . . . h3 immediately, and
he might increase the pressure further before
executing the attack. In short, I do not believe
that White is OK here.
Hikaru Nakamura
Sinquefield 2015
l.d4 ttlf6 2.c4 g6 3.ttlc3 .ig7 4.e4 d6 5.ttlf3
0-0 6..ie2 e5 7.0-0 ttlc6 8.d5 ttle7 9.ttlel
ttld7 10.f3 f5 l l ..ie3 f4 12 ..if2 g5 13.ttld3
ttlg6 14.c5 ttlf6 15.�cl �f7 16.�hl h5
17.cxd6 cxd6 18.tiJb5 a6 19.ttl a3 b5 20.�c6
g4 21 .%Vc2 %¥£8 22.�cl .id? 23.�c7 .ih6
24..iel
a
b
c
d
e
f
g
h
1 16
King's I ndian Warfare
Now Black sacrifices material in order to
open up the position for all his pieces, after
which White is defenceless.
25 ... f3! 26.gxf3 �xe4!!
3 1 .ixd2 ttJ f4 32.iel ctJ f2t!! 33.Wfxf2 Wfe4t
and everything falls.
The best chance, 27.ctJb4 �xf3+ still gives
Black an overwhelming advantage, as White
cannot play 28 .Wfxe4?! on account of 28 . . . �fl t
29.�g2 h3t! 30.�xh3 ttJ f4t and it is all over.
27.. J3xf3! 28.gxd7 �Ul t 29. ®g2 ie3! 30.ig3
30.ixfl h3t is the same as before, although
after 3 1 . �xh3 Wff3t 32.ig3 it is the knight
that delivers checkmate: 32 . . . ctJ g5#
30 ... hxg3 31 .�xfl ltJh4t
a
b
d
c
e
f
g
h
Dynamic line opening in action. All the
black pieces are coming to life.
27J�dl
It is easy to criticize this move, but hard to
come up with an alternative.
27.fxe4 �fl t 28.�g2 fails to 28 . . . ie3!!,
when the attack is decisive. The main point is
29.ixfl h3t! 30.�xh3 Wff3t 3 1 .ig3 ixg4#.
A more complicated variation is 27.�xd7 �xf3
28.ixf3 Wfxf3t 29 .Wfg2 Wfxd3 30.�dl .
a
b
c
d
e
f
g
h
The knights are terrific attackers. White
could resign, but he found it more
sportsmanlike to play out the attack.
32.®h3 %Vh6 33.g5 ltJxg5t 34.®g4 �hf3
35.ltJf2 %Vh4t 36.®f5 �f8t 37.®g6 �f6t
38.®xf6 ltJe4t 39.®g6 %Vg5#
0-1
a
b
c
d
e
f
g
h
White seems all right until you find 30 . . . id2!!
This chapter is about line opening, but I also
feel it is necessary to talk for a moment about
dynamic line blocking, with one last example
before we go from the glorious theory to the
reality of messy practical play. This game is like
something from out of this world, involving
both dynamic line opening and dynamic line
blocking.
Chapter 4
-
Per Auchenberg- Brian Elgaard
1 17
Line Opening
White continues with his plan.
Copenhagen 1992
35.cxd7 i.£8!!
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
White has made great progress on the
queenside, but Black is not without his
chances. The key task is to find a way to get
the queen into the attack. It all comes together
like a beautiful symphony, with three bizarre
moves.
34 ...YMg8!!
Yes, the queen is hanging, but it
immune on account of . . J�fl#.
IS
also
a
b
c
d
e
f
g
h
Closing the 8th rank for j ust a moment and
opening the g-file for the queen.
36J3xc8
It is a pity that White did not take with the
pawn, but at least he allowed his opponent to
execute the checkmate.
36... tiJhit
Knight checks
special.
m
the corner are always
37.®e2 YMxg2t 38.®d3 YMc2#
0-1
King's I ndian Warfare
1 18
My Games
Alexander Huzman
-
Ilya Smirin
Sverdlovsk 1987
l .�f3 �f6 2.c4 g6 3.�c3 .ig7 4.e4 d6 5.d4
0-0 6 ..ie2 e5 7.0-0 � c6
The most popular continuation among such
moves as 7 . . . ttJ a6, 7 . . . ttJ bd7, 7 . . . exd4, 7 . . . "!9e8
and 7 . . . c6.
8.d5 �e7 9.�el �d7 IO .ie3
•
One of the most aggressive setups for White
against the KID, and a pet variation of Viktor
Korchnoi.
1 o . ttJ d3 f5 1 1 .f3 f4 is another popular line; a
slightly more "quiet" one.
next day in the Russian language newspaper
"Soviet Sport". I was lucky, as during the lunch
break, only a few hours before the game, Boris
Gelfand enlightened me about the brilliant
idea played by Korchnoi.
1 3 . tD d3 ttJ f6 1 4.c5 tLlg6 1 5 .Ek 1 :gf7 was
played in Taimanov - Najdorf, Zurich 1 953,
as mentioned in the introduction.
13 ... b6!
Forewarned is forearmed, even though I did
not expect that my future Israeli teammate
Alexander Huzman would follow Korchnoi's
footsteps. Here is his game:
1 3 . . . a6 1 4.tLla7!!
1 o ... f5 I I .f3 £4 12 ..ifl g5
8
7
6
5
4
a
3
2
1
a
b
c
d
e
f
g
h
1 3.�b5!?
An interesting story is connected with this
move. It was first played by Korchnoi against
Hulak j ust two days before my game with
Huzman. There were neither databases nor
internet in those ancient days, so information
usually reached players much more slowly than
nowadays. However, the Korchnoi - Hulak
game was an exception - it was published the
b
c
d
e
f
g
h
A really beautiful move!
1 4 . . . :gxa7 1 5 .ixa7 b6 1 6.b4 ib7 1 7.c5 dxc5
1 8 .:gcl tLlc8 1 9 .bxc5 ia8 20.c6 ttJ f6 2l.ixb6
The bishop falls in the end, but the price is
very high.
2 1 . . . tLlxb6 22.ixa6 g4 23.tDd3 g3 24.h3 tDe8
25.tLlc5
White won in Korchnoi - Hulak, Zagreb
1 987.
I did not want to repeat Hulak's fate, and
found the best move 1 3 . . . b6 over the board.
14.b4 a6
1 19
Chapter 4 - Line Opening
Black starts his usual routine in this variation
- activity (attack) against the opponent's king.
However, in this case I would prefer 1 5 . . . �f6,
with the further transfer of this rook to g6 or
h6, or 1 5 . . . tt:J f6 1 6.c5 b5, which I like most. In
contrast, after 1 5 ... a5 ?! (activity on the wrong
flank) 1 6.c5! axb4 1 7.cxd6 cxd6 1 8 .tt:Jc4 lt:Jc5
1 9 .lt:Jd3 White's chances should definitely be
preferred - he is far ahead in developing an
initiative.
8
7
6
5
4
3
2
a
b
d
c
e
f
g
h
15.�a3
In my opinion, retreating this knight to c3
is more logical, even though Black is usually
doing OK after that. Below is another of
Korchnoi's games - this time he was on the
receiving end of Black's attack:
1 5 .lt:Jc3 h5 1 6. �h 1 tt:J f6 1 7.c5 g4 1 8 .cxb6 cxb6
1 9.Ek 1 g3 20.ig 1 gxh2 2l .if2 h4 22.lt:Ja4
�b8 23.b5 axb5 24.ixb5 lt:Jh5 25.�xh2 lt:Jg3
26.�g1 lt:Jg6 27.lt:Jd3 �h7 28.lt:Jb4 h3
16.c5 b5
Now it's evident that the knight on a3 is
preventing White from playing a2-a4, so he
must move it first - and time is very valuable
in such positions.
17.lLlac2 �f6 1 8.a4 bxa4 19J��xa4 �g6
20.b5 g4 2 1 .�b4
8
7
6
8
5
7
4
6
5
3
4
2
3
1
2
a
1
a
b
c
d
e
f
g
h
29.�xc8!?
29.gxh3 ixh3! 30.�xh3 �g5! ! gives Black a
winning attack.
29 . . . �xc8 30.gxh3 �h8 3 1 .ixb6 �e7
32.lt:Jd3 ?! �b8!
Black has a big attack and won in Korchnoi
-Ye Jiangchuan, Novi Sad 1 990.
15 ...h5?!
b
c
d
e
f
g
h
2 1 . .. g3!
After the famous game Taimanov - Najdorf
(see above) this pawn sacrifice became typical
in such positions.
22.hxg3 fxg3 23.�xg3 h4
Black's activity is based mainly on the weak
dark squares around the white monarch. Black's
knights strive to reach f4 and g3 ; the bishop
will support them from h6 and the g-file is free
for Black's heavy artillery. The question is: will
1 20
King's I ndian Warfare
Black have enough time to do all these things,
or will the opponent smash Black's queenside
and/ or centre first?
24.lbc6 V:Yd7 25 ..ih2 .ih6
After which the following complications are
possible:
27 . . .rnhs 2s .b6
28Jh 2 �g8 29.cxd6 cxd6 30.b6 lD 6h5
8
8
7
7
6
6
5
4
5
3
4
2
3
2
a
1
a
b
d
c
e
f
g
h
Now 26 . . .ie3t 27.rnhl lDh5 is threatened.
26.£4?!
It's easy to understand the motivations
behind this drastic decision: the black knight
is deprived of the h5-square and White's pieces
on the kingside have more space. On the other
hand, White has returned the sacrificed pawn
(though he will immediately take the pawn
on a6) , one black knight is already en route to
the excellent outpost at f4, and the e4-pawn
and g4-square have become much weaker.
I think the more principled (and stronger)
continuation was:
b
c
d
e
f
g
h
3 1 .�b2 (the following excmng vananons
show the strength and danger ofBlack's direct
assault against the white king: 3 I .lDa5 tD h3t
32.rnhl lD g3t 33 .ixg3 hxg3 34.b7 ie3
And now 35.bxa8=W Wh7-+ or 3 5 . bxc8=W
�axc8 36.Wb7 �c7 37.Wxc7 tD f2t 38.�fxf2
gxf2 39.�xf2 Wxc7.) 3 I . . . lD h3t 32 .rnhl
lD g3t 33 .ixg3 hxg3 34.b7 (34 .gxh3 Wxh3t
3s .rngi if4 36 .�g2 �g7-+)
8
7
6
5
4
3
2
26.Wb3 tD f4 27.ic4
8
a
7
b
c
d
e
f
g
h
34 . . ..ie3!! 35.bxc8=W �axc8 36.Wxe3 tD f4
37.Wa7 WeB and despite two extra pieces
White is helpless to prevent mate after
. . . Wh5 .
28 ... �g8 29.bxc7
29.b7 seems to be losing after 29 . . . .ixb7
30.Wxb7
6
5
4
3
2
a
b
c
d
e
f
g
h
121
Chapter 4 - Line Opening
34.gxh3
The only move in this wild and picturesque
position. If 34.�c3 ? �g7 3 5 .d7 ixd7
36.ixa6 .id2! White gets mated in 6
according to the silicon brains.
34 . . .ie3 3 5 . t2J f7t <;t>g7
And after the forced:
36.t2J g5 ! ixg5 37.�b2t if6 3 8.e5 �xh3t
39 .<;t>gi �h2t 40.�xh2 gxh2t 4 I .<;t>xh2
ixe5t 42.f4 �h8t 43 .<;t>g3 ixd6
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
30 . . . t2J xg2! 3 I .t2Jxg2 h3 32.�a2 ie3t
33 .<;t>hi hxg2t 34.�xg2 �xg2 3 5 .<;t>xg2 �g8t
36 . .ig3 �h7 37.�hl �xg3t 38 .<;t>xg3 �xhl
39 .�c8t <;t>g7 40.�xc7t <;t>g6 4 1 .�xd6
�gl t 42.<;t>h3
8
7
6
5
4
8
3
7
2
6
b
a
5
4
3
2
a
b
c
d
e
f
g
h
42 . . . if4 43 .t2J xe5t <;t>h5-+ Quite a journey
for the black king!
29 . . . ttJ 6h5 30.cxd6 t2J h3t 3 I .<;t>hi t2J g3t
32.ixg3 hxg3 33.t2Jxe5
33.t2Je7 t2J f4!-+
33 . . . �h7
d
e
f
g
h
There arises a complex endgame, the most
probable result of which is a draw, despite
White's extra pawn (the pawn on c7 will fall) .
I apologize for showing such long and possibly
not error-free variations. Usually I try to avoid
that, but on this occasion these variations
are very important for understanding the
whole picture. Also I hope you will find them
interesting and even paradoxical.
26 lttxf4
...
8
8
7
7
6
6
c
5
5
4
4
3
3
2
2
a
b
c
d
e
f
g
h
1
a
b
c
d
e
f
g
h
1 22
King's I ndian Warfare
27.bxa6?
In such a sharp and unbalanced position this
natural move, which would probably be made
by the vast maj ority of players, proves to be a
mistake.
After the immediate 29 . . . ttJ d7! (my opponent
saw this move during the game) 30.cxd6 ttJc5!
3 1 .tLle7t mh7 32.Wa3 tLlxa4 Black would
have a decisive advantage.
30J�xa6 ha6 3 1 .Y!lxa6 Y!lg3
Correct was 27.ttJd3!, with an unclear battle.
8
27... Y!lg7 28 ..ixf4
The knight on f4 is very annoying, so
Huzman decided to exchange it. But now the
bishop, which replaces the knight, becomes
really menacing. It's hard to suggest something
better, for instance:
2s.mh 1 tt:J g4 29.igl (29 .�a3 tt:J h3!-+) 29 . . .
h 3 , and from hl the king stares i n fear a t his
numerous enemies.
7
6
5
4
3
2
1
a
28 ....lxf4
b
d
c
e
f
g
h
32.Ylld3?
A serious mistake. After 32.if3 ie3t
(32 . . . h3? 33.ttJe7t mh7 34.lt:Jf5 Wh2t 35 .mf2
�g8 36.Wc4+-; 32 . . . ttJxe4 33 .ixe4 Wh2t
34.mf2 Wg3t 3 5 .mgi=) 33.mhl h3
8
7
6
5
8
4
7
3
6
2
5
4
a
b
c
d
e
f
g
h
29.Ylld3
3
2
Again, it's not easy to offer a better move.
a
After 29.cxd6 Wg3!? 30.�xf4 Wxf4 White is in
trouble.
29 .. J�xa6?!
A tempting but, alas, wrong sacrifice.
b
c
d
e
f
g
h
34.We2! (the queen is back in time) 34 . . . if4
35 .gxh3 Wxh3t 36.mgi tt:J g4 (36 . . . mh7
37.�f2) 37.tLlc2 ih2t 3 8.mhl the game
would have ended in a draw.
32 ... .le3t 33.g;,hl h3! 34J�gl?
Th e last error in time trouble.
1 23
Chapter 4 - Line Opening
Yuri Dokhoian - Ilya Smirin
34.E!f3 �xel t 3 5 .ifl still would have allowed
White to hold on, even though Black is much
better.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
34 ... <i>f7!
Now the rook will join the attack from h8
to threaten checkmate. White cannot prevent
this, so he resigned.
0-1
Sverdlovsk 1987
This game, as was my encounter with
Huzman, was played in a competition that
was very memorable for me. The first league of
the USSR Championship in Sverdlovsk (now
Ekaterinburg) was my first really big success.
Despite being a newcomer to tournaments
of this calibre, I took clear first place ahead
of such players as Tseshkovsky, Khalifman,
Gelfand, T ukmakov and Malaniuk. Before the
tournament I had not touched chess or even
thought about it for around two months. The
reason - I was in military service from 1 986-88.
By winning this first league I qualified for
the 1 988 USSR Championship. It was a
super-tournament with the participation of
Kasparov, Karpov, Yusupov, Salov, Ivanchuk,
Beliavsky, Vaganian and others, but that's
already a different story.
I .d4 tlJf6 2.c4 c5 3.d5 g6 4.tlJc3 i.g7 5.e4
d6 6.f4
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
The aggressive Four Pawns Attack (with
a slightly different move order) - a sign that
White is going for the maximum in the
opening. He would love to smash his opponent
with his powerful centre.
1 24
King's I ndian Warfare
6.. 0-0 ?.tl)f3 e6
8 ... fxe6
As a teenager, a couple of times I played
a pawn sacrifice in the spirit of the Benko
Gambit - 7 . . . b5 8.cxb5 a6 9.a4 W/a5 - but
without much success.
Now Black's pawn formation in the centre
becomes stronger.
I like the other possible capture 8 . . .ixe6
less, but perhaps it's just a matter of taste.
8.dxe6
9.i.d3
.
More common is:
8 .ie2 exd5 9.cxd5
9.exd5!?
9 ... ig4
9 .. J%e8 1 0 .e5 leads to very deeply analysed
complications. To play such lines requires
one to have a lot of theoretical knowledge
and a good memory.
1 0.0-0
The following is an inspmng (for Black)
game by the great Mikhail Tal :
1 0 . . .ixf3
Modern theory recommends 1 0 . . . ttJ bd7.
1 1 .ixf3 ttJ bd7 1 2 .\t>h 1 a6 1 3 .ie3 �e8 1 4.g4
h6 1 5 .g5 hxg5 1 6 .e5
8
7
6
5
4
3
2
1
a
b
d
c
e
f
g
h
9 .. e5!?
.
Over-the-board improvisation.
a
b
c
d
e
f
g
h
The main continuation here is:
9 . . . ltJc6 1 0.0-0 ltJd4
I'd like to mention a game by another World
Champion:
1 1 .ltJg5
More promising looks 1 1 .ltJxd4 cxd4
1 2 .ltJb5 and Black has to solve certain
problems.
ll. . . e5 1 2.f5 h6 1 3 .ltJh3 gxf5 1 4.exf5
1 6 . . . gxf4! 1 7.exf6 �xe3 1 8 .fxg7 lLl eS 1 9 .ig2
WigS 20.ltJe4 W/h4 2 1 .Wfd2 \t>xg7 22.Wff2
W/xf2 23.�xf2 f3 24.ltJxd6 �d8 25.lLlxb7
fxg2t 26.\t>xg2 �xd5 27.b3 ltJ d3 28 .�c2 �e l
0- 1 Doroshkievich - Tal, Yerevan 1 97 5 .
However, the move made b y Yuri Dokhoian
(a former long-time second of Garry Kasparov,
and captain of both the men's and women's
Russian national team) has its venom.
a
b
c
d
e
f
g
h
1 25
Chapter 4 - Line Opening
14 . . . b5! 1 5 .ie3 bxc4 1 6.ixc4t @h8 1 7.ixd4
cxd4 1 8 .lt:Jd5 ia6! 1 9 .lt:Jxf6 ixc4 20 .lt:Jh5
ixfl 2 1 .Wg4 Wd7 22.E!xfl d3 23.Wf3 d2
24.g4 E!ac8 25 .Wd3 Wa4 26. lt:J f2 Wd4
Black won in Christiansen - Kasparov,
Moscow 1 982.
8
7
6
5
4
10.fxe5
Interesting complications could have
resulted after 1 0.f5!? gxf5 l l .exf5 d5!? 1 2.cxd5
( 1 2.lt:Jxd5 is possible as well) 12 . . . e4 . I will
leave it to the reader to bring in a verdict.
3
2
1
a
10 ... dxe5 1 1 .0-0
b
c
d
e
f
g
h
14 ... g5!
On the greedy: l l .lt:Jxe5
Fighting for the initiative.
I S.iel
After 1 5 .ig3 lt:J h5+ Black's activity is
annoying, and it is not easy to find a plan for
White.
15 ... ig4 16.ic3 llJ d4 17.ixd4
a
b
c
d
e
f
g
h
Black can (and should) continue l l . . . lt:J g4!
12.lt:Jf3 E!xf3! 1 3 .gxf3 ( 1 3 .Wxf3? lt:J e5)
1 3 ... 'Wh4t 1 4.@d2 lt:J f2 1 5 .'Wel lt:J c6! (the
point) , with a devastating attack.
1 1 . .. liJc6
So the position is almost symmetrical, the
only difference being that the bishop on g7
is potentially slightly more active than its
counterpart on d3. Since it is White's turn to
move, chances should be roughly equal.
12.ig5 h6 13.ih4 VNd6 14.llJd5?!
In my opinion 1 4.ic2 lt:J d4 1 5 .lt:Jxd4 cxd4
16.lt:Jd5 lt:Jxd5 1 7.cxd5=, with a dull position,
was more to the point. But in that case I doubt
this game would have been in this book.
a
b
c
d
e
f
g
h
This bishop has made a long trip just to be
exchanged for the knight on d4 - a definite
achievement for Black.
17... llJxd5! 1 8.exd5
1 8.ixe5 ixe5 1 9.exd5 ixh2t under­
standably did not appeal to my opponent.
1 26
King's Indian Warfare
1 8 ... exd4 19 ..ie4
It seems that White has a solid and perfectly
safe position.
1 9.h3, to check the intentions of the bishop
on g4, was the better move. Now if 1 9 . . .ixh3
( 1 9 . . .ih5 20.ic2 �f4 2 1 .�d3 �e3t 22.�xe3
dxe3 23.g4 if7 24J�ae l �ae8 2 5. �e2 and
White is not worse in this ending; 1 9 . . . id7!?)
20.gxh3 �g3t 2 1 .�h l �xh3t 22.lt:Jh2 ie5
23.�c2 ixh2 24.if5 �xf5 2 5 .�xf5 �xf5
26.�xf5 if4 Black would remain slightly
better, with two pawns for the exchange,
but after, let's say, 27.�e l a draw is the most
probable outcome.
8
7
6
5
4
3
2
1
a
b
d
c
e
f
g
h
2 1 . ...ic8!
8
A very strong and unexpected manoeuvre,
with the idea . . . ia6. Also . . . g5-g4 may be an
important threat in some lines.
7
22JUcl
6
Slightly preferable was 22.�c2 ia6 23 .�fe l ,
but with energetic play - 2 3 . . . c4 24.ih7t
�h8 25 .�e6 �c5 ! 26.id3 (26.�xa6 d3t
27.�f2 �xf2t 28.�xf2 �xh7)
5
4
3
2
1
a
b
c
d
e
f
g
h
19 ... b5!
With the clear goal of opening up the
position; my bishop pair would appreciate that
very much!
20.%Vd3?
This natural move is a mistake, but it was
not easy to foresee Black's reaction.
20.cxb5 c4+ was barely playable, but after the
modest 20.b3 bxc4 2 1 .bxc4 �ab8 Black would
have only a little pressure.
20 ... bxc4 2 1 .%Vxc4
a
b
c
d
e
f
g
h
26 . . . �b5! 27.�xa6 cxd3 28 .�c6 �xb2
29 .�c l �c3 30.�c6 d2! 3 1 .�xc3 dxc l =�t
32.�cxc l d3 33.�ab l g4 - Black obtains a
nearly-decisive advantage in the endgame.
22 ... d3!
Chapter 4 - Line Opening
1 27
27.tDe4?!
8
More stubborn was 27.Wc3t Wf6 28.Wxf6t
El:xf6 29.ib5, but after 29 ... El:b8 30.El:xc5 a6
3 l .El:c7t @f8 32.ifl El:b2 Black is winning
anyway.
7
6
5
27 ...VMe5 28.h4
30.i.xe4 :g£4
4
3
i.xe4
29.VMxe4
VMxe4
The rest is simple and does not require
further comment.
2
1
a
b
c
d
e
f
g
h
I am sorry for using so many diagrams, but
this four-move span (starting from my 1 9th
move) is one of my favourites in my whole
career. It's hard to believe, but this powerful
push basically decides the game in Black's
favour. The dynamics of Black's position are so
strong that the opponent is helpless.
3 I .i.d3 :gd4 32.i.b5 :gcs 33.i.c6 c4 34.hxg5
hxg5 35.®£2 ®f6 36.:gel :gf4t
8
7
6
5
4
23.i.xd3
What else? 23.Wxc5 Wxc5t 24J:hc5 g4
25.ctJd2 id4t and 23 .El:ab l ia6 24.Wxc5
Wxc5t 25 . El:xc5 El:xf3 26.ixf3 id4t and Black
wins in both cases.
3
2
a
23 ...i.xb2
And a simple fork is the result. A decisive
loss of the exchange is unavoidable.
24-!�ab l i.xcl 25-!�xcl .ifS-+ 26.tiJd2 ®g7?!
26 . . . ixd3 27.Wxd3 El:ae8 would win faster.
8
6
5
4
3
2
a
b
c
d
e
f
g
h
c
d
e
f
g
h
37.®g3 c3 38.:ge6t ®f7 39.:ge2 :gc4 40.:gc2
:gbs 4I .i.d7 :gb2 42.:gcl c2 43.i.f5 ®f6
44.i.h7 :gxa2 45.d6 :gb2 46.:gf1 t ®g7
47.d7 :gd4 4S.i.f5 :gbl
What I like most about this game is the
speed with which White's seemingly solid
position fell apart after a few forceful strokes.
0-1
7
b
1 28
King's I ndian Warfare
Rafael Leitao - Ilya Smirin
New York 2000
This game was played in the penultimate round
of the New York Open. Just as it sounds, this
was a traditional big open tournament held
every year in New York. I managed to take
clear first place by beating Becerra, Leitao and
the young rising star 1 6-year-old Alexander
Grischuk in the last three rounds. Alas, up to
now this was the last New York Open held
in the centre of Manhattan - after 2000 it
ceased to exist. That's a pity, because I liked
the tournament's atmosphere, which made the
event very special among those I played in the
USA. As a small consolation, I remain the last
winner of this famous event.
l .d4 llJ£6 2.c4 g6 3.llJc3 .ig7 4.e4 d6 5 ..ie2
0-0 6 ..ig5 llJ a6
The main idea of this move is that after
6 . . . ttJ bd7 7.�d2 e5 S.d5 Black cannot play
S . . . c6? due to 9.dxc6 bxc6 1 0.�xd6.
1 1 . . . dxe4 1 2.tLlxe4 ig4 1 3 .�ad l �e7 1 4.tLlf2
ixf3 1 5 .ixf3 f5 1 6.d5 cxd5 1 7.cxd5 �d6
1 S .g3 tLlb5 1 9.ig2 �acS 20.�h 1 �c4 2 1 .tLld3
�fcS 22.�c l h5 23 .�e2 �c7 24.�xc4 �xc4
25 .�c l �xc l t 26.tLlxc 1 �xc l t 27.ifl ttJd6
b
a
c
d
e
f
g
h
2S.b3 �b l 29.�d3 �b2 30 .ig2 �xa2 3 1 .h4
a5 32 .�e3 b5 33 .�b6 ifS 34.if3 �a3
3 5 .�g2 b4 36.id 1 �a2t 37.�fl tLl e4 3S .�e l
ic5 39 .�bSt �h7 40.�eS �g7 4 1 .�e5t
�gs 42.�bst ifs 43.if3 ttJd6 44 .�b6
�a3 45 .id 1 tLle4 46.if3 �a1 t 47.�e2 tLld6
4S.�d3 �cl 49 .�xa5 �c3t 50.�e2 �xb3
5 1 .�a4 �b2t 52.�d3 b3 53.id 1 tLl e4
0- 1 Yakovich - Smirin, Munich 1 993.
7 ...e5 8.d5 c6
a
b
c
d
e
f
g
h
7.%Yd2
The more straightforward and aggressive 7.f4
was played in an old game by Yuri Yakovich.
7 . . . c6 S . tLl f3 tLlc7 9 .�d2 d5
9 . . . b5!?
1 0 .ixf6 exf6 1 1 .0-0
Better is 1 1 .exd5 cxd5 1 2.c5.
a
b
c
d
e
f
g
h
Now this advance is possible. Black starts
some activity on the queenside. There is always
1 29
Chapter 4 - Line Opening
a question in posltlons of this type in the
KID : which side will benefit more from that?
Quite often queenside play may boomerang
for Black, and White turns out to be superior
on this part of the board after all. Usually it
depends on nuances and the placement of
pieces. Here, for instance, in Black's favour is
the fact that after a likely pawn exchange on
d5 (White will recapture cxd5) the knight on
gl will not easily be able to reach its desired
destination on c4 - the f3-square, needed for
the manoeuvre lt:J f3-d2-c4, will be occupied
on the very next move.
knight on g1 to f2 via h3. After this is done,
White may even think about aggression with
h2-h4-h5 (the immediate h4 with the knight
still on g 1 may face the counterpunch
. . . h7 -h5 ) . 1 o . . h5!? This move disturbs the
opponent's plan, but Black's king may feel a
little uneasy. 1 1 .h3 cxd5 1 2.lt:Jxd5 ( 1 2.cxd5
id7 would lead to a complicated middlegame)
1 2 . . . �xd2t 1 3 .�xd2 lt:Jxd5 1 4.cxd5 f6 1 5 .ie3
f5 1 6.g5 id7 1 7.id3
.
9.f3
Of course 9. lt:Jf3 lt:J c5 would be great for
Black.
9.. VMa5
.
I have played this move a few times.
A more popular line here is:
9 . . . cxd5 1 0.cxd5 id7
Defi n ing the pawn structure in the centre.
For instance:
1 1 .ib5!?
A logical idea - to swap the light-squared
bishops and vacate the e2-square for the
knight on g1 - but it costs White some
tempos.
ll . .. ixb5 1 2.lt:Jxb5 �b6 1 3.lt:Jc3 lt:J c5 1 4.�bl
lt:Jh5 1 5 .ie3 lt:Jf4 1 6.g3 lt:J fd3t 1 7.�fl
With complicated play in Yusupov Gelfand, Dortmund 1 997.
c
d
e
f
g
h
1 7 . . . h4! ? 1 8.�cl lt:J c5 1 9.ixc5 dxc5 20.�xc5
�fc8 2 1 .�xc8t �xc8 22.exf5 gxf5 23.lt:Je2 if8,
and Black has sufficient (at least) compensation
for the pawn in the endgame; Yusupov Smirin, Yerevan 1 996.
IO ...ttk5 I I ..ic2
8
7
6
IO ..id3!?
5
White prepares the natural lt:J ge2 and short
castling. If Black does not create counterplay
quickly, he will j ust stand worse, with a lack of
space, as is often the case in the KID.
4
A couple of times I have faced 1 O.g4 here this move aims to grab some space on the
kingside and to prepare the development of the
b
a
3
2
1
a
l l . .. b5!?
b
c
d
e
f
g
h
1 30
King's Indian Warfare
An interesting over-the-board improvisation.
I wanted to use White's temporary lack of
development and to undermine the central
pawn chain. Strictly speaking, I did not have
much choice - after 1 1 . . . cxd5 1 2 .cxd5 i.d7
1 3.a3 all the black pieces would have to retreat
very soon, and I certainly wanted to avoid that.
1 3.exd5?!
The wrong capture. Rafael had to choose
another way, occupying the central square d5
with the knight rather than the pawn:
1 3 .i.xf6! i.xf6 1 4.ct:Jxd5 Wxd2t 1 5 .c;!?xd2
ig5 t 1 6.c;!?d 1
1 2.cxb5
1 2 .dxc6 bxc4 1 3 .�xd6? does not work:
8
7
6
5
4
3
b
a
2
b
a
c
d
e
f
g
h
1 3 . . . :gd8 1 4.�e7 i.f8 1 5 .�xf6 :gd6 and the
queen is trapped on account of 1 6.�xe5 ctJ d3t
1 7.i.xd3 �xe5.
c
d
e
f
g
h
And after 1 6 . . . i.b7 1 7.ctJe2 ( 1 7.ctJc3!? :gfd8
1 8 . ctJ h3 i.e3 1 9.c;!?e2 i.d4 20.ctJ f2 does not
look not bad either) 17 . . .i.xd5 1 8 .exd5 :gfc8
1 9.ctJc3 ie3 , transferring the bishop to d4,
Black would have compensation for the pawn,
but White would probably keep some edge the extra pawn is an extra pawn after all.
13....ib7 14J�dl a6!
8
With the intention of opening some files
on the queenside in the spirit of the Benko
Gambit.
7
6
5
15.a3 t!Jcd7 16.bxa6 .ixa6
4
3
8
2
7
1
6
a
b
c
d
e
f
g
h
1 2...cxd5
This pawn sacrifice was the idea behind my
1 1 th move.
After the naive 1 2 . . . cxb5 ? 1 3 .a3 Black would
remain much worse.
5
4
3
2
1
a
b
c
d
e
f
g
h
131
Chapter 4 - Line Opening
The c4-square is an excellent destination for
this knight.
He should has preferred the passive 22.@h 1
El:ab8 23.ic l , with a position that is still
defensible. 22.Ei:fc l would lead to very sad
consequences: 22 . . . E!:ab8 23.El:c2 lD xa3! 24.El:a1
(24.bxa3? El:xb 1 t 2 5 . tt:J xb 1 �e 1 #) 24 ... �c5t
25 .ie3 �xe3t 26.�xe3 tt:Jxc2 27.�c l tt:Jxa 1 +
and the queen is clearly no match for the black
rooks.
18 ..id3 .ixd3 19.V:Vxd3 �Uc8 20.0-0?!
22 ... V:Vb6t 23.®hl
The Brazilian Grandmaster is in a hurry
to castle, but it was better to offer a queen
exchange before it's too late - 20 .�b5 ! with
roughly equal chances. If 20 . . . �xb5 2 1 .tt:J xb5
tt:J fxd5 22.b3!? El:c2 23.@f2 White's pieces are
active enough.
The last mistake. The only move was 23.El:f2,
even though Black is clearly dominant after,
let's say, 23 . . . f5 24.tt:Jac3 e4 2 5 .fxe4 tt:J de5
26.�g3 tt:J g4.
Now Black will combine pressure against the
pawn on d5 with play on the queenside. A very
important factor is that White's king is still in
the centre.
17.llJge2 llJ b6
23 ... llJxb2 24.V:Ve4
24.�d2 tt:J c4
20 ... llJc4 2 1 .gbl
Bad would be 2 1 .ic 1? tt:Jxb2 22.ixb2 �b6t
23.@h 1 �xb2, with a clear edge.
24... gxa3
8
2 I . .. llJ d7
7
8
6
7
5
6
4
5
3
4
2
3
1
a
2
1
a
b
c
d
e
f
g
h
The knight moves closer to the battlefield
and now Black is ready for . . . f7 -f5, which will
considerably strengthen his position.
22.llJa2?
An oversight. Leitao wanted to transfer this
knight to c6 via b4, but he j ust has no time to
do it.
b
c
d
e
f
g
h
White's pawns are falling one by one. The
end is near.
25.llJac3
Or 2 5 .tt:Jb4 El:c4-+ .
25 ... V:Va6!
25 . . . El:axc3 26.tt:Jxc3 El:xc3 would win even
faster, but I did not want my knight on b2
to be pinned, and chose a more "human"
continuation.
1 32
King's Indian Warfare
Viktor Gavrikov- Ilya Smirin
26J�xb2 gaxc3-+
Now the white knight is the one that is
pinned.
This game remains very special in my
memory. I was playing against the well-known
theoretician Grandmaster Viktor Gavrikov
and he was clearly better prepared. Still, I was
not too impressed, managing to create huge
complications and get a winning position.
The way the game ended is unique for me as
well - more about this at the end. To this day
I consider this game (as well as Radjabov Smirin from this book) as one of my best ever.
8
7
6
5
4
3
I.ltJO ltJf6 2.c4 g6 3.ltJc3 .ig7 4.e4 d6 5.d4
0-0 6..ie2 e5 7.0-0 ltJc6 8.d5 ltJe7 9.ltJd2
2
1
a
b
c
d
e
f
g
h
The triumph of Black's strategy. It was very
pleasant to have such an ideal position.
30 ..ih4 e4 3 1 .fxe4 gal
White resigned - heavy material losses are
unavoidable.
0-1
Klaipeda 1988
This continuation was the most popular
from the 1 980s until the mid- 1 990s, when
Vladimir Kramnik started to play 9.b4 lLlh5
1 0.El:el with huge success. 9.lt:Jel is also played
often nowadays. But I would not be surprised if
the good old 9.lt:Jd2 were to regain popularity
in the near (or more distant) future.
9 ... a5 10.a3 ltJd7 l l .gbl f5 12.b4 �h8
13.V;Yc2
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
13 ... b6
Following in the footsteps of Carry
Kasparov - he played this move against the
1 33
Chapter 4 - Line Opening
same opponent j ust a few months before this
encounter.
1 3 . . . ctJg8!? became a popular continuation
after Kasparov played it several times, albeit
after 1 4.f3. In this case White may continue
1 4. f3 tt:J gf6. For example, it was played in the
following well-known game at the highest
level: 1 5 ..id3!?
active enough. This option seems to me to
be more promising for White, for instance:
1 5 . . . axb4 (or 1 5 . . . ctJ e7 1 6. ctJ f3 e4 1 7.ctJg5 tt:J f6
1 8 ..ie3 h6 1 9.ctJe6 .ixe6 20.dxe6) 1 6.axb4
exf4 (perhaps 1 6 . . . ctJ e7 is slightly more flexible
here) 1 7. ctJ f3 ctJ df6
8
7
6
8
5
7
4
6
3
5
2
4
3
a
2
a
b
c
d
e
f
g
h
This very interesting idea is to force . . . f5-f4
and release the pawn tension in the centre.
It became very popular. 1 5 .. .f4 1 6.ctJb5 b6
1 7.c5!? dxc5 1 8 .bxa5 El:xa5 1 9.ctJc4 El:a8 20.a4
ctJe8 2 1 .a5 .ia6 22 . .id2 .ixb5 23 .El:xb5 tt:J d6
24.El:b2 Wfe7 25.ctJxd6 Wfxd6; (Y2-Y2, 4 1 )
lvanchuk - Kasparov, Linares 1 992.
A different plan is 13 . . . ctJ g8 1 4.exf5 gxf5 1 5 . f4
b
c
d
e
f
g
h
1 8 . .id3 and, as practice shows, White chances
are preferable.
14.f3!?
Gavrikov improves his play compared to the
game with the World Champion, in which
he proceeded: 1 4. ctJ b3 axb4 1 5 .axb4 fxe4
1 6.tt:Jxe4 tt:J f6
8
7
6
5
4
3
2
a
a
b
c
d
e
f
g
h
Trying to apply pressure to Black's kingside
at a moment when Black's pieces are not yet
b
c
d
e
f
g
h
Gavrikov - Kasparov, Moscow 1 988, with
a draw after 40 moves. Now if the pawns are
exchanged on e4 White will take with the
f3-pawn, eliminating the opponent's idea of
activating the knight on e7 via f5 .
1 34
King's I ndian Warfare
I like the way White handled the opening in
the following game: 1 4.<tt> h l ! ? A useful waiting
move; White allows the opponent to show
his hand. 1 4 . . . ltJ g8 1 5 .exf5 gxf5 1 6.f4! ltJ df6
( 1 6 . . . ltJ e7 looks a more logical move, but I
think White is better in any case) 1 7. fxe5 dxe5
1 8 . ltJ f3 axb4 1 9.axb4 ltJ g4?!
8
with
approximate
equality)
2 1 . . . ltJdf6
22.'1Wd2 '.Wh4 23 .'1We3 White's chances are
to be preferred: the pawn on c7 is very weak
and some activity on the queenside does not
compensate for that.
15.ltJb3 axb4 16.axb4 g5 17.c5 t!Jf6 18.t!Jb5
8
7
6
7
5
6
4
5
3
4
2
3
a
b
c
d
e
f
g
h
20.ltJg5! h6 2 l ..ixg4 hxg5 22 ..ixf5 e4 23.g4!
with a decisive advantage; ( 1 -0, 32) Lputian ­
Frolov, Helsinki 1 992.
14... f4
A typical move, preparing the kingside
attack. The alternative was 1 4 . . . .ih6?!,
activating and exchanging the King's Indian
bishop. But it takes a lot of tempos. Then
1 5 . ltJb3 .ixc l 1 6.!!fxc l fxe4 1 7. fxe4 axb4
1 8 . axb4 ltJ g8 1 9 .c5 bxc5 20.bxc5 '.Wg5 looks
logical for both sides.
2
a
b
d
c
e
f
g
h
Gavrikov had spent barely a few minutes
until this moment, demonstrating his home
preparation, whereas for me all this was new.
But, being young, I was very optimistic!
18 ... g4
I decided to keep the b-pawns, even though
the pawn on b6 may potentially become a
weakness.
But now I think that probably it would not be
a bad decision to remove them - after:
1 8 . . . bxc5 1 9.bxc5 g4
8
7
6
5
8
4
7
3
6
2
5
4
a
b
c
d
e
f
g
h
2 l .c6! (2 1 . ltJb 5 ? '.We3t 22.<tt> h l !!f2 23.!!e l
!!b8!; 2 l .cxd6 cxd6 22.'1Wd2 '.Wd8 23 .'1We3 .ia6
3
2
a
b
c
d
e
f
g
h
13S
Chapter 4 - Line Opening
20.cxd6
20 . .id2 with the idea of putting a bishop
on aS does not really work due to: 20 . . . gxf3
2 l .gxf3 .ih3 22. 8:fd 1 tt:J xe4! 23.'1Wxe4 (or
23.fxe4 tt:JxdS! with a very strong attack)
forces. Bur in this case we would not have seen
the most fascinating part of the game - so I am
glad that I chose the more strategically double­
edged continuation.
19.cxd6 cxd6 20.�c7
8
7
8
6
7
5
4
6
3
5
2
4
3
a
b
c
d
e
f
g
h
23 . . . �f6! 24.�h 1 ifS 2S.\Wc4 .ixb 1
26.�xb 1 �h6 27 . .id3 '1Wd7 28 .\We4 �f8!
29.ltJc3 ltJ fS and Black's attack outweighs
the minimal material deficit.
20 . . . cxd6
The absence of b-pawns would be slightly
in my favour - after all, as I said, the weak
pawn on b6 no longer exists.
Mter a possible:
2 1 .'1Wc7 gxf3 22.'1Wxd8 f2t!?
8
2
1
a
b
d
c
e
f
h
g
realized that strategically my posmon
was very dangerous, especially in a possible
endgame.
20 ... gxf3 2 1 .gxf3
A variation in which keeping the b-pawns
on the board turns out to be in my favour
could have happened if: 2 1 .'1Wxd8
7
8
6
7
5
6
4
5
3
4
2
3
2
a
b
c
d
e
f
g
h
23.�xf2 �xd8 24.tt:Jd2 tt:J g6 2S . .ia3 .if8
Black is OK in this queenless middlegame
- the vulnerability of the pawn on e4
compensates for the weak d6-pawn, and
Black's pieces are no less active than White's
a
b
c
d
e
f
g
h
2 l . . . fxe2!
I do not like the consequences of 2 l . . . �xd8
22.gxf3 ih3 23.�d 1 �g8 24 . .ifl .if8t
2S.�h 1 .ixfl 26.�xfl .
1 36
King's I ndian Warfare
Back in 1 98 8 I was going to answer
2 1 . .. f2t?! 22.!hf2 �xd8 and thought that I
would have real counterplay against the pawn
on e4. But the artificial intellect insists that
after 23.tLld2 tLl g4 24.ixg4 ixg4 2 5 . tLl c4
tLlc8 26.g3! White has a tangible advantage
in the endgame. It is sometimes difficult to
argue with such a competent opinion.
22.Wixe7
Or 22.Wixb6 exfl =Wit 23 .�xfl lLlxe4
24.tLlxd6 �a6 25.tLlxc8 �xb6 26.tLlxb6 tLl c3
27.d6 tLl f5 28.�al lLl xd6, with a complex
but approximately equal endgame.
22 . . . exfl =Wit 23 .�xfl ia6!
28 .ib2 tLld2 29.�d l tLlc4 30.Wlc7 �f7 3 1 .tLld7
tLle3
Black is at least not worse. Still, I think that
considering the problems that White faced
in the middlegame, it was probably the right
decision to exchange the queens immediately.
8
7
6
5
4
8
3
7
2
6
1
5
a
4
b
c
d
e
f
g
h
2 1 . ..�e8!
3
2
a
b
c
d
e
f
g
h
Here it is! Because of the b-pawns, the rook
on b 1 is unable to protect the pinned knight
on b 5 .
24.Wixd6 ixb5 t 2 5 .�gl lLl xe4
The position it really messy, but it seems that
after the probable continuation:
26.Wlxb6 ic4 27.tLl c5 ixd5
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
As I mentioned before, I was not eager to
exchange queens, even though the computer
does not think that Black has real problems
after:
2 l . . .Wixc7 22.tLlxc7 �a2 23.tLld2
If23.�b2 �a4 24.id2 Black has: 24 . . . tLl fxd5!
An excellent sacrifice to activate the King's
Indian bishop. 25.tLlxd5 lLlxd5 26.ib5
(bad is 26.exd5 e4 27.�bb l e3 28.ic l ifS
29.�b2 �g8 30.�h l �c8 3 1 .id l �a7, and
despite having an extra piece, White is on
the ropes) 26 . . . �a3 27.exd5 e4 28.ic l e3,
with great compensation.
23 . . .ih3 24.�d l tLle8 25.tLlb5 �g8 26.�h l
if6 27.�gl �xgl t 28.�xg l
I still suspect that Black's position is not so
rosy, and that the weaknesses on b6 and d6,
together with a lack of space on the queenside,
might soon tell. Fortunately, I had a strong
alternative at my disposal, and I went for it.
Objectively speaking this is the strongest move
in the position - I failed to find even equality
for White after that.
1 37
Chapter 4 - Line Opening
22.ci>hl
Suddenly my opponent needed to solve a
concrete problem connected with his king
facing some serious threats. It's not easy, from
a psychological point of view, to switch from
dreams of a pleasant endgame to the necessity
of calculating numerous quite complex
variations, or to call on one's intuition for
help. It is a really hard task, and Gavrikov did
not manage to cope with it during the game.
But even now, not being limited by time and
having a very powerful assistant (you know
what I mean) , I find it hard to show the right
path for White. For example, both the possible
captures of the "doomed" pawn on d6 are
inadvisable for White:
22.'Mfxd6? .ih3 23.�f2 Wg6t 24.<tt> h l tLlxe4!
and White may as well resign.
22.tLlxd6 'Mfg6t 23. <tt> h l tLlexd5! 24.exd5
A nice variation may arise after 24.'Mfc4
tLl h5! 25.tLlxc8 (25 .'Mfxd5 tLl g3t 26.hxg3
'Mfh5t 27.<tt> g l fxg3 28.�f2 gxf2t 29.<tt> xf2
.ig4-+) 25 . . . �axc8! 26.'Mfxd5 tLl g3t 27.hxg3
'Mfxg3 and you may check for yourself that
the attack decides the outcome.
24 . . . 'Mfxb l 25 .�gl id7 26.tLld2 'Mfa2
Black has an extra exchange without allowing
any compensation.
a
b
d
c
e
f
g
h
23J�b2?
Analysing the game without a computer
back in 1 98 8 , I thought this move was mistake
and that White should have played:
23.�f2
But now the silicon creature considers both
moves to be equal to each other. Moreover,
to my sincere amazement, it cannot find
a completely satisfactory defence for my
opponent in the complications starting after
2 l . .. 'Mfe8! . If that's true, we may consider
White's 2 1 st move to be a serious and almost
decisive mistake: Gavrikov should have
exchanged queens instead.
After the move in the game it seems for a
moment that the white king has found safety
in the corner, but it's j ust an illusion.
22 ... llJh5!
Black's pieces start to get closer to the white
monarch. This knight is clearly in sacrificial
mode now - the j ump to g3 is constantly
hanging in the air (and indeed will happen
soon) .
a
b
c
d
e
f
g
h
OK, if White were to proceed on move 23
with the other rook, the knight sacrifice
seems to lead to a decisive advantage for
Black:
23 . . . tLl g3t!
1 38
King's I ndian Warfare
But not 23 . . . if6 ? 24.lt:Jxd6 '!9g6 25 .El:g2
'!9h6 26.'!9d8! ie6 27.'!9xa8 El:xa8 28.dxe6+­
(a variation from 1 988) .
24.hxg3 fxg3 25 .El:g2 '!9h5t 26.�gl '!9h4!
And the defence is broken. This beautiful
variation shows that Black's pieces (the rook
in this case) may also penetrate from the
queenside.
23 ....th3
8
7
6
8
5
7
4
6
3
5
2
4
a
b
c
d
e
f
g
h
The key move, protecting both the pawn
on g3 and the knight on e7. White's main
problem, in this and similar positions, is
that he is unable to regroup his pieces fast
enough to defend his king.
27.lt:Jd2
Or 27.id 1 ih3 28 .El:bb2 El:ac8 29 .'!9xd6
lt:Jg6! 30.lt:Jc7 El:f6 3 I .lt:Je6 El:xe6 32.dxe6
ixg2 33.El:xg2 El:d8 34.e7 '!9xe7! 3 5 .'!9xe7
El:xd 1 # nice!
27 . . . ih3 28.lt:Jfl ixg2
The simplest.
29 .�xg2 El:a2! 30.El:b2 El:xb2 3 I .ixb2 lt:J g6
32.ic l
3
2
1
a
b
c
d
e
f
g
h
24J�el
Now I would like to present some lengthy
computer analysis. Sorry about that, but I
hope you will enjoy the variations. White had
two other major options at his disposal:
-
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
32 . . . ih6! 33.ixh6 '!9xh6 34.lt:Jxg3 lt:J f4t
3 5 .�f2 '!9h2t 36.�e3 '!9xg3
a) 24.El:g l lt:Jg3t 25.hxg3 '!9h5 26.El:g2
Or 26.'!9xe7 if6-+ .
26 . . . ixg2t 27.�xg2 fxg3 28.�xg3 if6
29 .�f2 El:ac8
The strongest.
I was very happy when I discovered
a checkmate of rare beauty in 1 98 8 :
29 . . . '!9h 1 30.lt:Jxd6 El:ac8 3 1 . lLl xc8 El:xc8
(objectively stronger is 3 1 . .. '!9h4t 32 .�fl
El:g8 33 .id l '!9h l t 34 .�e2 El:g2t 3 5 .�d3
'!9xd 1 t 36.lt:J d2 El:f2 37.'!9d8t lLl g8
38 .'!9xb6 El:xf3 t 39 .�c4 '!9e2t 40.�c5
El:c3t 4 1 .�d6 El:xc8 and White is helpless)
32 .'!9xb6 (alas, this line is not forced and
after 3 2.'!9xc8t lt:J xc8 33 .El:c2 lt:J d6 34.El:c6
White is still in the game) 32 . . .ih4t
33.�e3
1 39
Chapter 4 - Line Opening
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
33 . . . Ek3t 34.id3 '1We l t 3 5 . �e2 if2#
8
7
6
5
4
3
2
28.f5
If 28.if3 exf4 29.�d2 ixfl 30.@xfl
1Wh3t 3 l .�g2 (3 I .ig2 '1Wg4 32.'1Wxe7 f3-+)
3 1 . .. lt:J g6 White has no defence. After the
most stubborn 32.lt:Jf5 �xf5 33.exf5 �e8!
34.lt:Jd2 1Wxf5 3 5 .ie4 �xe4 36.lt:Jxe4 '1Wxe4
37.'1Wc8t lt:J f8 38 .'1Wc2 1Wxd5 39.ib2 @g8
40.ixg7 <±>xg7 he is able to escape the
mating attack, but the resulting position
would be easily winning by technical means
for Black.
28 . . . ixf5 29.ib5
Bad is 29.if3 ih3 30.ig2 ixg2 3 l .�xf8t
�xf8 32.�xg2 '1Wg4 33 .1Wc2 1Wf3 34.1We2
'1Wxb3.
And 29.id3 ih3 30.lt:Jf7t @g8 3 l .d6 ixfl
would lead to the same position.
29 ...ih3 30.lt:Jf7t @g8 3 I .d6 ixfl 32.ixfl
lt:Jd5! 33.1Wc4 �xf7 34.1Wxd5 �af8 3 5 .ic4
@h8 36.1Wxf7 �xf7 37.ixf7 '1Wxe4 38.d7 if6
39.�d2 @g? 40.id5 1We l t 4 I .@g2 h5
8
a
b
c
d
e
f
g
h
30.'1Wxd6 ih4t 3 I .@gl �xc l t 32.lt:Jxc 1 '1Wg5 t
With the total collapse of White's position.
b) 24.lt:Jxd6 lt:Jg3t
Again this knight sacrifices itself in kamikaze
fashion.
25.hxg3 1Wh5 26.@gl fxg3 27.f4
The only move.
27 . . . 1Wh4
8
7
6
5
4
3
7
6
5
4
3
2
a
b
c
d
e
f
g
h
After a huge mess, an unbalanced position
has arisen. Black is still on top, but, thanks
to the pawn on d7, White is certainly not
doomed, and may count on a positive
outcome. It might well be that this position is
the best White may hope for as a result of the
complications.
24...%Yg6 25.�fl
2
a
b
c
d
e
f
After 25 .id l 1Wf6! White has no defence:
26.lt:Jxd6 1Wh4 27.�gl lt:Jg3 t 28.hxg3 fxg3
g
h
140
King's I ndian Warfare
29.!hg3 l.Wxg3 30 .l.Wxe7 l.We l t 3 1 .�h2 l.Wxd l
32. 4J f7t �xf7 33.l.Wxf7 �f8 34.l.We7 �xf3 and
his position j ust falls apart.
25 ... �g3t! 26.hxg3 Wxg3
8
3 t .Wg2
3 1 .�f2 does not help due to 3 1 . . . �g8t
32.�g2 �xg2t 33.l.Wxg2 �g8-+ .
If 3 1 .l.Wd3 �g8t 32.�f2 4J xd5! 33.exd5 e4
34.l.Wxe4 �ce8 White should resign despite his
two extra pieces.
3 1 ..J�g8 32.�xd6 gcf8 33.ge2 Wd7 34.�c4
gxg2t 35.gxg2 �g6
7
6
5
8
4
7
3
6
2
5
4
a
b
c
d
e
f
g
h
As we have already seen in a few variations,
the knight on h5 has completed his mission
on g3 and the queen and some other black
pieces have made a rendezvous with the white
king. The few next moves are basically forced
for White.
27.Wc3 gac8 28.Wd2 .ixfl 29J�xfl Wh3t
30.i>gl .1£6
8
7
3
2
a
b
c
d
e
f
g
h
The rest, as they say, is a matter of technique:
Black has a material advantage and the white
king is still unsafe.
36 ..1b2 �h4 37.gh2 Wc7 38.�bd2 b5
39.�a5 Wc2 40 ..ial ggst 4I.i>hl gg2!
42.gxg2 �xg2 43.i>xg2 Wxd2t 44.gf2 Wel
45.gfl Wg3t 46.<it>ht h5
6
8
5
7
4
6
3
5
2
4
3
a
b
c
d
e
f
g
h
White must give up his queen to avoid
immediate disaster.
2
1
a
b
c
d
e
f
g
h
141
Chapter 4 - Line Opening
Here, as was the custom i n the good old days,
the game was adjourned. The position is easily
winning for Black - the forcing line 47.ttJc6
mh7 48 .ic3 ih4 49 .ttJxe5 if6 50.d6 ixe5
5 I .ixe5 '1Wh4t 52.mg2 '1Wg5t 53.mf2 '1Wxe5
54.E!.d l '1Wc3 demonstrates that convincingly.
But the game was destined not to be resumed
- for some reason Gavrikov withdrew from the
tournament a few days later. He did not resign,
so the position was evaluated by tournament
officials and declared a win for me. However,
I did not receive a point in the tournament's
crosstable - all Gavrikov's results were
cancelled. But despite that, the game brought
me enormous creative pleasure.
Igor Novikov- Ilya Smirin
Las Vegas 1999
Las Vegas . . . one of the most surreal cities in the
world. For me, playing chess in this gambling
empire always added some more adrenaline
than usual. The following is my favourite game
from all the five tournaments (including the
1 999 World Cup) that I played in Vegas.
l .d4 ttJf6 2.c4 g6 3.tiJc3 .ig7 4.e4 d6 5.8
0-0
0-1
a
b
c
d
e
f
g
h
6 ..ig5
This is a more ambitious move compared to
the most popular 6.ie3. In my opinion it has
one drawback - the d4-pawn (or square) may
become more vulnerable.
6 ... a6 7.Y!Yd2 tlJc6 s.tlJge2 gbs
Preparing . . . b7-b 5 , j ust as in the line where
the bishop is developed on e3 instead of g5 .
9.d5
Sometimes White chooses the prophylactic:
9 .E!.c l
Here I have played:
9 . . . h5!?
This may look a bit strange, but in fact it's
quite a logical move: Black grabs space on
1 42
King's Indian Warfare
the kingside and prepares to play . . . ltJ h7 in
some cases.
9 . . . b 5 ? does not work due to the rook on c l :
1 0.cxb5 axb5 l l .ltJxb5
l O. ltJ d l
A fashionable move i n similar positions
nowadays, but it does not impress me much.
In a few games White has chosen 1 0.b3 ! ?
which i s a kind o f waiting move.
1 0 . . . e5 l l .d5 tLl e7 1 2.c5
Black has better development and can
easily meet the opponent's activity on the
queenside.
1 2 . . . dxc5 1 3 .!hc5 �d6 1 4.Ek2 c6 1 5 .dxc6
�xd2 t 1 6J�xd2 ttJ xc6 1 7. ttJ ec3 ie6
Yz-Yz Kaidanov - Smirin, Mallorca 2004.
Actually I could continue playing in the final
position.
- and he will consolidate his edge. So I have to
hurry to create counterplay. The only way to
do that is to try to undermine White's central
pawns on c4 and d5.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
l l . .. b5 1 2.cxb5 cxd5!
9 ... ttle5 10.tilg3
Mter 1 0 .ltJd4 c5 1 1 .tLl c2 b5?! ( l l . . . ltJh5!?
with the idea . . . f7 -f5 is an interesting
alternative) 1 2.cxb5 axb5 1 3 .ixb5
A positional pawn sacrifice; a similar motif
may be seen in the game Leitao - Smirin above.
In the case of the modest 1 2 . . . cxb5 1 3 .0-0 it
would be hard for Black to find a reasonable
plan.
1 3.exd5
Mter 1 3.ixf6 ixf6 1 4.ltJxd5 ig7 the two
strong bishops and advantage in development
provide Black with clear compensation for the
missing pawn.
13 ... axb5 14.llJxb5 .ih7
a
b
c
d
e
f
g
h
I was convincingly beaten by Dreev after:
1 3 .. Jhb5?! 1 4.ltJxb5 ltJ c4 1 5 .�e2 �a5t
1 6.@f2 ltJxb2 ( 1 6 . . .�xb5 1 7. b3± - I did not
notice that in time) 1 7.a4; ( 1 -0, 39) Dreev ­
Smirin, Biel 1 993.
1 o.. c6 l l ..ie2
.
White has a clear space advantage and no
visible weaknesses. One more move - castling
The aim - the pawn on d5.
15.Lf6
Mter the knight retreat 1 5 .tLlc3 I was going
to continue 1 5 . . . ia8! ? with the idea 1 6.0-0 ?
�b6t.
143
Chapter 4 - Line Opening
17 ...YMb6
8
Keeping the enemy's king in the centre.
7
1 8.tlJe4
6
It is very important that long castling is j ust
losing: 1 8 .0-0-0? .ia6 1 9.ltJc3 .ih6! - the
"passive" bishop delivers a lethal blow.
5
4
3
8
2
7
a
b
c
d
e
f
g
h
1 5 ... exf6!
Perhaps
Novikov expected
1 5 . . . .ixf6
1 6.ttJe4 .ig7 1 7.0-0 with an extra pawn and
a stable position - White's knights coordinate
efficiently.
16.£4
1 6.0-0 f5 would lead to a kind of dream
position for Black - the bishop on g7 is terrific
and the b- and e-files are open for the black
rooks, and most importantly the knight on
g3 is very passive since the e4-square is under
Black's control.
The greedy 1 6.ttJe4!? was j ust about possible.
For example: 1 6 .. . f5 1 7.ttJexd6 .ia6 1 8 .a4
'tl9h4t 1 9 .g3 'tl9e7 20.0-0 �fd8 2 1 .f4 �xd6
22.fXe5 'tl9xe5 Black has fine compensation for
the missing pawn. And of course 23 .ttJxd6??
would lose t o 23 . . . �xb2.
16 ... tlJd7 17.£5!?
Very logical. By blocking the pawn on f6,
White turns the g7 -bishop into a passive piece,
while the knight on g3 is looking hopefully at
e4. Again, if White can castle he will be more
than OK.
IfWhite were to capture another pawn instead
by 1 7.ttJxd6?! f5 1 8 .ttJxb7 �xb7 1 9.�b 1 'tl9b6
he would be in serious trouble.
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 8 ....ia6! 19. ttJ bxd6?
In such a sharp position this natural move
proves to be a decisive mistake, which I failed
to exploit.
If 1 9 .ttJexd6 .ixb5 20.ttJxb5 �fe8 2 1 .tDc3 .if8!
only a computer would have chances to defend
White's position.
1 9.a4!, protecting the knight, was the only
correct move. Mter 19 . . ..ixb5 20.axb5 �a8
2 1 .�xa8 �xa8 22.ttJf2 �a l t 23.ltJ d 1 seemingly
White can hold, even though his position does
not look aesthetically pleasing.
19 ... tlJe5?
For many years I thought this move deserved
a "!", but now I have discovered that 1 9 . . . gxf5!
20.ttJxf5 �fe8 2 1 .ttJ eg3 ttJc5 (with the threat
. . . ttJ d3t) was just a killer. All Black's pieces are
attacking the "centralized" king and the pretty
knight on f5 and two extra pawns do not help.
The following nice variation: 22.'!9d4 'tl9a5t
23 .'!9c3 ltJd3t 24.cj;lfl
144
King's Indian Warfare
2 1 . .. �xc4 22.bxc4 ge8 23.c5!?
The best option - Novikov gives up the piece
but keeps two central passed pawns and finally
castles. Otherwise he would be doomed to a
silent death, for example:
23 .�d3 gxf5
b
a
c
d
e
f
g
h
24 . . J�b4! 25 .ixd3 (2 5 . ttJ h 5 ttJ f4-+) 25 . . .!�f4t
26.cj:;lgl �b6t 27. ttJd4 f5 ! demonstrates this.
20.b3
Now this pawn is protected and it in turn
will protect the knight if it goes to c4.
20.a4 ixe2 2 l .�xe2 E!:fdS 22.ttJb5 El:xd5
23 .El:d l E!:bdS did not look promising for my
opponent.
Or: 23.ttJc3 ifS! 24 .cj:;>dl ih6!! (a really
beautiful move; the prosaic 24 . . . ib4 is not as
powerful, but should also be enough for a win)
25 .�xh6 g5! This is the idea - the white queen
is cut off and the king is absolutely helpless.
26.E!:b l (or 26.�h3 El:e3) 26 . . . �d4t 27.cj:;>c2
El:xe2t 2S.ttJxe2 �xc4t 29.ttJc3 �d3t
23 .. J�Ma7 24.0-0 gxe4 25 ..ixa6 YMxa6
26.gfel
If 26.d6 El:e2 27.�d5 E!:bb2-+ .
8
7
20 .. JUd8
6
5
8
4
7
3
6
2
5
4
a
3
2
a
b
c
d
e
f
g
h
2 1 .�c4?
After the strongest move 2 I .ixa6 �xa6
22.�e2! (22.El:d l ifS! 23.ttJxf6t cj:;>hS-+)
22 . . . El:xd6 23.ttJxd6 �xd6 24.0-0 �xd5 I
would keep an edge, but two connected passers
on the queenside offer White some chances.
b
c
d
e
f
g
h
The position remains tense and is far from the
technical stage of converting an extra piece. I
had to find precise and energetic continuations
to prove Black's (decisive) advantage.
26 ... gxf5!
I like this capture. The tripled pawns look
ugly, but Black gets control over e4.
27.gacl
After 27.c6 ifS or 27.d6 �c6 the pawns
would be blocked.
145
Chapter 4 - Line Opening
29 .. J�be2!
8
The point. The white king has to leave his
shelter.
7
6
30J�xe2
5
Or 30.:8:fl �e3t 3 1 .<tt> h l �xc l is the end.
4
30 ...\Wxcl t 3 I .<bfl \Wf4t 32.<bel gd4-+
3
The rest was easy and very pleasant for me
- I always like to attack my opponent's king
when I'm a piece up.
2
a
b
c
d
e
f
g
h
27 ...\Wa3!
The winning move; Black creates a direct
attack against the king. Because of that, White
has no time to advance the c- or d-pawns far
enough.
33.\Wast
Or 33.�f3 �c l t 34.<tt> f2 �xc5 3 5 .:8:eSt
ifs 36.�g3t El:g4t.
33 ....if8 34.\Wa3
8
28.d6
If 2S.c6 El:b2 29 .�d l Black would win
after: 29 . . . �xa2 30.�f3 El:g4 3 1 .g3 El:e4! 32.c7
El:xe l t 33.:8:xe l �a7t 34.<tt> h l �xc7 3 5 .:8:eSt
ifS 36.:8:xfSt <tt> g7! A nice variation!
2s .. J�b2 29.\Wds
Just as fatal for White is 29.�d l �xa2
30.�f3 El:g4! 3 1 .g3 ifS 32.d7 El:d4 33 .c6
:8:dd2.
7
6
5
4
3
2
1
a
8
b
c
d
e
f
g
h
34 ...\Wxh2!
7
Just grabbing some more material, preventing
the possible exchange of queens after �g3t,
and threatening checkmate simultaneously.
6
5
35.\We3 ge4 36.\Wfl g£4 37.\We3 .ih6
38.\West
4
3
Or 3S.d7 �h l t 39.<tt> d2 El:d4t.
2
a
b
c
d
e
f
g
h
38 ... <bg7 39.<bdl \Wgl t
0-1
1 46
King's I ndian Warfare
Evgeny Postny - Ilya Smirin
Maalot-Tarshiha 2008
I .d4 tiJf6 2.tiJf3 g6 3.c4 .ig7 4.tlJc3 0-0
5.e4 d6 6 . .ie2 e5 7.0-0 tlJc6 8.d5 tlJe7 9.b4
aS 10.bxa5
I have faced this move quite a few times in
my career. Still I think that after it Black's life is
easier, and the "more theoretical" 1 O.ia3 poses
more problems. But maybe it's j ust a matter
of taste.
1 o .. J�xa5 l l .a4
1 1 .lDd2 c5 1 2.a4 �a6 1 3 .�a3 happened in
a rapid game with Bareev. We can compare
this position with one from the game Aronian
- Radjabov, Sofia 2008: 9 . . . ttJ h 5 1 0.�e 1 a5
1 1 .bxa5 �xa5 1 2 .lDd2 ttJ f4 1 3 .ifl c5 1 4. a4
�a6 1 5 .�a3 lD h 5 It's obvious that I have gained
few tempos - my knight on f6 did not travel to
h5, f4 and back to f6. No wonder Black got an
excellent position after the opening: 1 3 . . . ttJ d7
1 4.ttJ b5 f5 1 5 .exf5 gxf5 1 6.ib2 lD g6 1 7.g3
ttJ f6 1 8 .f4?! exf4 1 9 .gxf4 ih6! 20.�g3 lD g4;
(Y2-Y2, 45) Bareev - Smirin, Moscow 2002.
l l . .. c5
a
b
c
d
e
f
g
activity there. On the other side, Black too will
have difficulties trying to start his kingside
play - White, as often in the KID, enjoys a
space advantage and may place his pieces
comfortably to meet such attempts.
12J3a3
I faced a slightly different setup by White
in the following encounter with Victor
Mikhalevski: 1 2.id2 �a6 1 3 .�b 1 lDd7 1 4 .Wc 1
�h8 1 5 .ig5 if6 1 6.ih6 ig7 1 7.ig5 if6
1 8 .id2 ig7 1 9 .�a 1 lD f6 20.�a3 tDe8 2 1 .lDb5
tD f6 22.tD c3 lDe8 23.lDb5 tD f6 24.Wb 1 ttJ fg8!?
Preparing the next move. 25.id 1 ih6 26.ic3
ig7 27.id2 ih6 28.ic3 ig7 29.tDd2 lD h6
30.Wa 1 g5! Vacating g6 for the knight on e7.
3 1 .ic2 lD g6 32 .ib2 ttJ f4 33.�e 1 f5 And after
several repetitions of moves, I managed to start
some offensive actions against the opponent's
king; (Y2-Y2, 65) Mikhalevski - Smirin,
Philadelphia 2008.
12 ... tlJe8
Another of my games continued: 1 2 . . . �a6
1 3. tD e 1 lD d7 1 4 .lD d3 tt:J b6!? (the immediate
1 4 . . .f5 was also not bad at all, but I wanted
to play something "more original") 1 5 .�b3 f5
1 6.exf5 ctJxf5 1 7.lDb5 tDd7 1 8 .f4 e4 1 9 .ctJf2
�e8 20.Wfc2 ttJ f6 2 1 .ib2 id7 22.ixf6 ixf6
23.lDxe4 ttJd4 24.tt:Jxd4 ixd4t 25 .�h 1 ixa4
26.�xa4 �xa4 27.Wxa4 �xe4 28 .id3 �e3
29.Wfc2 Wh4 30.Wb 1 Y2-Y2 Lputian - Smirin,
Kemer 2007.
h
Now we have a typical pawn structure
for this and similar lines - the queenside is
stabilized and it's not easy for White to develop
a
b
c
d
e
f
g
h
1 47
Chapter 4 - Line Opening
13.tiJel?!
I do not like this move. In my opinion the
knight should stay on f3 longer to meet . . . f7-fS
with lDgS, making use of the weak square on e6.
I think that White has stronger options here,
such as 1 3 .!d2 �a6 1 4.'\Wc l ! ? (aimed against
a possible . . . h6 to prevent the knight's leap to
gS, as in my game with Mikhalevski above) ;
or the prophylactic 1 3.g3!?, intending to play
lD h4 and wait for Black to play . . .fS, after
which White's reaction would be exfS . . . gxfS
f4!, stopping the opponent's activity thanks to
the excellent knight on h4. Actually there are
a few similar examples of this manoeuvre in
the book.
13 ... f5 14.exf5
favour: the square d4 for the black knight
weighs more heavily here than the square e4
for the white pieces.
16 ..id3
Another second-best move; my opponent
does not feel the danger yet.
Probably he should have continued 1 6 .!g5 h6
(in my opinion 1 6 . . . "1Wc7!? is more accurate)
1 7.!xf6 !xf6 1 8.lDd3 !g7 1 9.!g4 and due to
the strong knight on e4 the chances are about
equal.
16 ... tlJxe4 17 ..ixe4 .id7
8
7
8
6
7
5
6
4
5
3
4
2
3
1
2
a
a
b
c
d
e
f
g
h
14... tiJxf5!
I think Postny expected 14 . . . gxf5 1 5 .f4 lD g6
1 6.lDd3 e4 1 7. lDf2 (or 1 7.lD e l ) with mutual
chances, and underestimated the capture with
the knight.
15.tlJe4 t2Jf6
I would suggest that Black already has a
pleasant position: all his pieces are active
enough, which cannot be said about the
misplaced knight on e 1 . Moreover, the a4pawn, as we will see very soon, may become
vulnerable. There is yet another factor in my
b
c
d
e
f
g
h
Now White's problems are obvious. His
pieces lack coordination and the knight on e 1
is evidently misplaced.
18 ..id2 ga7 19.tiJf3
Back to the action.
19 ...'1Na8?!
Tempting, but not the best decision.
The position demanded the far more energetic
1 9 . . . b5! 20.cxb5 !xb S 2 1 .�e l and only now
2 1 . . . "1Wa8. After that the best continuation for
White seems to be 22.lD g5 !xa4 23."1Wg4 !d7
24.�xa7 "1Wxa7 25 ."1Wh3 h6 26.lDe6, hoping
148
King's Indian Warfare
to obtain compensation for the pawn. Still
after 26 . . ..ixe6 27.dxe6 liJ e7 28 ..ixh6 (bad
is 28 .'\Wd3 c4 29 . .ie3 WaS) 28 . . . d5 29 . .ixg7
<i>xg7 30.if3 c4 White would have to work
hard to hold the resulting position - Black
clearly dominates, but the limited material
leaves decent drawing chances.
24..ixd4 exd4 25.tiJg5
20.a5?
The other capture was less strong: 25 . . . �xa5
26.�xa5 '\Wxa5 27.liJe6, with the sample
variation 27 . . . �b8 28 .h4! ie5 29.h5 d3
30.�fl ! and unclear complications.
This unlucky pawn had to be sacrificed:
20.ttJg5 �xa4 (20 . . . liJ d4? 2 1 .ttJxh7 <i>xh7
22.'\WhSt) 2 1 .�xa4 ixa4 22.'\Wb l ttJ d4 23.ttJe6
liJxe6 24.dxe6 ic6 25 .ixc6 bxc6 26.e7
�e8 27.ig5 and here the drawish variation
27 . . . h6 28.'\Wxg6 hxg5 29 .'\We6t <i>h7 30.'\WfSt
<i>h8 3 1 .'\Wh3t <i>g8 32 .'\We6t looks a logical
conclusion.
8
7
As I mentioned above, this is the only real
attempt to get some counterplay.
25 ...hd5
The right choice: this pawn had to be
removed.
26.Y;Vg4 .ies 27J3f3 gxa
27 . . . �e8 was probably a cleaner way to
convert the advantage, for instance: 28.h4 h5
29.'\Wh3 ixe4 30.ttJxe4 c4 and the end is near
- the duo of pawns on d4 and c4 are too close
to the prize rank.
28 ..ixf3 d3
6
5
8
4
7
3
6
2
5
1
4
a
b
c
d
e
f
g
h
3
20 ... b5!
2
Now Black has managed to take over on the
queenside - a rare feat for this opening line.
1
a
b
c
d
e
f
g
h
2 1 .cxb5 .ixb5 22J3el .ic4
29 ..ixd5t
Suddenly the pawn on d5 has become very
weak and White has no time for the manoeuvre
ttJ f3-g5-e6. Without this, he can hardly hope
for counterplay.
Houdini "thinks" the move 29.�d l is the
most stubborn defence. This is indeed so, but
after some time the same Houdini shows a
forced win for Black:
23 ..ic3 tiJd4
29 . . . c4
Less precise is 29 . . ..ixf3 30.gxf3 (30.liJ xf3
'\Wd5) 30 . . . '\WdS 3 I .f4 if6 32.'\Wcst <i>g7
This knight j umps to d4 at the best possible
moment - now the d5-pawn is doomed.
149
Chapter 4 - Line Opening
33 .'\Wh3 (33.t2J e6t �h6 34.'\Wf8t ig7
35.l2Jxg7 El:xg7 36.'\Wf6 c4-+) 33 . . . ixg5
34.B:xd3 '\Wc4 3 5 . fxg5 '\Wc l t 36.'\Wfl '\Wxg5t
37.El:g3 '\Wf6 38.a6 with chances for survival.
30.ixd5t '\Wxd5 3 1 .'\Wc8t @g7 32.t2Je6t @f7
33.t2Jd8t
33.l2Jg5t �e7 is similar to what happened
later in the game.
33 . . . @f6 34.l2Jc6 B:f7 35 .a6
32.ttlg5t
Or if: 32.t2Jd8t @f6 33.t2Jc6 d2 34.B:d l
ixh2t 35 .�xh2 (35.�h l B:f7 36.'\Wh8t �g5
37.@xh2 B:xf2 3 8 .'\Wd8t @h5 39.El:g l d 1 =1W)
35 . . . '\Wh5t 36.@g3 '\Wxd l 37.'\Wh8t El:g7
38 .'\Wd8t @e6 39.'\We8 t @d5 The black king
travels quite a long way to victory.
32 <it>g7 33.ttl e6t <it>f7
•••
Time-trouble repetition.
8
34.ttlg5t <it>e7 35.ttlxh7 d2
7
5
8
4
7
3
6
2
5
b
a
c
d
e
f
g
4
h
35 . . . '\Wc5!
The only winning move.
36.a7
Or 36.'\Wh8t @g5 37.'\Wd8t @h5 and White
has no more good checks.
36 . . . @g7 37.El:fl d2 38.a8=1W '\Wxf2t 39.El:xf2
d 1 =1Wt
With mate.
29 %\'xdS 30.%\'cSt <it>g7 3 1 .ttle6t <it>f7
•.•
8
3
2
a
b
c
d
e
f
g
h
The black king is in safety, and the game is
essentially over.
36J�dl c4
36 . . . ixh2t would win too, but the move I
made is simpler.
37.%\'fBt <it>d7 38.ttlf6t hf6 39.%\'x£6 gxa5
7
This pawn did not escape its fate.
6
40.%\'xg6 ga3 4 1 .h4 c3 42.h5 %Yb3
5
4
3
2
a
b
c
d
e
f
g
h
King's I ndian Warfare
1 50
43 ...fie6 44.fig7t <if?c6 45JUJ. c2
8
7
8
6
7
5
6
4
5
3
4
2
3
1
a
b
c
d
e
f
g
h
43.fig4t?!
2
1
a
Shortening the suffering.
b
c
d
e
f
g
h
46.fib2 d l =fi 47.fixa3 fixfl t
43 .'Wg7t �c6 44.'Wg4 would prolong the
game, but after: 44 . . . 'Wa4 45 .'Wc8t �d5
46.'Wf5 t �d4 47.'Wg4t �c5 (Just not:
47 ... �d3 48.'Wf3t �c2? 49 .'Wf5t �xd l ??
50.'Wb l t �e2 5 l .'Wfl # Quite a picture!)
48.'Wc8 t 'Wc6 49.'Wf5 t 'Wd5 50.'Wc8t �d4
5 1 .h6 'Wh5 52.g4 'Wxh6 The inevitable would
happen anyway.
In view of 48.�xfl 'Wc4t White resigned.
0-1
Chapter 5
Destruction of Pawn
Structure
Ifyou cannot find a way, go deeper!
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
8
... 8
6
6
7
7
5
5
4
4
2
2
3
3
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
A difficult one. How can Black
Time to do something!
(see page 1 57)
improve his position?
(see page 1 66)
... 8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
Can you calculate this till the
What brilliant refutation of
end? (see page 1 5 3/4)
my idea had I missed?
(see page 1 62)
8
7
6
4
3
... 8
7
5
a
b
c
d
e
f
g
h
c
d
e
f
g
h
5
6
4
4
2
2
3
2
b
White has a direct win. Can you
find it? (see page 1 67)
... 8
7
6
5
a
3
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
How can Black strengthen
Time for the big combination!
Sometimes one nice move is all
the attacking resources?
(see page 1 56)
(see page 1 63)
you need. . . (see page 1 67)
1 53
Chapter 5 - Destructio n of Pawn Structure
In this chapter I will show three of my own
games with different variations on the theme
of the destruction of pawn structures. In the
first of my games I break through on the g-file,
though in a slightly different way than in our
first example below. In the second I remove
the structure on the dark squares, and in the
third we have a very complex example, which
I would suggest that you take more than just a
few ideas from.
In general, the destruction of a pawn structure
takes place on one colour of squares. By
removing the defence there, we create a way to
access the opponent's position. The following
two examples are pretty simple, which should
be a good way to come to grips with the basic
idea before we look at it in a dynamic scenario.
Aleksej Aleksandrov - Alexei Fedorov
The main defence of the white king is on the
light squares, so elimination of the defence of
these squares is essential.
30 ... tlJxg2!
Fedorov has played the King's Indian for
decades; this move will have come very easily
to him. The immediate 30 . . . f3 ? 3 1 .�xh4
would be very poor, of course. And defending
the knight would be a waste of time.
3 I.J.xg2 f3
The attack on the light-squared defenders
continues.
32.a5
32.�fl �h3t would contribute to the
complete destruction of the defence on the
light squares.
32 ...�h5!
The final finesse; Black does not want a
pawn on g2.
Minsk 2016
33. i>fl gxg2
0-1
8
7
The next example is also very basic. Black
comes crashing through on the dark squares.
6
5
4
Eesha Karavade - Pierre Ballet
3
Nancy 2016
2
8
1
a
b
c
d
e
f
g
h
The first example is elementary for a seasoned
King's Indian player. Black has positioned
himself for a classical breakthrough on the
g-file, set up already by . . . g5-g4, . . . hxg4 and so
on. We have reached an obvious crunch point:
the black pieces are as active as they are likely
to get, so anything other than direct action
would be a waste of time.
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 54
King's Indian Warfare
My Games
32 .. J��xb2t! 33J!xb2 �xb2t 34.'tt>xb2 Y«d2t
35.'tt> b l
Alexei Barsov Ilya Smirin
Or 3 5 .ci>a3 'Wb4#.
-
Khanty-Mansiysk Olympiad 2010
35 ... a3!
With a simple threat of mate in one.
35 . . . '\Wb4t 36.ci>c2 would be the wrong
direction, although Black can still repeat with
36 . . . '\Wd2t 37.ci>b l and continue as in the
game.
36.Y«e2 Y«b4t 37.'tt> a l
Or 37.ci>c2 'Wb2t 38.ci>dl 'Wc l #.
37 ...Y«c3t
l .d4 ttlf6 2.ttlf3 g6 3.c4 .ig7 4.g3 0-0 5 ..ig2
d6 6.0-0 ttlc6 7.ttlc3 .ifS 8.b3
37 . . . id2!? also wins immediately.
The curious-looking move 8 .ie3!?, which
has been played by Jon Hammer, among
others, definitely deserves attention.
38.'tt> b l
a
The 20 1 0 Olympiad in Khanty-Mansiysk was
different for me from other such events. It was
my ninth Olympiad playing for the Israeli
team and for the first time we won a medal bronze. In 2008 in Dresden our team had even
greater success - the silver medal - but for the
first time since 1 992 I had not participated in
the Olympiad. (Perhaps that was the reason for
such a great team result? Well, never mind.)
b
c
d
e
f
g
h
38 ....id2!
Cutting off the queen and deciding the
game.
0-1
a
b
c
d
e
f
g
h
8 ... e5
More than twenty years ago I used to play
the logical-looking 8 . . . ltJ e4, but the following
tough defeat changed my mind about that
move: 9.ltJd5!? id? 1 0.ib2 fS 1 l .e3 It's not
easy for Black to find a plan connected with
counterplay here. 1 l . .. a6 1 2.Ek1 bS 1 3.ltJd2
ltJgS 1 4.'\We2 b4 1 5 .c5 ci>h8 1 6.cxd6 cxd6
1 7.h4 ltJ f7
155
Chapter 5 - Destruction of Pawn Structure
9 ... tLle7 IO.tLlel tLle8 l l .e4 .id?
8
7
6
8
5
7
4
6
3
5
2
4
a
b
c
d
e
f
g
h
1 S .ltJf4 lDa5 1 9.ixaS �xaS 20.Ek7 �dS 2 1 .�fc l
e5 22.dxe5 lDxe5 23.h5 g5 24.h6 ifS 25.4Jd3
<i>gS 1-0 Cvitan - Smirin, Tilburg 1 993.
9.d5
The most principled choice, but perhaps not
the best one.
9 .ib2 leads to a dull and equal posmon
after 9 . . . 4J xd4 1 0.4Jxd4 exd4 1 1 .�xd4 ltJ e4
1 2.�e3 lDxc3 1 3 .ixc3 �e8 1 4.�d2 ixc3
1 5 .�xc3 ie4.
9.dxe5 dxe5 1 0.ig5 ! ? is not as innocent as it
looks. For instance: lO . . . �cS 1 I .lDd5 lDxd5
1 2.cxd5 ltJd4 1 3 .lDxd4 exd4 1 4.�c l �d7
1 5 .�c4 �fe8
a
b
c
d
e
f
g
3
2
1
a
b
d
c
e
f
g
h
Black's pieces have withdrawn from the f-file
to play . . f7 -f5 as soon as possible. It looks as
though Black has j ust lost a tempo compared
to the line with the immediate 7 . . . e5 S.d5
ltJe7, but in fact the "extra'' move b2-b3 turns
out to be in Black's favour.
.
12.tLld3 f5 13 ..id2
A similar scenario as in this game
happened in my blitz encounter with
Shakhriyar Mamedyarov from the World
Blitz Championship: 1 3 .c5 f4 1 4.a4 g5 1 5 .g4
Trying to close the kingside. 1 5 . . . h5 1 6.h3
hxg4 1 7.hxg4 ltJg6 1 8.f3 ltJ h4 1 9 .<j{f2 �f6
20.�h l �h6 2 l .ia3 ifS 22.b4
h
1 6.�xd4 ( 1 6.if4!?) 1 6 . . . �xe2 1 7.�xe2 ixd4
1 S .�c l �eS With equality, but I lost in the
end; ( 1-0, 35) Durarbayli - Smirin, Plovdiv
20 1 2.
a
b
c
d
e
f
g
h
22 . . . c6! Now the position opens up even more
and the king on f2 feels unhappy about that.
1 56
King's I ndian Warfare
23.i.fl ctJ f6 24.i.e2 cxd5 25.exd5
b
a
c
d
e
f
g
to see that, without making obvious mistakes,
White is clearly worse after just 1 4 moves! I
would again like to stress the critical role of
the "small" detail of the pawn being on b3 - it
completely changes the assessment.
h
25 . . . e4! 26.ctJxe4 ctJxe4t 27.fxe4 i.g7 With
a crushing attack; (0- 1 , 38) Mamedyarov Smirin, Khanry-Mansiysk (blitz) 20 1 3 .
8
a
7
b
c
d
e
f
g
h
15.c5 llJg6 16J�el �f7
6
The standard regrouping - the rook vacates
f8 for the bishop.
5
4
17.a4 i.f8 1 8.llJa2?
3
This manoeuvre looks artificial, and indeed
this is a weak move.
2
1
a
b
c
d
e
f
g
h
13 ... £4!
Starting a direct attack against White's king.
14J�cl
The simple variation 1 4.gxf4 exf4 1 5 .ctJxf4?
i.xc3 1 6.i.xc3 :gxf4 shows the drawback of the
white pawn being on b3 instead b2: it does not
protect the knight on c3 .
14 ...g5
Black's play is very simple: soon almost all of
his pieces will be involved in the attack. White
will find it hard to defend or to create his usual
counterplay on the queenside. It's surprising
Barsov should have continued 1 8 .b4, creating
some activity on the left Bank, planning to
put his queen on b3 and trying to deflect the
opponent from a direct kingside assault. In this
case the game could have continued 1 8 . . . ctJ f6
1 9 .Wfb3 h5 20.c6!? (20.ctJb5 also deserves
attention) 20 . . . bxc6 2 1 .dxc6 i.xc6 22.ctJb5
i.xb5 23.axb5 :gbs 24.:ga l :gb7 2 5 .ctJb2,
and at least White has active pieces and some
compensation for the missing pawn.
1 8 ... llJf6 19.c6
This pawn push was the aim of the previous
move. It proves to be harmless for Black
though.
19 ... bxc6 20.dxc6 i.g4 2 l .VMc2 h5!
1 57
Chapter 5 - Destruction of Pawn Structure
24 ... llJxd5 25.exd5 hxg3 26.hxg3 f3
8
This pawn is a bone in White's throat.
7
27J�e4
6
27.ifl ih3 28.�e4 ttJ f4! 29.ixf4 gxf4
would change nothing.
5
4
27... llJf4!
3
2
8
a
b
d
c
e
f
g
h
Black's centre is stable, so he can concentrate
on the attack. Now it's hard to give White any
good advice.
22.�c4 �c8 23.liJdb4 h4 24.llJd5?!
This loses virtually by force.
The only chance to prolong resistance consisted
of: 24.ttJ c3 hxg3 25.hxg3 ttJ h5 26.ttJ bd5
Occupying the strong outpost on d5 with the
knight. Even in this case, after: 26 .. .f3 27.ifl
lLlxg3 28.ttJf6t <j;;> g7 29.fxg3
7
6
5
4
3
2
a
b
c
d
e
f
g
h
A simple but pretty move, sealing the
opponent's fate.
28.gxf4 gxf4
28 . . . fxg2 was even simpler, for instance
29 .fxg5 �f5 30 .ie3 �h7.
29.i.xf4 exf4 30J�xf4
8
7
6
a
b
c
d
e
f
g
h
29 . . . ie7! This computer suggestion decides
the outcome, but it would be not so easy
to find this move over the board. 30.ttJxg4
(or 30."1Wa6 �xf6 3 l .ie3 �h8 and White is
doomed) 30 . . . �xg4 3 l .<j;;> f2 �h8 Black has an
irresistible attack.
5
4
3
2
1
a
b
c
d
e
f
g
h
1 58
King's I ndian Warfare
30 .. J�g7
30 .. Jhf4 3 1 .Wxf4 fxg2 32.Ek4 .id l would
be enough for the win, but I wanted to win
more quickly and decided to proceed with my
attack.
3 I..ihl
Slightly more stubborn was 3 l . .ifl :
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
33J��xf3
33 .�h4 .ie7 34.�h5 WeB! would bring no
relie(
33 ....ixf3 34..ixf3 'iMf5 35 ..ig2
a
b
c
d
e
f
g
h
And I would have to choose between:
3 I . .. .ih3t 32.�h l .ixfl 33 .Wxfl �h7t
34.�gl .ih6 3 5 .�xf3 �g7t 36.�h2 Wg4
37.Wh3 .if4t 38 .�h l Wxh3t 39.�xh3 .ixc l
40.ctJxc l �e8 4 1 .ctJd3 �e4
And the more resolute computer method
3 1 . .. �g5 32.�h2 .ih6 33.�el �g7 34.�xg4
Wxg4 3 5 .Wxg4 �xg4, both leading to a win.
This is not surprising - the white knight on a2
is situated too far from the battle.
3 I. ...ih3t 32.<it>h2 .ig2
If 3 5 .We4 Wg5 36 . .ig2 �e8 37.Wf3 Wd2
38 ..ih3 Wxa2 the unfortunate knight would
be lost, leaving his army lacking a whole rook.
35 ...YMxf2
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
But now the situation is bad enough to cause
immediate capitulation.
0-1
1 60
King's Indian Warfare
Mikhail Nedobora - Ilya Smirin
Israel 2008
This game was played in one of the numerous
(about 25!) Israeli Team Championships
I participated in. It's included in the book
because of a nice trap that I set, and its excellent
refutation, which my opponent failed to fi n d.
The opening phase is of some importance as
well in my opinion.
I.d4 �f6 2.c4 g6 3.�c3 �g7 4.e4 d6 s.�e2
0-0 6.�g5 �a6
8
b
a
d
c
e
f
g
h
And I managed to win this slightly more
pleasant endgame (two bishops) ; (0- 1 , 6 1 )
Milov - Smirin, Haifa 1 995 .
8.ie3 e5 9.0-0, as was played for instance in
Gelfand - Naiditsch, Wijk aan Zee 20 1 4, is
possible, but in my opinion the "extra" move
. . . h6 should be in Black's favour here. In my
game with Maxim Rodshtein, which you will
find in this book on page 278, I reached this
position with White to move!
7
6
5
4
3
8 ... g5 9.�g3 �hS
2
a
b
c
d
e
f
g
h
7.�f3
White's move may be seen as some kind
of hybrid of the Averbakh System with the
Classical System (5 .ie2 and 6.lLlf3). It looks
natural, and is hardly worse than more popular
continuations such as 7.iWd2 or 7.h4.
7... h6 8.�h4
This is definitely more logical than: 8 .if4
e5! 9.dxe5 lLl h5 1 O.ie3 (after the more
ambitious 1 O.g3 lLlxf4 1 1 .gxf4 lLlc5 I think
Black has enough compensation for the pawn)
1 o . . . dxe5 1 1 .0-0 c6 1 2 .iWxd8 �xd8 1 3 .�fd 1
�e8 1 4.g3 lLl f6 1 5 .lLl d2 lLl g4 1 6.ixg4 ixg4
1 7 .f3 ie6
8
7
6
5
4
3
2
,
���
)��-�J!� i �� �� i :t)u-----%·----%------��
�- %� �-�
�u !D !a u
�r.% � ��
"
�
�
r.;...
� !:.ll.J
�
;�_j
� rJ
RiJtJ!i"JJ
�f "� l�r %•�
a
b
c
d
e
f
g
h
The principled approach - Black is going
to exchange the dark-squared bishop despite
weakening his own kingside pawn structure.
This would work perfectly were it not for
another drawback - the knight on a6 is rather
poorly placed in these circumstances (even on
161
Chapter 5 - Destructio n of Pawn Structure
b8 i t would b e better, I suppose) . Still I feel
that Black should be 0 K - the bishop on g7
usually feels happy in such positions.
10.0-0
Back in 1 998 I played an interesting game
with Alexey Aleksandrov: 1 0.Wfd2 e6 (with the
idea . . . f7 -f5 , threatening j ust to win the bishop
on g3 after .. .f5-f4) 1 1 .e5 A very aggressive
move (probably 1 1 .0-0, transposing to the
game Nedobora - Smirin was objectively
better) .
1 7.lt:Jxf7! �xf7 1 8 .ih5 t �g8 1 9.ixe8 Wfxe8
20.Wfxg5 And Black was on the verge of defeat.
In the end I was lucky to escape; (Yz-Yz, 40)
Aleksandrov - Smirin, New York 1 998.
10 ...e6 l l .Y:Yd2 f5
Going for the bishop on g3.
Today I would probably choose another plan
- 1 1 . . . c5!? 1 2.d5 lt:Jxg3 1 3.hxg3 exd5 1 4.cxd5
( 1 4.ctJxd5!? El:e8 1 5 .id3 ie6 1 6.ctJe3 '2J b4
1 7.a3 lt:J c6!? seems to lead to a complex and
double-edged position) 1 4 . . . lt:J c7 Transposing
into a Benoni-type position. It should be
perfectly playable for Black. One small detail
may work in his favour - on d2 the queen
occupies a square which often belongs to the
f3-knight.
12.exf5 exf5 13.tlJel
Black has to decide what to do with the
knight on h5 .
a
b
c
d
e
f
g
h
1 1 . . . lt:Jxg3 ?! (As I found out after the game, I
had to play: 1 1 . . .f5! 1 2.exf6 Wfxf6 1 3 .0-0-0
ltJ f4 This is the idea - the knight on h5 has
found another, better, job. 1 4.h4 g4 1 5 . ctJ e 1
e 5 And Black i s doing great - I would be
happy to play this position.) 1 2.hxg3 dxe5
1 3.ctJxe5 c5 14.lt:Jg4! cxd4 1 5 .lt:Jb5 !!e8
1 6.ctJxh6t �f8
8
7
6
5
4
3
2
8
1
7
a
6
b
c
d
e
f
g
h
1 3 . g4
5
..
I thought this move was the most principled
one - the hunt for the g3-bishop continues.
4
3
2
a
b
c
d
e
f
g
h
The timid 1 3 . . . Wfe8 does not solve all the
problems: 1 4.ixh5 Wfxh5 1 5 .f4 id7 1 6.lt:Jc2
!!ae8 1 7.El:ae l with somewhat better prospects
for White - the black bishops are not very
1 62
King's Indian Warfare
active, White has a good central formation,
and we have already spoken about the knight
on a6.
But the simple 1 3 . . . ttJ xg3 1 4. hxg3 ttJ b4!? was
the optimal solution - the black knight has
joined the battle! After let's say 1 5 .a3 tD c6
1 6.tDc2 f4! ? 1 7.gxf4 El:xf4 1 8 .ttJ d5 El:f8 1 9 .El:fe 1
tD e7!? 2 0 ..if3 tD xd5 2 l ..ixd5t �h8 Black's
position seems to be fully acceptable.
It's interesting that even now after 1 6 . . . '\WgS !?
1 7.'1Wxg5 hxg5 1 8 .tDe7t �h7 1 9.tD dxf5 ixf5
20.tDxf5 El:xf5 2 l ..id3 �g6 22 . .ixd6 tD f4
23.ixf4 gxf4 24.El:fd 1 b6 Black would have
reasonable chances to hold this ending. But of
course, I did not even look in that direction.
17.tl:Jb5?
Mikhail trusts me. If, after 1 7.tDxf4 '1Wg5 , he
had found the only (but winning) move:
14.tl:Jd5 c5?!
Played in the same spirit; I found a possibility
to set a nice trap and went for it, not sensing
the danger.
Correct was 1 4 . . . El:e8 1 5 .tD c2 '1Wg5 1 6.El:ad 1
'lWxd2 1 7. El:xd2 .ie6 1 8 . tD f4 tD xf4 1 9 .ixf4
El:ad8 and White's advantage is minimal if any.
1 5.tl:Jc2 cxd4 16.tl:Jxd4
a
8
b
c
d
e
f
g
h
1 8 .h4!! gxh3 1 9.ixh5 El:xf4 20 .'\Wxf4 '1Wxh5
2 l .gxh3 he would have delivered a serious
blow to my ego. Alas, it remained big after the
game. More seriously, 1 8 .h4!! would have been
a brilliant refutation, but luck smiled on me
this time.
7
6
5
4
3
Of course, if 1 7.ixf4? then 17 . . . tDxf4 1 8 .tDxf4
'\Wf6! is a nice double threat; Black would win
fair and square.
2
a
b
c
d
e
f
g
h
Black's position is seriously compromised
- there are too many weaknesses. Still, I was
very inspired by the idea and made my next
(beautiful, I thought) move without hesitation.
16 ... f4?
1 6 . . . tD c5 1 7.tDb5 tD e4 1 8 .'1We3± would
lead to a difficult situation for Black - he has
insufficient dynamic options to compensate
for his structural disaster.
17 ... fxg3 1 8.hxg3
Certainly the resulting posmon should
be evaluated as winning for Black - White's
compensation for the piece is clearly
insufficient. But the win is not automatic here
- White's forces are quite active and Black still
has numerous weaknesses.
18 ... �h8?!
A prophylactic move of the "just in case"
kind.
1 63
Chapter 5 - D estruction of Pawn Structure
More accurate, logical and strong would be
1 8 . . . ltJ f6.
19.�adl .ie6 20.�e3
Mter 20.a3 (simply preparing 2 1 .b4) the
following variation is possible: 20 . . . .ixd5
(20 . . . ttJ f6? 2 I . ltJ f4 and suddenly White is
better; 20 .. .'1Wd7 2 1 .b4 �ad8 looks like a
good alternative) 2 l .Wi'xd5 ltJ f6 22.Wi'xb7 ltJ c5
23 .Wi'c6 �c8 ! ? 24.WI'xd6 Wl'e8 25 .Wi'd2 ltJ fe4
26.Wi'e3 a6 27 . .ixg4! ltJxf2! 28 .WI'xe8 �cxe8
29.�xf2 �xf2 30.�xf2 axb5 3 1 .cxb5 .ixb2
32 . .if3 �b8 33 . .ic6 .ixa3 I believe Black
should win this, even though it will not be at
all easy.
20 ...�b6 2 1 .b4?
Missing some simple but pretty tactics.
Mter the correct 2 1 .Wi'xd6 Wi'xd6 22.�xd6
�ae8 23.ltJxg4 (23 ..ixg4 .ixg4 24.ltJxg4 �e4)
23 . . . .ixb2 the extra bishop should tell in the
end, but again Black would have to overcome
some obstacles.
2 1 . .. �xb4!
Giving back the extra piece to destroy the
shelter ofWhite's king.
22.�xb4 �xg3 23.�fel �xfl!
This second blow leads to an irresistible
mating attack.
24..ixg4
Or 24.�xf2 ltJe4t.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
24 ... �fl t! 25.<it>h2 .ie5 26 ..ixe6 �e2t
The elegant 26 . . . �h l t 27.�xh l ltJ fl t
28.�h3 Wl'xe3t 29.�g4 Wi'f4t 30.�h5 Wi'g5#
would checkmate a bit faster.
8
7
6
27.g3 �xe3
5
White resigned. Maybe the game is not of
the very highest level, but it was an interesting
struggle nonetheless.
4
3
0-1
2
a
b
c
d
e
f
g
h
1 64
King's I ndian Warfare
Kiril Georgiev Ilya Smirin
-
Elenire 1 994
The strong tournament in the Bulgarian Black
Sea resort Elenite (near Burgas) went well
for me - I tied for first place with Topalov,
Khalifman and Epishin. This game contributed
a lot to this success.
l .d4 tlJ£6 2.c4 g6 3.tiJc3 .!.g7 4.e4 d6 s.a
tlJc6
This time I wanted to experiment a bit and
to try to postpone castling.
Kiril Georgiev chooses the most active move.
The more popular and cautious 7.ie3 (the
d4-pawn!) would be another option.
c
It's time to castle.
d
e
l l .g3 b5 would be good for Black. For
instance, 1 2.cxb5 axb5 1 3 .d5 ? tt::l e5 1 4 .tt:Jd4
b4, and 1 5 .tt::l cb5 ? ixb5 1 6.ixb5 !!xb5
1 7.tt::l xb5 tt::l x f3t is impossible of course.
l l . .. bS
7.. J�b8 SJ�cl .!.d7 9.�d2
b
10 ... 0-0
II.tiJdS
6. tlJ ge2 a6 7 ..!.gS
a
In my opinion this move plays into Black's
hands - White's kingside becomes more
vulnerable.
Grandmasters of today would probably
continue here 1 0.b3!? 0-0 l l .h3 with the
idea g4, or l O.tt::l d l ! ? with the intention
of transferring this knight to e3 or f2 and
meeting . . . b5 with c4-c5. But in 1 994 such
subtle prophylactic manoeuvres were mostly
undiscovered.
f
g
h
a
b
c
d
e
f
g
h
9 ... h5!?
12.c5
White's setup is aimed against 9 . . . b5?. It
would not be good here because of 1 O.cxb5
axb5 l l .d5 tt::l e5 1 2. tt:J d4 b4 1 3 .tt::l cb5 . That's
why I decided to open a second front on the
kingside, even though this move looks rather
abstract.
The pawn exchange 1 2.cxb5?! axb5 would
weaken White's control over the centre.
IO.h4?!
The correct decision.
1 2 .. J�e8
White's play is not without some tactical
venom.
1 65
Chapter 5 - Destruction of Pawn Structure
1 2 . . . ltJ h7? would lead to immediate disaster:
1 3 .cxd6 cxd6 1 4J!xc6 ltJxg5 1 5 .Wxg5t .ixc6
1 6.ltJxe7t @h7 1 7.ltJxc6
Now . . .f6-f5, undermining White's central
pawn chain, may be very unpleasant for him.
13.g3
The undermining . . . f5 can be seen in the
variation 1 6 . .if4?! f5! 1 7 . .ixd6? fxe4 1 8 . fxe4
:gxe4 1 9 . .ixb8 'Wxb8, and Black has an
overwhelming position.
With the modest objective of completing
development. If White manages to do so,
and keeps his strong central grip, he will have
an edge. That's why Black has to organize
counterplay rather quickly.
16.Ae3
a
b
c
d
e
f
g
h
16 . . f5
.
a
b
c
d
e
f
g
h
13 ... a5!?
Also possible was something like: 1 3 . . . ttJ h7
14 ..if4 ( 1 4 . .ie3?! dxc5 1 5 .Ei:xc5 e6 1 6.ttJ df4
e5! 1 7.dxe5 ltJxe5) 1 4 . . . Ek8 (the immediate
14 . . . e6?! 1 5 .ttJxc7 'Wxc7 1 6 ..ixd6 leads to
White's advantage) 1 5 . .ig2 e6
But Black made a more concrete and ambitious
move - to grab some space on the queenside.
Also, the potential knight j ump to b4 (after an
exchange on d5) may cause worries for White.
14.cxd6
Kiril decided to release the pawn tension.
14 ... cxd6 1 5.llJxf6t
1 5 . .ig2? was bad due to the simple 1 5 . . . ttJxd5
1 6.exd5 ttJ b4.
15 ... exf6
I need not have rushed with this obvious
move.
As my opponent pointed out after the game,
1 6 . . . a4! ? , with the idea . . . ttJ a5 would have
posed serious problems for him. For example:
1 7 . .ig2 ttJ a5+ 1 8 ..if4 'Wb6 1 9 ..ixd6 ( 1 9 . 0-0
f5) 1 9 . . . ttJ c4
1 7.b3 axb3 1 8.axb3 ttJa5 1 9.:gb 1 ( 1 9.'Wd3 b4!
20 ..ig2 .ib5) 1 9 . . . d5! 20 . .ig2 .ic6!+
In both cases White's position is on the verge
of collapse.
1 7. @f2! This computer suggestion seems to be
best, as is often the case.
Still, after 1 7 . . . ttJ e7!, supporting the pawn
pushes .. .f5 or . . . d5, I would love my prospects.
1 7 .. .f5 1 8 .exf5 would be welcomed by
White; 1 7 . . . ttJa5 is not effective either because
of 1 8 .ttJc3 .
1 66
King's I ndian Warfare
17.1g2?!
This very natural developing move proves to
be inaccurate in this tense situation.
Better would be the more principled though
risky 1 7.exf5 , which would promise an
interesting fight ahead.
8
7
2 l .�f4 the pawn on f7 is under attack - a
direct consequence of 1 7 . . . fxe4? .
Still I could have found some prudent
continuation like 1 8 . . . b4!? 1 9.0-0 lD a7,
intending to bring this knight to b5 with a
somewhat worse but reasonable position. But
I was already driven by adrenaline, and wanted
to play as actively and riskily as possible.
19.exf5
I began to feel that something had gone
wrong, but fortunately I did not have much
choice but to try to keep things messy.
6
5
4
19 .. J!c8!?
3
Relatively the best decision.
2
1 9 . . . �e7 20.cj{f2 �bc8 2 l .fxg6 �f6t 22.lDf4
would be hopeless.
1
a
b
c
d
e
f
g
h
17 ... fxe4?
But this premature exchange is clearly
wrong.
1 7 . . . a4! would probably be even stronger than
on the previous move. After that I do not see
an entirely satisfactory continuation for White.
For instance: 1 8 .exf5 lDa5 1 9.fxg6 fxg6
20.ig5 �b6 2 1 .0-0 lD c4
Or 1 8 .0-0 lD a5 1 9 .Ek2 lD c4 20.�c l lDxe3
2 l .�xe3 �b6.
And finally 1 8 .b3 axb3 1 9.axb3 �a8!? 20.0-0
�a3 . And in all three variations it's hard to
envy White's position.
1 8.fxe4
White is going to castle, after which he
would be happy, with a good pawn centre and
well-placed pieces.
1 8 ... f5?!
1 8 . . . a4 1 9 .0-0 lDa5 would have not nearly
the same effect as before - after 20.ih6 lD c4
1 9 . . . �b6 20.0-0 �xe3 2 l .�xe3 lDxd4
22.id5t cj(hs 23.cj( h 1 �e8 24.�g5 also leads
to disaster.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
20.0-0!
Completing development.
After the less strong 20.fxg6 �f6 Black would
manage to keep the white king in the centre,
with practical chances.
1 67
Chapter 5 - Destruction of Pawn Structure
23J�xc8 hc8 24.llJf4?
20 ...YNe7 2 I.�d5t ct!/hs 22J�f3?
Oops! This is wrong. Now my bluff pays off.
Georgiev fails to find the precise and powerful:
22.ih6! Wffxe2
This obvious move, which was made in the
approach to time trouble, immediately brings
White very dose to defeat.
24.f6! ixf6 25.lt:Jf4 was required, with the
small but vital difference that after 25 . . . if5
(better is 25 . . .Wff g7 with approximate balance)
26.a3 CLJxd5 27.CLJxd5 the bishop on f6 will
be hit by the white knight, and the position
arising will be in White's favour.
8
7
6
5
4
24...�xf5 25.�e6
3
2
b
a
c
d
e
f
g
h
23.f6 is tempting, as 23 . . . if8? 24.ig7t! leads
to checkmate, and 23 . . . ixf6 24J�xf6 Wffxd2
25 .ixd2 CLJ xd4 26Jhc8 ixc8 27.ixa5 ifS
leaves Black with only slim chances to hold.
But 23 .ixg7t! @xg7 24.f6t @h7 25 .ig8t!!
is even more convincing. Black can only resign.
The key point is 27.f8 =Wit!.
Obviously Georgiev has been counting on
this move. Objectively speaking his position
was already worse than "so-so". For instance,
25 .ib3 would j ust lose after 25 . . . Wff e4.
25 ...YNb7!
Exposing the rook's awkward position.
26.YNg2
8
7
8
6
7
5
6
4
5
3
4
2
3
1
2
a
1
a
b
c
d
e
f
g
h
22 ... llJ b4!
This unexpected knight j ump changes the
picture. Now White has to find a way to hold
his own in tactical complications.
Probably Kiril was counting on 22 . . . gxf5 ?
23 . lt:Jf4, with a big advantage.
b
c
d
e
f
g
h
26 ...�g4!
In my opinion, this is the best move from a
practical point of view, especially considering
that my opponent was experiencing a serious
lack of time. It required some calculation.
26 ... ixe6? 27.CLJxe6 �xe6 28.�f8t ixf8
29 .Wixb7 �xe3 30 .Wif7 would lead nowhere.
1 68
King's Indian Warfare
26 . . . .ie4 27.d5! would bring unnecessary
complications. After let's say 27 . . . gxe6!?
28.dxe6 .ixf3 29.e7! Vfixe7 (29 . . ..ic6??
30.e8=Vfit) 30.Vfixf3, White would keep good
drawing chances.
28.llJxd3 .ixB
28 . . ..ixe6!-+ says the computer.
8
27.<i>f2?
The final mistake, but the choice was very
tough.
7
27.lt:Jxg6t mh7 28 ..ixg4 hxg4 would lose on
the spot.
5
27.d5 gxe6! 28.lt:Jxe6 Vfixd5-+ is the same
outcome.
Only 27.ixg4 hxg4 28.gfl Vfixg2t 29.lt:Jxg2
lt:J c2 30 ..if2 lt:J xd4 3 1 .lt:Je3 would have
prolonged resistance. But I believe that even
in this case Black should gradually win after
3 1 . . . ge4.
6
4
3
2
a
b
c
d
e
f
g
h
The complications are over and a prosaic
ending with an extra exchange has arisen. Its
evaluation is clear - Black should win easily.
32.<i>e4 <i>gs 33.b3
8
33 .md5 would be slightly more stubborn,
but after 33 . . . mf7 34.mc6 b4 3 5 .mb5 gfst
36.mc6 .ih6! 37.mxd6 g5 White is doomed
anyway. Now it's even simpler.
7
6
5
4
33 ....ih6 34.a4 bxa4 35.bxa4 <i>f7 36.d5
.ix£4 37..ixf4 ci>e8 38.ci>d4 <i>d7 39.ci>c4
�f5 40.ci>d4 �£8 41 .ci>d3 �b8
0-1
3
2
1
a
b
c
d
e
f
g
h
27 ... llJ d3t
The simplest.
As usual, the artificial intellect insists on
stronger options, like 27 ... gxe6! 28.lt:Jxe6
lt:J d3t 29.me2 Vfic6 30.lt:Jxg7 Vfic2t 3 l ..id2
mxg7, with total domination. But after all,
you can't win the same game more than once!
Chapter 6
Knight Agility
Also with White I have always been able to
make good use ofthe kn ights...
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering
these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
... 8
7
6
5
4
3
2
2
a
b
c
d
e
f
g
h
a
d
e
f
g
8
6
6
6
5
4
4
2
2
3
a
b
c
d
e
f
g
h
a
White is quite weak on the
dark squares (see page 203)
2
2
2
3
d
e
f
g
h
As a start, don't lose the queen
(see page 1 8 1 )
f
g
h
5
4
c
e
6
4
b
d
What should Black play?
(see page 2 1 O)
4
a
c
Another day at the office . . .
5
3
b
... 8
7
6
5
h
7
... 8
7
7
g
did? (see page 208)
... 8
Train you KID muscle
(see page 1 76)
f
(see page 200)
... 8
h
e
Can you navigate the
3
c
d
complications better than I
5
b
c
Positional play. What is
Black's best move?
7
a
b
3
a
b
c
d
e
f
g
h
Simple tactics (see page 205)
a
b
c
d
e
f
g
h
What is White's only move?
(see page 2 1 2)
171
Chapter 6 - Knight Agility
When I was talking about the games for this
book with my publishers, they pointed out that
some of the games were centred on the knights.
I have to admit I had never thought about this
and do not have a great secret to share about
the knight's role in the King's Indian - only
that it is quite natural that the knights would
be prominent pieces in some games, as the
King's Indian often leads to closed positions.
Yes, we want the bishop to break out, often
with dynamic line opening, but sometimes
this is not possible and the knights will have to
field the majority of the activity.
As you will see from my own games in this
chapter, the main playing field for the knights
is the dark squares. A knight comes to d6,
blocking a pawn, and is very powerful. It
makes it to the more active square e5, where
it has a big range. Or it goes to f4 and aims at
targets behind enemy lines. Or it can go to d4
and dominate the opponent.
On the way it will go mainly via h5 and
f5 , but these squares have a tendency to be
transition squares. The same is the case for
Kasparov's knights in the lightly-annotated
example below. This is quite natural when you
stop to ponder for a moment. Black's pawn
structure in the King's Indian is on the dark
squares and White dominates the light squares
with his pawns. A knight needs support to
do its best (and worst!) and who is better at
supporting a knight than a pawn?
Do pay attention to the dynamic potential
of the knights in the King's Indian. Without
the knights there would be far fewer tactics in
chess; and without tactics the King's Indian
would not be the King's Indian . . .
But I digress. Let's have a quick look a t some
textbook knight-handling from a great King's
Indian player.
Alexander Khalifman - Garry Kasparov
Paris (rapid) 1 99 1
I n this rapid game from his King's Indian
prime, Kasparov manages to make 1 3 knight
moves before crashing through on the kingside.
As the former World Champion's favourite
number was 1 3 , this must have pleased him
immensely.
I.d4 tl)f6 2.c4 g6 3.tl)c3 i.g7 4.e4 d6 s.tl)f3
0-0 6 ..ie2 e5 7.0-0 tt)c6 8.d5 tl)e? 9.tl)d2
a5 IO.a3 tl) d? l l .�bl f5 12.b4 i>h8 13.8
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
I3 ... tl)gs
The move Kasparov introduced against
Karpov (see page 1 8 3).
14.�c2 axb4 1 5.axb4 tl) df6 16.c5 tl)h5
17.tl)c4 tl)f4 1 8.cxd6 cxd6 19.tl)b5 �a6
20.�b3?!
This move is thematic in some lines, but here
it does not really work.
20 .ie3 was preferable, but after 20 . . . lt:J f6
Black clearly has a good game.
20 ... fxe4 21 .fxe4
1 72
King's Indian Warfare
25 ... tiJh6!
8
The knight aims for the e5-square.
7
25 .. .f3 26.gxf3 '.Wg5 t 27.<it>h l '.Wc l was also
strong, but by no means a very human solution.
6
5
26.h3 tiJf7 27.tiJe2
4
3
2
1
a
b
c
d
e
f
g
h
2 1 . ...id7! 22. ttJ c3
22.ttJcxd6 �xd6 23.ttJxd6 '.Wb6t 24.ie3
'.Wxd6 also looks very pleasant for Black.
22 .. J�al
The position is already a bit uncomfortable
for White. He decides to get rid of the strong
black knight, but this dramatically weakens his
dark squares.
23 ..ixf4?!
23.if3 was better, but Black has a promising
position nonetheless.
23 ... gxfl t 24.-ixfl exf4 25.%Vf2
27.'.Wxf4 would lose very quickly to some
rather natural moves: 27 . . . id4t 28.<it>h l ttJe5
29 .'\Wc l '.Wf6, and the attack is decisive. We
can add 30 .ie2 ixh3 to the line if you like. I
am certain that it would be within Kasparov's
powers to see such a tactic, even in a rapid
game.
27 ... tlJg5!
8
7
6
5
4
3
2
1
a
8
b
c
d
e
f
g
h
We should always stay flexible. When the
opponent prepares for one idea, it is good to
have another he had not anticipated ready.
7
6
5
28.tiJxf4?
4
This loses immediately, but the position after
28.ttJd2 '.We7 would also not hold for long.
3
2
1
a
b
c
d
e
f
g
h
28 ...%Vf6 29.g3 tlJxe4 30.%Ve3 tlJxg3 3 1 .%Vxg3
%Vxf4 32.%Vxf4 gxf4
0-1
1 73
Chapter 6 - Knight Agility
My Games
When you play through my games you will
probably realize that the knights do an awful
lot of work, only to be eliminated once it is
time to reap the rewards. This is perfectly
natural. Chess is not only a game of piece
coordination and harmony, it is also about
destruction and ruthlessness. So do not get
too attached to your knights, j ust because you
realize they have great potential.
Besides this move, Black can choose from
three main different plans: 6 . . . tLl bd7 and
. . . e7-e5, 6 . . . c5, and 6 . . . c6. I have played all of
them in my practice, but in the last few years
I have switched almost exclusively to 6 . . . tLlc6.
7.0-0
Mesgen Amanov - Ilya Smirin
Kin g of Prussia 2010
l.d4
The World Open (sounds good, doesn't it?)
in Philadelphia is a tournament I played in
more times than any other event in my career.
In 20 1 0 the World Open took place in a town
near Philadelphia with the romantic name of
'King of Prussia'.
a
b
c
d
e
f
g
h
7 ...i.f5
The system with a fianchetto of the light­
squared bishop has a very solid reputation. Its
main aim is to restrict the opponent's potential
activity and obtain a small but lasting edge.
Here too Black had a wide choice: 7 . . . a6,
7 . . . e5 or 7 . . . ig4. I like the move in the game
- Black continues with development and
stays flexible with his pawn structure. Besides,
. . . ttJ e4 becomes a possibility. Of course, how
to continue here is a matter of taste.
3...i.g7
8.d5
l. .. tiJf6 2.c4 g6 3.g3
Perhaps the simplest way to fight for equality
is 3 . . . c6 with . . . d7 -d5 to follow, transposing
into the Griinfeld. The then World Champion
Carry Kasparov played it a few times against
Anatoly Karpov in their matches in 1 986 and
1 987. All the games ended peacefully. I have
also played this line quite a lot, mostly when I
did not mind a draw. But in this game I was in
a more aggressive mood.
4.i.g2 0-0 s.tlJc3 d6
After 5 . . . c6 White plays 6.e4.
8.b3 happened in my game with Barsov,
which is also included in this book.
8 ... tlJa5 9.tiJd2
9.tLld4 is another option here. Mter 9 . . . id7
(9 . . . tLl xc4? 1 0.tLlxf5 gxf5 1 l .�d3) 1 0.b3 c5
l l .dxc6 bxc6 ( l l . .. tLlxc6 was played a few
times by Boris Gulko, among others) 1 2.ib2
:gb8 1 3 .:gb 1 �c8 1 4.:ge1 :gd8 1 5 .�d2 tLlb7
1 6.e4 e5 1 7.tLlde2 ih3 1 8.f3 ixg2 1 9.�xg2
tLlc5 20.ia3 tLl e6 2 l .:gbd 1 if8 22.�e3
h5 23.:gd3 :gb7 24.:ged 1 :gbd7 I had good
position and later managed to win, Greenfeld
- Smirin, Haifa 1 993.
1 74
King's I ndian Warfare
9 . c5
..
I came to dislike 9 . . . c6 after the following
game:
a
a
b
c
d
e
f
g
h
1 0.b4 ttJxd5 l l .cxd5 ixc3 1 2.e4 ixa l 1 3 .exf5
ig7 1 4 J�e l ?! ( 1 4.bxa5 cxd5 1 5 .ttJ b3 ±)
1 4 . . . cxd5 1 5 .bxa5 gxf5 1 6.ixd5 W/xa5 1 7.ttJc4
Wlc7 1 8 .if4 El:ad8 1 9.Wlb3 b6 20.if3 El:c8
2 l .id5 El:cd8 22.h4 e6 (22 . . .W/c5 !) 23 .ixe6
fxe6 24.ttJxd6 �h8 2 5 .El:xe6 Wfc3 26.W/d5 El:d7
27.W/b5 El:dd8 28 .El:e7 W/c5 ? 29 .Wfxc5 bxc5
30.ttJf7t El:xf7 3 1 .El:xf7 c4 32.El:xf5 c3 33 .El:c5
id4 34.El:c4; ( 1 -0, 50) Ivanchuk - Smirin,
Odessa 2007.
Objectively speaking this line is not clearly
bad. I decided to give it another try in my
game with Michalik, where I played the better
1 o . . . ttJxc4. This game is mostly interesting
for the endgame, so I give the game in full:
1 l . ltJxc4 ttJxd5 1 2 .ttJxd5 ixa 1 1 3 .ih6 ig7
1 4.ixg7 �xg7 1 5 .ttJ f4?! (better is 1 5 .ttJde3)
1 5 . . . e5 1 6.e4 ixe4 1 7.ixe4 exf4 1 8 .W/xd6
fxg3 1 9.hxg3 b5 20.W/xd8 El:axd8 2 l . ttJ a5 El:d4
22.if3 El:xb4 23.ttJxc6 El:b2 24.ttJxa7 El:d8
2 5 . a3 El:d3 26.ic6 El:c3 27.El:a 1 El:cc2 28.El:fl f5
29.ig2 El:c3 30.El:a 1 El:cc2 3 1 .El:fl El:c4 32.El:e 1
f4 33.ttJxb5 fxg3 34.fxg3
b
c
d
e
f
g
h
34 . . . El:g4 3 5 .ifl El:xg3t 36.�h 1 El:g4 37.ttJd6
El:b6 38.ttJc4 El:f6 39.ttJd2 El:d4 40.ttJb3 El:a4
4 l .�g2 El:g4t 42.� h 1 El:f3 43.ttJd2 El:xa3
44.ig2 El:a2 45 .El:e7t �h6 46.El:d7 El:b4 47.ttJfl
El:bb2 48.ttJe3 El:e2 49.El:d3 �g5 50.�h2 El:ad2
5 1 .El:xd2 El:xd2 52.�g3 h5
a
b
c
d
e
f
g
h
53 .�f3 �h4 54.ifl El:d6 5 5 .ttJg2t �g5
56.�g3 El:b6 57.ie2 El:b4 58.if3 El:b3
59.�f2 h4 60.ie4 El:b2t 6 l .�e3 h3 62.ttJe1
h2 63 .ttJf3t �h5 64.ttJe5 g5 65.ttJd3 El:b 1
66.ttJf2 El:e l t 67.�d2 El:xe4 68.ttJxe4 �h4 0- 1
Michalik - Smirin, Jerusalem 20 1 5 .
I O.e4 �d7
It seems that Black has simply lost time
by moving this bishop twice in the opening.
However, it is unclear if e2-e4 is so useful for
White. There are three reasons for answering
in the negative: the scope of the bishop on g2
is restricted, the d4-square can no longer be
controlled by the e-pawn, and the e4-square is
Chapter 6
-
unavailable to the knights for the foreseeable
future. Of course, the strong pawn centre built
by White may become a cause for concern for
Black.
1 75
Knight Agility
13.b4 tiJb7 14.�b2 tlJg4
A pretty standard manoeuvre in this
variation; Black prepares for the . . . f7 -f5 push.
1 5.h3 tiJh6
l l .�c2
Both black knights occupy far-from-central
positions, but the closed character of the game
allows them to do so without substantial
damage.
8
7
16.tlJe2 f5 17.£4
6
5
8
4
7
3
6
2
5
a
b
c
d
e
f
g
h
l l . .. e5
In this game I decided to close the centre
and prepare to play on the kingside.
A year earlier I tried the more popular 1 1 . . . a6
in a rapid game against Boris Avrukh - 1 2.b3
b5 1 3.ib2 !!b8 1 4.!!ae 1 e5 1 5 . 'Li d 1 'Li g4 1 6.h3
'Lih6 1 7.ic3 f5 1 8 .f4 b4 1 9.ib2 exf4; (V2-V2, 33)
Avrukh - Smirin, Netanya 2009 . Black had the
worse position in the course of this game, so I
chose a different line against Amanov.
12.a3
4
3
2
1
a
b
c
d
e
f
g
h
It's becoming interesting - the pawn tensions
are growing, and even more so after Black's
next move.
17... b5!
Now the battle is raging in the centre and on
both flanks.
18.fxe5 dxe5 19.bxc5?!
In 20 1 2 Mesgen Amanov played the
definitely stronger 1 2 .b3 in one of his games
and won pretty quickly. The main problem
for Black in this system (as in the analogous
Yugoslav Variation) is the rather unfortunate
position of the aS -knight on the edge of the
board. It is hard for him to join the kingside
battle from there.
This wins a pawn, but leads to a serious
activation of Black's forces. More prudent was
a waiting move like 1 9 .l.Wb3!?, not releasing
the tension prematurely. In my opinion the
question "to keep or to release the pawn
tension?" and the timing of it is one of the
subtlest problems in chess.
12 ... b6
19 ... tlJxc5 20.cxb5 �c8
Vacating b7 for the knight.
1 76
Ki ng's I ndian Warfare
26.tt:Je4? �c2 27.�f2 'lMb6 28.@h2 ltJ e3 would
lead to a bad position for White - Black is too
active and material losses are unavoidable.
a
2 1 .ltl c3 'ffg5
b
d
c
e
f
g
h
26 ... e4!
2 1 . . . ltJ f7!? deserved serious attention. The
knight will go to d6 and later Black can decide
where to put the queen - on g5 or perhaps
somewhere else.
By sacrificing the second pawn Black opens
the a l -h8 diagonal for the g7-bishop, and the
pin on the knight at c3 becomes very sensitive
for White.
22JU3 fxe4
27.'ffxe4 ges
Another plan was 22 . . . f4 23.�afl with a
complicated game.
Stronger was the computer's suggestion:
27 . . . ltJd6 28.�xf8t 'lMxf8 29.'lMb4 �c4! 30.'lMb3
23.ltl dxe4 ltlxe4 24.'ffxe4 ltlf5 25.'ffe l
'ftd8!?
Suddenly the queen retreats. She switches
her action from the kingside to the opposite
side of the board.
25 . . . e4? does not work because of 26.ltJxe4,
but now this move is threatened, which my
opponent did not notice. In addition, the
queen may now become active via b6 or a5.
26.a4?!
Better was the subtle prophylactic move
26.�b 1 !, protecting the bishop on b2 if 26 . . .
e 4 27.ltJxe4.
a
b
c
d
e
f
g
h
30 . . . ixh3! 3 l .�fl id4t 32.@h2 if5 ! with a
devastating attack, but I failed to spot it.
28.'ffd3 ltld6 29.gdl
The attempt to free himself from the pin
by means of 29.ltJ d l ? does not work due to
1 77
Chapter 6 - Knight Agility
29 . . . El:e l t 30.El:fl "1Wb6t 3 1 .�h2 ifS , with a
total collapse.
29 ... tlJc4 30 ..ial tlJe5 3 1 .�d2?
Thanks to the active rooks, Black has good
winning chances in this endgame, but the
outcome is far from obvious. After all, White
is OK in terms of material.
Leading to an immediate loss.
35 ...Lal 36.gxal gf8
Much more stubborn was 3 1 ."\Wfl ltJxf3t
32."1Wxf3 "1Wb6t 33.<;t>h2 "1We3 (33 . . . El:e3 34.a5!)
34.ltJe4 "1Wxf3 3 5 .ixf3 El:c2t 36.�g l ixh3
37.d6 and White is not yet doomed.
Or 36 . . . ixh3t 37.�xh3 El:xf3 38.ltJd4 El:d3
39.ltJc6 El:c7 40.El:e l and this position is hard
to win - the knight on c6 is very strong.
37.tlJgi gcs
An interesting psychological moment - I
hoped that Amanov would not repeat moves
but would try to find something better. My
hope proved to be j ustified.
8
7
38.ga2?!
6
Mter the correct 38 .ltJ e2 Black would
probably have nothing better than 38 . . . ixh3t,
transposing into the previous line.
5
4
3
8
2
7
a
b
c
d
e
f
g
h
32 ...�b6t?
6
5
I briefly lost my concentration. 32 . . . "1Wf6!
33.ig2 El:xc3 would win easily.
4
33.�g2 �e3 34.�xe3 gxe3 35.tlJe2
2
3
1
8
a
7
b
c
d
e
f
g
h
38 ... gcc3! 39. �fl �f8
Now Black's plan is to exchange one pair
of rooks and collect the defenceless a- and
b-pawns after that. It's difficult for White to
prevent this with his passive pieces.
6
5
4
3
40.h4?!
A better try was 40.El:d2 El:ed3 4 1 .El:b2 El:b3
42.El:e2.
2
a
b
c
d
e
f
g
h
1 78
King's I ndian Warfare
40 ... <tt> e7 4l .g4?!
Boris Chatalbashev - Ilya Smirin
Still 4 l .El:b2 El:b3 42.El:c2 was more stubborn.
Mter the king reaches d6 the win becomes
rather simple.
41 ... <tt> d6 42 ..ie2 �a3 43.�b2 �eb3 44.�xb3
�xb3
8
7
6
5
I had won two tournaments in Athens, the
Acropolis Open, in 2007 and 2008. Alas, after
that the tournament was held in Khalkidhiki
in 2009 and my result was poor. Probably the
ancient spirit of Athens (and also an excellent
swimming pool on the roof of the hotel, which
I used every day before rounds) influenced me
in a very positive way. This game was played
during my first successful Athens event.
l .d4 tt)f6 2.c4 g6 3.tlJc3 .ig7 4.e4 d6 s ..id3
4
3
2
a
b
c
d
e
f
g
h
Black has fulfilled his plan while White's
pieces remain inactive. The rest is a simple
matter of technique.
45.h5 �b4 46.hxg6 hxg6 47.<tt> g3 �xa4
48.<tt> h4 �b4 49.<tt> g5 .ixb5 50 ..ixb5 �xb5
The a-pawn is out of reach of the white
knight, so White resigned.
0-1
Athens 2007
White chooses the system with the
development of the light-squared bishop to
the more active square d3 compared to e2. The
only drawback of this setup is that the pawn
on d4 becomes more vulnerable without the
queen's protection.
5 ... 0-0 6.tlJge2
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
6... tlJc6
I like this move - Black immediately takes
aim at d4. Actually there are a few different
paths to choose from. It's possible to transpose
into the Benoni with:
6 . . . c5 7.d5 e6
1 79
Chapter 6 - Knight Agility
An example between high-level opponents
is:
8.h3
Instead one of my games went:
8.0-0 exd5 9.cxd5
a
b
c
d
e
f
g
The other way is the immediate: 7 . . . e5 8.d5
ttJd4 9.ttJxd4 exd4 1 0.ttJb5 ( l O.ttJe2!?)
1 0 . . . El:e8 1 1 .El:e 1 ttJg4 (also quite possible is
1 l . .. a6 1 2.ttJxd4 ttJ xdS 1 3 .cxd5 .ixd4 1 4.'\Wc2
.id? 1 5 . .ie3 .ig7 1 6.El:ad 1 c5 1 7 .dxc6 Yz-Yz
Seirawan - Benjamin, Los Angeles 1 99 1 )
1 2.h3 a6 1 3.hxg4 axb5 1 4.cxb5 '\Wh4 1 5 .g3
'I.Wxg4 1 6.'\Wxg4 .ixg4 with good chances to
equalize.
h
9 . . . ttJ a6 (9 . . . ttJ g4!? is an interesting
alternative,
immediately
looking
for
concrete counterplay) 1 O.f3 El:b8 1 1 .a3 .id?
1 2.El:b 1 '.WaS 1 3 .�h 1 b5 1 4 . .id2 c4 1 5 . .ic2
'.We? 1 6 . .if4 El:fe8 1 7.ttJd4 ttJ c5 1 8 .ttJc6 El:b6
1 9 .a4 .ixc6 20.dxc6 b4 2 1 .tLlb5 'I.Wxc6 22.b3
ttJ hS 23 . .ie3 c3 24.ttJxa7 '.We? 25.ttJb5
'I.Wd8 26.'\WdS Yz-Yz Khalifman - Smirin,
Sverdlovsk 1 987.
8 ... exd5 9.exd5 ttJfd7 1 O.f4 f5 1 1 .0-0 El:e8
12 ..id2 ttJ a6 1 3 .El:e 1 ttJ f6 1 4.tLlg3 El:xe 1 t
1 5 ..ixe 1 .id?
With approximate equality - Black has
no bad pieces or real weaknesses; (0- 1 , 56)
Malakhov - Grischuk, Moscow 20 1 0.
6 . . . e5 7.d5 ttJhS!? is a different plan. As in
the Samisch System, Black prepares a quick
. . . f7-f5 . I played this once - 8 . 0-0 tLl d7 9 ..ic2
f5 1 0.exf5 gxf5 1 1 .tLl g3 tLlxg3 1 2.fxg3 e4
1 3 .g4 ttJeS 1 4.gxf5 .ixf5 1 5 .ttJxe4 tLlxc4, with
an unclear position; (0- 1 , 39) Lev - Smirin,
Haifa 1 993.
7.0-0 llJ d7
In this game I wanted to postpone . . . e7 -e5
for one move.
a
b
c
d
e
f
g
h
s .ie3
Or: 8 . .ic2 e5 9.d5 ttJ d4 1 0 . .ie3 ( l O.ttJxd4!?
exd4 1 1 .tLlb5 ttJe5 1 2.ttJxd4 ttJ xc4 1 3.El:b 1
happened in a few games, with good results
for White; among others, Chatalbashev
played it in 2009) 1 O . . . ttJxc2 I believe that
after exchanging this important bishop for the
knight, Black has no problems at all. 1 1 .'\Wxc2
aS 1 2.a3 b6 1 3.b4 f5 1 4.exf5 gxf5 1 5 .f3 .ia6
1 6.b5 .ic8 1 7.f4 ttJcS 1 8.El:ad 1 .id? 1 9.tLlg3
'\Wh4!? 20.ttJce2 El:ae8 2 l . fxe5 .ixe5 22 . .if4
.ig7 23.ttJd4 Yz-Yz Moiseenko - Smirin,
Maalot-Tarshiha 2008.
.
8 ...e5 9.d5 llJ d4 IO.Y;\'d2
Now 1 0.ttJxd4 is impossible and the knight
will remain on d4 for some time.
1 0.El:c l and 1 0.ttJb5!? are other popular
options in this position.
1 80
King's I ndian Warfare
1 o ... c5 1 1 .dxc6
Practically forced from a positional point of
view, as otherwise the knight on d4 would be
too strong.
1 1 . .. bxc6 1 2.b4 c5
Or if 1 6.lt:Jd5, then at once 1 6 .. .f4 and again I
would prefer Black - he has gained some space
on the kingside and the b6-bishop is cut off
from the rest of its forces and may become a
target.
16 ...�b7
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
13.b5?!
This decision to release the pawn tension
is too optimistic in my opinion. Yes, in an
endgame such a pawn structure would favour
White (he will potentially be able to create
a dangerous passer on the queenside) but in
the middlegame it's a strategically risky move.
Now the d4-knight occupies an excellent, well­
reinforced outpost. It will help a lot in Black's
upcoming activity against the white king.
1 3 .El:ab 1 , with a complex position, should be
preferred.
1 3 ... £5
Of course.
14.�g5 �f6 1 5.�h6 �H7 16.£4
True to his aggressive style, the Bulgarian
grandmaster Boris Chatalbashev tries to
play as actively as possible. If he were to
continue 1 6.exf5 gxf5 1 7. lt:J d5 instead, then
after 1 7 . . . lt:J b6!?+ the strong pawn centre
would promise Black excellent prospects.
Black's posltlon makes a nice harmonious
impression: his pieces interact very well.
Both fianchettoed bishops control important
diagonals, and we have already spoken about
the knight on d4. White should be careful
now.
17.fxe5?
The wrong choice, helping Black to activate
the dark-squared bishop.
The dangerous knight on d4 should have
been immediately exchanged: 1 7.lt:Jxd4 cxd4
1 8 .lt:Jd5 ixd5 (or 1 8 . . . lt:Jc5 1 9 . fxe5 ixd5
20.cxd5 ixe5 2 l .ig5) 1 9.cxd5 fxe4 20.ixe4
lt:J c5 (20 . . . exf4 2 l .�xf4 with an approximately
equal position) 2 l .fxe5 lt:Jxe4 (2 l . .. dxe5
22.�c2 El:c8? 23 .ixg6) 22.�xd4 ixe5
23 .�xe4 El:f6! A nice trick. 24.E!:xf6 (24.E!:ae l ?
g5-+) 2 4. . . �xf6 25 .E!:fl ixh2t 26.c±>xh2 �xfl
27.if4! The draw is inevitable.
17...Le5
181
Chapter 6 - Knight Agility
Black has a clear advantage now - his pieces
are quickly becoming very active.
22 ....ixd5 23.cxd5 h4 24..ixe5 tlJxeS 25 ..ie2
8
18.exf5 gxf5
Premature would be 1 8 . . . 1Wh4 1 9.if4
ttJxe2t 20.ixe2 id4t 2 1 .�h l �xf5 22 .ig3
'Mfg5 23.'Mfxg5 �xg5 24.�ac l lt:Je5 25.lt:Jd5,
with approximate equality.
7
6
5
4
19.tlJg3
Or 1 9 .�ae l 'Mfh4 20.if4 �h8 2 I .lt:Jd5
�g8, and all Black's forces will participate in
the upcoming attack on White's king. I doubt
White will withstand that.
19 ...�h4 20 ..ig5
3
2
1
a
b
c
d
e
f
g
h
25 ...�g7!
This wins the exchange with the help of a
little geometric trick.
8
7
26.tlJh5 tlJef3t!
Just another example that in a position with
a big advantage, tactics usually work in your
favour.
6
5
4
27..ixf3 tlJxf3t 28J�xf3 �xal t
3
2
8
a
b
c
d
e
f
g
h
20 ...�g4!
The queen enters the midst of events and
feels perfectly safe - the opponent is unable
to harm it. Meanwhile, Black's assault is now
speeding up considerably.
2I ..if4 hS!
After the inclusion of this pawn in the attack,
my advantage became decisive - White cannot
survive the storm.
22.tlJd5
The other options are no better, for instance
22.h3 'M!g6 etc.
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Not quite the famous 'Mfb2-h8t from
the game Petrosian - Spassky, World
Championship ( 1 O) 1 966, but it was a very
pleasant "long" move!
1 82
King's I ndian Warfare
Rustam Kasimdzhanov - Ilya Smirin
29.<i>f2 �d4t
The computer prefers 29 . . . �e8 30.lLlf4 �e4
3 1 .lLle6 f4, but my decision to swap queens is
easy to understand.
30.�xd4 cxd4
8
7
6
5
4
3
Elista Olympiad 1998
The Olympiad in Elista was memorable for
me. It was our first success after quite a few
tries - the first time that the Israeli team tied
for third place - and it happened before Boris
Gelfand joined our team in 1 999. Two other
Boris's - Avrukh and Alterman - plus myself
and Lev Psakhis made a big contribution to
this result. This game has also remained special
for me. I found an original manoeuvre and
the very strong player Rustam Kasimdzhanov
collapsed shortly after that.
l .d4 tL)f6 2.c4 g6 3.tL)c3 .ig7 4.e4 d6 s.tL)f3
0-0 6..ie2 e5 7.0-0 tL)c6 8.d5 tL)e? 9.tL)d2
a5
2
1
a
b
c
d
e
f
g
h
This endgame with an extra exchange is
technically won, of course.
3 1 .gd3 gcs 32.gxd4 gc2t 33.<i>f3 gfc7!
The last precise move in the game; now
White cannot avoid the exchange of rooks. The
rest does not require any further comment.
34.a4 <i>f7 35.tL)f4 g7c3t
36.tL)d3 gc4 37.gxc4 gxc4 38.a5 gd4
39.tL)f4 gb4 40.b6 axb6 41 .axb6 ci>f6 42.g3
gb3t 43. <i>g2 hxg3 44.hxg3 gxb6
0-1
9 . . . c5 became very popular after the
following famous game: 1 0.�b 1 lLl e8 1 l .b4
b6 1 2.a4 f5 1 3 .a5 lLlf6 1 4.'!Ma4 id? 1 5 .'1Ma3
ih6 1 6.id3 '1Mc7 1 7.bxc5 bxc5 1 8 .exf5 gxf5
1 9.ic2 a6 20.lLlde4 ixc l 2 1 .lLlxf6t �xf6
22.�fxc l �af8 23 .�b6 ic8 24.lLle2 f4 25 .ie4
lLl f5 26.Ek6 '!Mg7 27.�b 1
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
27 . . . lLlh4 28.'1Md3 ifS 29.@h 1 f3 30.lLlg3
fxg2t 3 l .@g 1 ixe4 32.'1Mxe4 lLlf3t 33.@xg2
lLld2 0-1 Larsen - Fischer, Denver (4) 1 97 1 .
I have played 9 . . . c5 a few times, but mostly I
have followed another great champion - Carry
Kasparov. He usually played 9 . . . a5 and made
huge contributions to this line.
1 83
Chap ter 6 - Knight Agility
IO.a3 .id?
1 0 . . . lt:Jd7 was played by Kasparov several
times. Perhaps the most memorable game was
a clash with his formidable "eternal" opponent
- the great Anatoly Karpov. In this game Garry
was close to winning, but Anatoly managed to
hold: 1 1 .El:b 1 fS 1 2.b4 <tt> h 8 1 3 . f3
can neutralize the knight. 36.lt:Je2! <tt> g7
37.�h 1 lt:Jd4 38.lt:Jbxd4 exd4 39.�d 1 �eS t
40.<tt> f3 �f6t 4 1 .<tt> g3 �eS t 42.<tt> f3 �f6t
Yz-Yz Karpov - Kasparov, Skelleftea 1 989.
I have also played this line a few times (once
against Kasparov himself in 1 988, and it was
quite a game - you will find it elsewhere in the
book) and another time against Gavrikov in
the same year. For a while I considered it one
of my best creative achievements, and you will
also find that game in this book. However, in
the game against Kasimdzhanov I had chosen
other continuation, introduced to practice by
the outstanding KID expert Efim Geller in
1 974.
l l .b3
a
b
c
d
e
f
g
h
1 3 . . . lt:J g8 ! This move became very popular; by
playing it Black solves his usual problem in
this line: how to coordinate his forces and to
make the e7-knight an active piece. 1 4. lt:J b3
axb4 1 S .axb4 tt:Jdf6 1 6 . .id2 lt:J hS 1 7.g3 lt:Jhf6
1 8 .El:f2 lt:Jh6 1 9 .El:a 1 El:xa 1 20.�xa 1 tt:J f7
2 1 .�c l f4 22.g4 h S 23.h3 lt:J h7 2 4. .ie 1 .if6
2S.<tt> g2 <tt> g7 26.El:fl lt:J hgS 27.El:h 1 lt:Jxh3
28.El:xh3 lt:JgS 29.El:h2 hxg4 30.fxg4 El:h8
31 ..ih4 f3t 32 . .ixf3 lt:Jxf3 33 . .ixf6t �xf6
34.El:xh8 <tt> xh8 3 S .<tt> g3
8
7
6
5
4
3
2
1
a
8
b
c
d
e
f
g
h
The most natural continuation, preventing
the blockading . . . aS-a4.
7
6
l l . .. llJcS
5
This was an original idea of Geller - Black
transfers this knight to b6, to be able to jump
to a4 if White starts activity on the queenside
by means of b3-b4.
4
3
2
a
b
c
d
e
f
g
h
35 . . . gS?! (3S . . . <tt> g7! gives Black a big advantage)
Now that . . . �h4 t is no longer possible, White
1 1 ... cS or 1 1 ... c6 are other popular lines here ­
it shows that Black has a lot of versatility in
this system.
1 84
King's Indian Warfare
12J�b l tlJb6 13.Y:Yc2
Mter the immediate 1 3 .b4 axb4 1 4.axb4
lt:J a4 1 5 .Wfc2 lt:Jxc3 1 6.Wfxc3 ih6 (Langeweg
- Geller, Amsterdam 1 974) Black has no
problems at all - he has good piece play and
no real weaknesses.
1 3 ...Y:Ye7
I played this move for the first time against
M. Gurevich (Elenite 1 994) .
14.b4
Or 1 4.ib2 ih6 1 5 .�be 1 ig4 1 6.ixg4
lt:Jxg4 1 7.a4 lt:J d7 1 8 . lt:J f3 lt:J c5 1 9.ia3 b6
20.ixc5 bxc5 2 l .g3 lt:J f6 22.lt:Jh4 Wfd7 23.mg2
ig5 with approximate equality; (Yz-Yz, 54)
M . Gurevich - Smirin, Haifa 1 99 5 - this was
our second battle in this particular variation.
( 1 8 .f3 lt:Jh5) 1 8 . . . c6 1 9.lt:Jf3 ig7 ( 1 9 . . . ixcl
20.�fxc l cxd5 2 l .cxd5 �fc8 22.lt:Jd2 Wfd8!?
was also perfectly playable with approximate
equality - Black easily can match the
opponent's activity on the queenside) 20.ig5
cxd5 2 1 .cxd5 h6 22.id2 (22 .ih4 g5 23 .ig3
lt:J h5f!) 22 . . . id7 23.�bc l Wfd8 24.Wfb 1
lt:Jh5 2 5.g3 ig4 26.mg2 Wfd7 27.lt:Jg 1 ixe2
28.lt:Jxe2 f5 Black was OK in the other game;
(Yz-Yz, 50) M. Gurevich - Smirin, Elenite
1 994.
16 ... tlJg4!
8
7
6
5
14 ... axb4 1 5.axb4 tlJa4
4
8
3
7
2
6
1
5
a
4
3
2
a
b
c
d
e
f
g
h
16.tlJdl
After some thought Rustam came up with
a fresh idea, keeping the knight from being
exchanged.
b
c
d
e
f
g
h
This move is the main reason why
included this game in the book. It looks rather
paradoxical at first sight - the knight j umps
into the air, and White may attack him with
the f- or h-pawns. But it transpires that both
these moves have drawbacks. Meanwhile I
want to proceed with . . . f7-f5 , when the knight
on d 1 is temporarily unable to move to e3.
I love this idea, even though objectively it
deserves a ! ? rather than ! .
17.£3
Useless is 1 6.lt:Jb5 c6! (this is the idea behind
1 3 . . . Wfe7 - now the d6-pawn is protected)
1 7. lt:Jc7? �a7.
Mter: 1 6.lt:Jxa4 ixa4 1 7.Wfc3 ih6 1 8 .Wfd3!
If 1 7.h3 I would continue 17 . . . lt:Jh6 and after
1 8.lt:Jb3 f5 1 9 .f3 lt:Jf7 Black would feel great ­
his initiative on the kingside is obvious, and the
h3-pawn may be an additional weakness in the
near future, after let's say . . . Wfh4 and . . . lt:Jg5 .
1 85
Chapter 6 - Knight Agility
17... tiJf6!
1 7 . . . tb h6 would be less to the point - the
pawn is still on h2 and after 1 8 . lLl b3 f5 1 9 .c5±
I would probably prefer White. Instead the
knight returns to f6, but the position has
been changed - because the f3-pawn limits
the bishop on e2, Black can play . . . tbh5
next. I used a similar idea in my game with
Zilberman, which you will find in this book,
but in that game this idea was well known,
while in the present game it was over-the­
board improvisation, which I value much
more highly.
the attack on the kingside, whereas Black's
queenside lies in ruins.
20.c5 bxc5 2 1 .bxc5
8
7
6
5
4
3
2
b
a
c
d
e
f
g
h
2 1 . . . �fb8! ?
2 1 . .. f5 22.c6 .ie8 23.�a l ±
2 1 . . . .if6!? 22.c6 .ih3 23.�e l .ig5
22.c6 .ih3
The consequence of 1 9 .g3 .
23.�el .if6! 24.tbe3 .ig5 2 5 .�al lb b6
And despite the strong white pawn on c6,
which divides the board into two parts, all
Black's forces are rather well coordinated and
he has a fully acceptable position.
8
7
6
5
4
3
2
19 ... tlJf4!
a
b
c
d
e
f
g
h
19.tlJa5?!
The future FIDE World Champion
underestimates Black's idea. He would have
been better off playing the prophylactic:
1 9.g3
When I would have to choose between the
defensive:
1 9 . . . b6
I think this is the best move in the position.
Or the more energetic: 19 .. .f5 20.lLla5 f4
2 1 .g4 lLl f6 22.lbxb7 h5 23.gxh5 (23.g5
'Llh7 24.h4 'Llxg5 25.hxg5 �xg5t 26.�h l
�h4t=; 23.h3 'Ll h7) 23 . . . 'Llxh5 But after
24.�f2! I cannot find a way to continue
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Both knights have reached active positions,
and it seems that the knight on f4 does a more
effective job than its white counterpart on a5.
1 86
King's Indian Warfare
20.�xb7??
A blunder which immediately decides the
outcome.
It was much better to eliminate the annoying
knight: 20.ixf4 exf4 2 l .ltJxb7 Now Black
totally dominates on the dark squares and
can choose between 2 l . .. EHb8 22.ltJa5 �e5!
23J�b3 �d4t 24.mh l �b6 2 5 . ltJ c6 (what
else?) 25 . . .ixc6 26.dxc6 ltJ c5 (26 . . . �xc6+)
27.El:b l ltJ e6 28.b5 El:a3 29 .�d2 El:ba8 with
better chances, or 2 l . .. �h4 22.�d2 El:ae8
23.�e l �h6, with rich play and a better
position for j ust one small pawn.
But the best decision in this situation was the
modest 20.El:f2, acknowledging that White's
1 9th move was a mistake, but hoping for a
stubborn defence in a worse position. This
would be an easy decision for a computer
program, but not for humans who usually do
not like to confess their mistakes.
a
b
c
d
e
f
g
h
20 ... �xe2t 21 .YNxe2 �Hb8 22.c5
Alas, 22 .ltJa5 is impossible due to 22 . . . El:xa5,
but the game is lost in any case.
22 ... dxc5 23.bxc5 �xc5 24.�xc5 �xbl
25.�xd7
My opponent resigned, not waiting for
25 . . . El:xc 1 .
0-1
Chapter 6 - Knight Agility
Tigran Kotanjian - Ilya Smirin
European Championship, Yerevan 2014
greatly enjoyed the atmosphere of the
European Championship in Yerevan. I also
came away satisfied with my own performance,
especially its creative aspect. A notable case was
the following game from the second round,
against the Armenian Grandmaster Tigran
Kotanjian.
l.d4 liJf6 2.c4 g6 3.liJc3 .ig7 4.e4 d6 5.ltJf3
0-0 6..ie2 ltJ bd7 7.0-0 e5
There were two mundane reasons why
I opted for developing the knight on d7,
which today is quite a rare variation: (a) I was
counting on the surprise effect, and (b) In my
view, the variation itself is not bad.
8.�c2 lDh5
I intended this sideways jump to confuse
my opponent somewhat, but he didn't bat an
eyelid and made his next two moves quickly.
9J�dl liJf4 lO ..ifl
8
7
6
5
4
3
2
1 87
l l .d5?!
I was glad to see this premature release of the
tension. After the tougher l l .ie3 Wle7 1 2.g3
lt:Jh5 (or 1 2 . . . lt:J e6 1 3.d5) 1 3 .d5 c5 1 4.a3 ,
a standard closed structure would arise in a
form highly favourable to White. In our post­
mortem analysis, Kotanjian described that
situation as a "catastrophe" for Black. That is
putting it strongly, but in broad terms it isn't
far from the truth.
l l . .. cxd5 12.cxd5
In the case of 1 2.lt:Jxd5 lt:Jxd5 1 3 .cxd5 lt:J f6
1 4.h3 id7, I would have nothing to complain
about.
12 ... ltJf6
Black's .. .f5 i s postponed for now, b u t his
pieces are coming into play - for instance, his
bishop is already poised to j ump out to g4 with
an awkward pin.
13.liJd2
This knight is headed for c4, with subsequent
activity on the queenside and against the pawn
on d6 - but that is still some way off. For the
moment the bishop on cl is shut in, and in
its turn it blocks the rook on a l . Taking an
unbiased look at the position as a whole, you
notice that all White's pieces, except for the
knight on c3 , are placed on the first two ranks.
Of course that doesn't mean anything yet,
but all the same . . . In this sense Black is more
active. However, his knight on f4 won't achieve
anything all on its own; it needs support.
13 ... ltJg4!
a
b
c
d
e
f
g
h
1 0 ... c6?!
This move deserves a question mark, but I
wasn't keen on the theoretical 1 0 . . . c5 , which
is stronger, but less aesthetic according to my
taste.
One knight is good, two are better. Moreover,
everything is now ready for . . .f5 .
14.h3
What's this - an oversight, underestimation
of the reply, provocation, or the result of
exact calculation? This last supposition was
1 88
King's Indian Warfare
practically out of the question; as to the other
three, during the game I reckoned it was a case
of underestimation.
A more cautious line was: 1 4 .lt:Jc4 f5 1 5 .f3
(or 1 5 . h3 ttJ h6) 1 5 . . . lt:J f6 (again 1 5 . . . ttJ h6
isn't bad) 1 6.exf5 (on 1 6.g3 fxe4 1 7.gxf4
exf3 1 8 .Wf2 e4 1 9. h3 ifS , we reach a
complicated position with Black holding the
initiative) 1 6 . . .ixf5 1 7.Wb3 At this point,
a piece sacrifice looks interesting: 17 . . . g5 ! ?
1 8.g3 g4! 1 9 .gxf4 gxf3 20.ttJe3 id? 2 1 .f5
ih6 And what is clear is that nothing is
clear; but over the board I would take Black's
side.
An attractive move. Th e queen has penetrated
to the very base of the white kingside. Its
main role at the moment is to stop the enemy
king from taking refuge in the calm haven
of h2.
8
7
6
5
4
3
2
8
7
a
6
17.tiJe2
c
d
e
f
g
h
This again is strictly White's only move. One
of the short variations I had calculated before
resolving on my 1 4th was 1 7.ttJf3 ttJ h5t
1 8 .�h4 if6t 1 9.ig5 ixg5t 20.ttJxg5 We3
2 1 .El:d3 Wf4 t 22.g4 h6, with an inevitable
quick mate.
5
4
3
2
1
a
b
c
d
e
f
g
h
14 . . tiJxf2!
.
The exclamation mark is an award for
boldness. I took about 1 5 minutes to perform
some preliminary calculations and reject
the good positional alternative 1 4 . . . lt:J h6
1 5 .lt:Jc4 f5 . The temptation to play inventively
outweighed all else, and the knight sacrificed
itself to draw the white king out of its shelter.
I S.<i>x£2 YNb6t I6.<i>g3
The only move. Not 1 6.�e l We3t 1 7.ttJe2
ttJd3t (a pleasing case of a boxed-in king) or
1 6.�f3 f5 with a crushing attack.
I6 YNgi !
••.
b
17 ... tlJh5t I S.<i>f3 YNh2
So for the moment the white king is stuck
in the danger zone. It isn't under direct attack
right now - but in just one move's time, after
the typical King's Indian break with . . . f5 (rarely
so effective as here!) virtually all Black's forces
will be joining in the attack.
19.<i>f2
The king could have moved away into
the centre under fire from the black pieces:
1 9.�e3 f5 (or 1 9 . . . ih6t 20.�d3 f5 2 1 .�c4!
- wow!) 20.ttJf3 Wh l 2 1 .ttJc3 fxe4 22.ttJxe4
ixh3 23.Wff2 And in the computer's opinion,
White has everything in order - but not many
of us mere mortals would have gone in for
something like this.
1 89
Chapter 6 - Knight Agility
19 ... £5 20.tiJf3 fxe4 21 .YMxe4 .ixh3
23 ... tlJxg3 24.hg3 .if5
Reinforcements have reached the queen in
time, but there are quite a few defenders round
the white king too.
8
7
8
6
7
5
6
4
5
3
4
2
3
1
a
2
b
c
d
e
f
g
h
25.YMa4?
a
b
c
d
e
f
g
h
22 ..if4!
I had missed this strong move in my
calculations, hoping for 22.@e l ? �h i with an
overwhelming position. I now had to extricate
myself somehow as I went along.
22 ...YMht
Fortunately the queen is not lost, and Black
retains quite good chances in the tactical fight.
23.tiJg3?!
A first step in the wrong direction.
Tigran and I had both seen the variation:
23.gxh3 ltJxf4 24.ltJxf4 (24.@g3 �ae8!)
24 . . . �xf4 25.�xf4! exf4 26.ig2 id4t!
27.ltJxd4 �h2 At this point my opponent
broke off his calculations. I saw j ust a little
further: 28.�d3! f3 (the only way to save the
queen; otherwise 29J�h 1) 29 Jhf3 �e5 30J�d l
�c8 3 I .@gl In the resulting situation, queen
and pawn are fighting against rook, bishop
and knight. White has a small material plus,
and his chances overall must be preferred; but
in my view Black shouldn't lose with precise
play - White's king is exposed and demands
protection from his pieces.
This is already a losing error. I think the
reason for it was that Kotanj ian overrated his
position. He didn't want to go in for 2 5 .�h4
�xh4 26.ixh4 (26.ltJxh4? ic2t 27.@ e l
ixd l 28.�xd l �ac8) 2 6 . . .e 4 27.ltJd4 ig4t
28.@el ixd l 29.�xd l ie5 , with approximate
equality in a complex ending - White's pieces
are active, but Black has a material edge.
Instead of this, White makes a move based on
a decisive oversight.
25 ...e4 26.hd6
My opponent evidently hadn't seen that
after 26.ie2 e3t! 27.@xe3 �xg2, Black would
win at once.
26 ... exf3 27 .gxf3
The only move. White loses immediately
in the event of 27.ixf8 �xf8 28.gxf3 �h2t
29 .ig2 ih3 30.�g l ie5 .
27... .ic2
An obvious-looking strike, but here I missed
27 . . .id7! 28 .�xd7 �xf3t 29.@e2 �h5!
30.�e6t @h8 3 I .@d2 �f6, winning in all
variations - as indicated by the computer.
28 ..ie2
1 90
Ki ng's Indian Warfare
32.d6!
8
A correct pawn sacrifice. Now it is the black
king's turn to feel uncomfortable.
7
6
32 ... hd6 33.�h3t <i>£8
5
If 33 . . . <j{h8 then 34.�xd6!, and Black has
to give perpetual checks with 34 . . . �h l t
(not 34 . . . �xd6 35 .�c3t <j{g8 36 . .ic4t <j{fs
37.�h8t) 3 5 .<j{f2 �h2t 36.<j{e J �g3t
37.<j{fl �h3t .
4
3
2
a
b
c
d
e
f
g
h
28 ...hdl?
And this i s j ust wrong. Th e correct line
was the prosaic 28 . . . �xd l 29 .-ixd l .ixa4
30 ..ixa4 .ixb2 3 l .�b l .id4t 32.<j{g2 �f6 and although the strong passed pawn on d5
and White's active bishop pair give Black some
difficulty in winning, his material advantage
is great; playing the position is already, in
essence, a mere matter of technique. However,
I was bent on continuing the attack. These
things happen . . .
29J�xdl �h3 3 0..txf8 �h2t 3 I .<i>fl .tx£8
There is material equality on the board,
but Black has an obvious initiative - and the
opposite-coloured bishops, as usual, help in
the attack. Not everything is so simple, though.
34.�e6
The bishop on d6 is under attack and White
also threatens �f6t, mating; in addition, the
prospect of his bishop coming out to c4 gives
cause for concern. Black solves these problems
with the aid of checks.
34 ...�hl t 35.<i>f2 �h4t 36.<i>g2
Not 36.<j{fl ? which loses
37.<j{g2 �e8 .
to 36 . . . .ic5
36 ...�g5t
8
7
6
5
4
8
3
7
2
6
5
a
4
37.�g4
3
2
1
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
Better was 37.<j{fl .ic5 38 . .ic4 �gl t
39.<j{e2 �f2t 4o.<j{d3 �xf3t (after 40 . . . �d8t
4 I ..id5 �d4t 42.<j{c2 �g7 43.�e l , White is
not at all worse) 4 J .<j{c2 �f5 t 42.�xf5 t gxf5
43.�d7, and the game heads rapidly towards
a draw.
Chapter 6
-
37 ...Yfe5!
It's useful to provoke White's next move.
38.£4 Yff6 39.Yfg5?
He shouldn't have given up a second pawn
like this. Mter 39.f5!? gxf5 40.�fl f4 4 l .�dl
White would have had excellent drawing
chances.
39 ...Yfxb2
It's always pleasant to pick up a pawn along
the way.
8
7
6
5
4
3
2
1
a
40.<ibfl
191
Knight Agility
b
c
d
e
f
g
h
4 I . .. <ibg7! 42.Yfd5
Not 40 .�h6t? <±>g8-+.
8
After 42.�xd6 �c l t 43.<±>f2 �xc4 44.�d7t
<±>g8-+ 45 .�f6? �e2t, Black gets in first with
the mate.
7
42 .. J�e3!
6
This dots all the "i"s and crosses all the "t"s.
5
43.Yfd4t
4
3
2
a
b
c
d
e
f
g
h
40 .. J�e8!
After this rook's entry into the game, my
optimism increased significantly.
4I .i.c4
White would lose with 4 1 .�h6t <±>g8
42 . .ic4t <±>h8 43.�xd6 �cl t (or 43 . . . �b4!)
44.<±>g2 �xc4. As it is, his threats are dangerous
but the situation is clarified by my next two
accurate moves.
This hastens the finale, but the game was no
longer to be saved. Mter 43.�xd6 �f3t, the
white king would be mated. There would be
the same result after 43.�f7t <±>h6 44.�xd6
�b 1 t 45.<±>g2 �e4t 46.<±>h2 �c2t 47.<±>g 1
�g3t 48.<±>fl �c l t .
43...Yfxd4 44J�xd4 gat
White resigned, as after 45.<±>e2 �xf4 there
is only Black left on the board.
0-1
1 92
King's Indian Warfare
Boris Alterman
-
Ilya Smirin
Israeli League 2004
I .d4 �f6 2.c4 g6 3.�c3 .ig7 4.e4 d6 5.8
0-0 6..ie3 a6 7.V!!d2 b6
1 3.�e2 1his is a good version ofthe "hedgehog"
for White - the weakness on d6 is significant.
1 3 . . . �e8 1 4 .CLJ c2 �f8 1 5 .�f4 ctJe5 1 6 .�g5
ctJ ed7 1 7. 'it> h 1 �c7 1 8 .�f4 �e7 1 9 .�h4 �d8
20.�f4 �e7 2 1 .�d2 CLJh5 22 .�e3 �f8 23.f4
CLJ g7 24.�fd 1 f6 25 .�h4; ( 1 -0, 39) Razuvaev
- Zagrebelny, Moscow 1 983.
9 ..id3
Of course, in the event of the active 9 .e5
CLJ e8 1 0.f4 I would play 10 ... c5, undermining
the strong white pawn centre with good
counterplay.
9 ... c5 IO.d5
Now after 1 O.Ct:J ge2 Black can develop the
b8-knight to the much more active square c6 1 o . . . Ct:Jc6 with the idea 1 1 .0-0?! Ct:Jg4!.
a
b
c
d
e
f
g
h
8
This time I wanted to deviate from my usual
Samisch treatment and to play something new
and unexpected.
7
s.gdi
5
Alterman quickly and confidently made this
move, which is considered to be the best.
8 . CLJ ge2 c5 9.d5 e6 would lead to Benoni
positions. In the following game Black
managed to solve his problems successfully:
1 0.a4 exd5 1 l .cxd5 Ct:J bd7 1 2.CLJg3 CLJ e5
1 3.�e2 h5 1 4.0-0 h4 1 5 .CLJ h 1 Ct:J h7 1 6. CLJ f2
f5 1 7.exf5 gxf5 1 8 . CLJ h3 CLJ g6 1 9.'it>h 1 �a7
20.�ae 1 �e7 2 1 . f4 �f6 22.�c4 �fe8 23.�f2
�xe 1 24.�xe 1 �xe 1 t 25 .�xe 1 �e7 26.�d 1
�xc3 Yz-Yz Dydyshko - Vokac, Ostrava 200 5 .
8 ....id7!?
I like this original move more than the natural
8 . . . Ct:J bd7. Here is a game in which Grandmaster
Yuri Razuvaev played convincingly in
the opening and middlegame: 9 .�d3 c5
1 0 . CLJ ge2 cxd4 1 1 .CLJ xd4 �b7 1 2.0-0 e6
6
4
3
2
1
a
b
c
d
e
f
g
h
10 b5!
...
Making use of the absence of the white
rook from its initial position, Black carries out
this important move, striving for queenside
activity.
l l .b3
Of course not 1 l .cxb5?! axb5 1 2.�xb5 �xb5
1 3 .CLJxb5 �xa2.
1 1 .�h6 was played against me by the
Romanian Grandmaster Mircea Parligras,
1 93
Chapter 6 - Knight Agility
but it did not bring him an opening success:
l l . .. ixh6! 1 2.'1Wxh6 bxc4 1 3 .ixc4 ib5
1 4.ib3 '1Wb6 1 5 . lDh3 ttJ bd7 Black was fi n e:
he has easy and pleasant play against the b2pawn along the b-file, while the white king
is stuck in centre for the moment. 1 6. ttJ f2
lDe5 1 7.�d2 a5 1 8 .ttJxb5 '1Wxb5 1 9 .id l ttJ c4
20.ie2 '1Wb4 2 I .ixc4 '1Wxc4 22.a3 �ab8+;
(V2-V2, 44) Parligras - Smirin, Skopje 20 1 4.
1 1 . .. b4
Grabbing some space on the queenside and
driving away the knight on c3 .
After, let's say, 1 5 .lD g3 exd5 1 6.cxd5 a4
1 7.ttJ 1 e2 ic8! ? 1 8.0-0 ttJ bd7 Black would
have a good position - the knight on g3 is
slightly misplaced in my opinion .
15 ...fxe6 16.i.b1
8
7
6
5
4
12.lbce2 aS
3
2
8
7
a
6
b
c
d
e
f
g
h
16 .. J�� a6!
5
Here we see another benefit of 1 2 . . . a5 (besides
activity on the queenside) - now it's possible to
defend the d6-pawn in this unusual way.
4
3
2
1 7.£4 e5
a
b
c
d
e
f
g
h
Now my opponent faces a problem - how to
complete the development of his pieces?
13 ..ih6
1 3 .lDg3?, to vacate e2 for the other knight,
was impossible due to the cheap trick
13 . . . lDxd5 1 4.cxd5 ic3 . That's why Boris
Alterman decides to exchange the dark-squared
bishops, but it leads to some weakening of the
dark squares, as is often the case in the Samisch
System.
13 ... e6 1 4.i.xg7 Wxg7 1 5.dxe6
I think this decision to change the pawn
structure is correct.
The main purpose of this move is to fix the
white pawn on e4, thus making the bishop on
b 1 rather passive. Also Black fights to dominate
on the dark squares.
1 8.lb£3 �e7 1 9.0-0 exf4 20.l!Jxf4 i.g4
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 94
King's I ndian Warfare
I decided to let my opponent activate his
e2-knight, but in exchange Black is now free to
develop his knight from b8 to d7 and e5.
24.tlJe3?!
White could have played the nice: 24.�f5!
.ixf5 25.exf5 �e8 (the only move)
2 1 .tiJg5 tiJbd7 22J�del
Or 22.h3 .ixd 1 23.ltJge6t <±>h8 24.tt:Jxf8
'1Wxf8 25 .'1Wxd 1 '1We7 and Black is definitely
better because of the poor bishop on b 1 .
22 ... tlJe5 23.tiJ d5
8
7
a
6
5
4
3
2
1
a
b
d
c
e
f
g
h
23 ...�d8
With hindsight, more accurate would have
been: 23 . . . tt:Jxd5 24Jhf8 (24.exd5 �aa8 and
Black is OK at least) 24 . . . tt:J f6!? (24 . . . '1Wxf8
25.exd5 �a8 is good as well) 25 .�b8 h6 26.h3
'1Wa7 (26 . . ..ih5 27.�fl !) 27.�b5 .id?
b
c
d
e
f
g
h
26.ltJe6t (very interesting and perhaps stronger
would be 26. tt:J f4!? <±>g8 27.�fl gxf5 28 . .ixf5
with annoying pressure against the black
king - the white knights are very dangerous
and even more importantly the light-squared
bishop is now active) 26 . . . �xe6 27.fxe6 ltJxd5
28.cxd5 '1We7 29.�fl �a8 And chances are
about even - the e5-knight is still stronger
than the white bishop, but the protected pawn
on e6 is definitely a force Black has to reckon
with.
24 ... i.c8 25.tiJf3 i.e6?!
Having consolidated the position, I commit
another inaccuracy.
After the correct 25 . . . tt:J fd7! 26.ltJxe5 ltJxe5
27.�xf8 '1Wxf8 28.�fl '1Wd8 29.ltJd5 .ie6
30.ltJf6 h6 (prophylactic against 3 1 .'1Wg5)
Black would have nothing to worry about his king is well covered, the knight on e5 is
pleasant to see, the b 1 -bishop has no bright
future. I would definitely prefer to play as
Black in this situation.
26.tlJxe5 dxe5
a
b
c
d
e
f
g
h
28.ltJf3 .ixb5 29.cxb5 �b6 30.ltJxe5 dxe5
3 1 ..id3 with roughly equal chances.
1 95
Chapter 6 - Knigh t Agility
looks for two reasons: Black has an active plan
with . . . a5-a4 at the appropriate moment and
(again) the black bishop is j ust better than its
white counterpart. In my opinion White has
enough resources to draw, but it seems that
Boris Alterman underestimated the potential
danger.
29 ..ic2 ga7 30.gfl �d6
The knight continues its route towards d4.
a
b
c
d
e
f
g
h
27JWxd8?
My opponent had to continue more
aggressively with 27.ctJd5!, when this knight is
annoying. If Black were to exchange it right
away with 27 . . .i.xd5?! 28.exd5 he would end
up in the worse position - the e5-pawn is
weak, the white bishop has improved its status
a lot, and d5 is a protected passer.
27.. J�xd8 28.� d5
Without queens, this j ump has much less
effect - White does not have enough pieces
to support the knight's activity and I can j ust
ignore it.
3 1 .�b6 �c8 32.�d5
Probably 32.ctJa4 �c7 33.ctJb2 should
have been preferred. Then after 33 . . . ttJ b6
(or 33 . . . �cd7 34 .ttJa4 �c7 3 5 . ctJ b2) 34.ctJ d3
ctJxc4!? 35.bxc4 i.xc4 36. ctJ c 1 i.b5 I would
have enough compensation for the piece, but
hardly more, for instance: 37.ctJ b3 a4 38.ctJd2
c4 39.h3 �b8 40.�b l b3 4 1 .axb3 axb3
42.ctJxb3 cxb3 43.�xb3
32 ... �e7 33.h3
Perhaps White should have preferred to
defend passively: 33.ctJxe7 �xe7 34.�d l �xd l t
35 .i.xd l �d7 36.i.c2 a4 37.<±>fl axb3 3 8.axb3
�a7 39 .�d2 �al t 40.�d l �a2 4 1 .�d2 mf6
42. <±>e 1 And in my opinion he should hold
this fortress.
33 ... �c6 34 ..idl �d4
a
b
c
d
e
f
g
h
28 ... �e8
Starting the plan of bringing this knight to
d4. This endgame is worse for White than it
a
b
c
d
e
f
g
h
1 96
King's I ndian Warfare
The knight has arrived.
35.�e3 gf7
I decided to exchange a pair of rooks to
prevent any possible counterplay.
White still holds the last line of defence.
My next move aims to clear the space for the
black pieces to be able to penetrate through
the kingside.
49 ... h4!
36.gen gxfl 37 .gxn .id?
The next step is breaking through on the
queenside after the inevitable . . . a5-a4.
38.�d5 gas 39.i>fl ga7
8
7
6
The final preparation.
5
Of course, not the immediate 39 . . . a4? 40.tt:J b6.
4
40.®el a4
3
2
8
7
a
b
c
d
e
f
g
h
so.®hu
6
This seems to be a decisive mistake Alterman hopes to save the game by defending
passively.
5
4
3
2
1
a
b
c
d
e
f
g
h
4 1 .bxa4
My opponent chooses to spoil his pawn
structure (now the pawns on a2 and c4 become
weak) but not to allow me to exchange pawns
on b3 and penetrate to al with my rook. It's
hard to say which is the lesser evil.
4 1 . ...ixa4 42.La4 gxa4 43.�b6 ga7 44.h4
h5 45.�d5 g5!
He should have taken a more active approach
instead:
50.ctJb6! El:c3
What else?
50 . . . h3 5 l .ctJd7
Or 50 . . . El:a7 5 1 .\t>h3 <±>h5 52.\t>h2 <±>g4
53.ctJd5.
5 l .ctJd7 El:xc4 52.ctJxe5 E!:cl 5 3 .tt:Jf7t \t>h5
54.ctJe5 c4 5 5 .E!:d2 b3 56.axb3 cxb3 57.El:xd4
b2 58 .E!:b4
Another way is 58 .E!:d5!? b 1 =Wf 59.ctJd3t
<±>g4 60.ctJxc l Wfxc l 6 1 .El:d3= with a
well-known theoretical fortress.
58 . . . b l =Wf 5 9.El:xb l E!:xb l 60. ctJ f3 E!:b4 6 1 .g3=
Now it's time to activate the king.
46.hxg5 ®g6 47.®fl ®xg5 48.®gl ga3
49.i>h2
50 ... h3 5 1 .®h2 hxg2 52.®xg2 �e6
White is doomed - he has to defend too
many pawns and squares.
1 97
Chapter 6 - Knight Agility
53JU5t ®h4 54J!f2
If54Jhe5 'Dd4! White would be completely
lost despite the small amount of material
remaining.
59 ... tLld3?
The natural 59 . . . <kt>f3! would have won
immediately.
60.®g2?
54 ... tLlg5 55J!e2
Again Boris missed the chance to complicate
my task: 60 .'Dc7 with the idea 60 . . . <kt>f3 ?
6 1 .'D b5 . Instead correct i s 6 0 . . . 'D f4 6 1 .'D b5
�h3!? 62.<kt>f2 'D e6 63 .'Dd6 'D g5 and Black
should still win this.
8
7
6
60 ... tLlcl
5
The rest is simple; Black's pieces came
through.
4
3
6I.tlJf6t ®f4 62.tLld7 tLld3 63.tLlf6 ®e3
64.gc2 tLlel t
2
a
b
c
d
e
f
g
h
55 ... tLl h3! 56J!d2 ®g5 57.gc2 ®g4 58.gd2
8
7
Slightly more stubborn was 58.'Df6t, even
though it would not change much.
6
58 ... tLlf4t 59.®fl
4
5
3
8
2
7
6
a
5
0-1
4
3
2
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
1 99
Chapter 6 - Knigh t Agility
Yaacov Zilberman Ilya Smirin
-
Tel Aviv 1 999
I played quite a few KID encounters with the
Israeli Grandmaster Yaacov Zilberman. In
almost all of them the same position arose after
1 1 moves. This game is the most interesting in
my opinion, but I will briefly mention some
other games between Yaacov and myself.
l .d4 �f6 2.�f3 g6 3.c4 .ig7 4.�c3 0-0 5.e4
d6 6..ie2 �a6 7.0-0 e5 8.d5 �c5 9.'\Wc2
Or 9 . ltJ d2 ih6.
axb4 1 9.axb4 �g7 20 .'t!Md3 c5! 2 1 .dxc6
bxc6 22.El:fd 1 c5 23.ltJb3 cxb4 24.ltJb5 ltJ e8
2 5 .ltJxd6 ia4 26.c5 lDxc5 27.ltJxc5 ixd 1
28.El:xd 1 El:d8 29.ltJf5 t gxf5 30 .'t!Mxd8 't!Mxc5t
3 1 . �fl ttJ f6 32. '!Md2 f4
8
7
6
5
4
3
2
9 ... a5
a
8
b
c
d
e
f
g
h
33.g3 El:g8 34.El:c 1 't!Mb6 3 5 .El:b 1 El:b8 36.gxf4
�h7 37.ic4 't!Mc7 38 .id3 ltJ d7 39.El:cl 't!Md6
40.El:d 1 El:g8 4 I .'t!Mf2 b3 42.ie2 't!Mc6 43 .'t!Ma7
b2 44.'t!Ma2 't!Mg6 4 5 .�e 1 El:b8 46.�d2 exf4
47.�c2 ltJe5 48.�b 1 't!Mc6 49.'t!Ma3 't!Mc7
50.'t!Md6 't!Ma7 5 1 .ia6 El:e8 52.'t!Mf6 El:e6 53 .'t!Mf5t
�g7 54.'t!Mg5 t ltJg6 0- 1 Zilberman - Smirin,
Israel 20 1 1 .
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
IO ..ig5
1 O.ltJd2 ih6! would lead to excellent play
for Black, as has been known since the game
Petrosian - Geller, Moscow 1 949.
10 ...h6 l l ..ie3 b6
This secures the knight's outpost on c5.
1 1 . . . ttJ g4 would lead to a different type of
position. Once it was played by the great
Robert James Fischer: 1 2.ixc5 dxc5 1 3 .h3
ltJ f6 1 4.ltJxe5
8
7
6
5
4
3
1 1 . . . 't!Me7!? is another option here. Although
much less popular, this move is probably j ust
as good as 1 1 . . . b6. I played it once against well, you guessed right: 1 2.ltJd2 id7 1 3 .b3 h5
14.f3 �h7!? A different plan - Black goes for
the exchange of dark-squared bishops. 1 5 .a3
ih6 1 6.ixh6 �xh6 1 7 .b4 ltJ a6 1 8 .El:ab 1
2
a
b
c
d
e
f
g
h
1 4 . . . ttJxd5 1 5 .cxd5 ixe5 1 6.f4 id4t 1 7.� h 1
't!Mh4; (0- 1 , 6 5 ) Gligoric - Fischer, Siegen (ol)
1 970.
200
King's I ndian Warfare
I think it's purely a matter of taste which line
to choose.
1 2.tiJd2
8
7
6
5
a
4
3
2
1
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
1 6.exf5 (the sample variation 1 6.f3 lD g5 1 7.b4
f4 1 8.if2 [better is 1 8 .ixc5 bxc5 1 9.bxc5 , but
Black is fine, of course] 1 8 . . . lDxh3t 1 9.gxh3
ixh3 20.�h2 ixfl 2 1 .8:xfl axb4 22.axb4
lD d7 shows clearly the drawback of the pawn
being on h3) 1 6 . . . gxf5 1 7.f4 exf4 1 8 .ixf4
Wh4 1 9.8:ae 1 lD g5 20.�h2 E:ae8 2 1 .Wc1
1 2 ...1g4
Another encounter with Zilberman saw the
"simpler" (and weaker, I should add) 1 2 . . . id7
1 3 .b3 lD h7 1 4.a3 f5 1 5 .f3 f4 1 6.if2 g5 1 7. b4
and I was much worse, as White is ahead with
his attack; Yz-Yz Zilberman - Smirin, Israel
2007.
1 3.h3
For 1 3 . f3 see the game Naumkin - Smirin
on page 1 0 1 .
a
b
c
d
e
f
g
h
We will see later in the game that on h3 the
pawn in general stands worse than on its initial
position, and this factor will help Black to
develop an initiative on the kingside.
2 1 . . . ie5! 22.ixe5 E:xe5 23.b4 axb4 24.axb4
f4! 2 5 . lD f3 lDxf3t 26.8:xf3 lD b3 27.Wd l
lDd4! 28.8:d3 tDxe2 29.8:xe2 ig4 30.Wc2
ixe2 3 1 . tDxe2 Ei:e3 0- 1 Zilberman - Smirin,
Panormo 200 1 .
14.b3 tLlh7 15J�ael
1 5 ...�h4!?
1 3...1d7!?
In our next "thematic opening duel" Yaacov
preferred a more traditional approach, but
without success: 1 5 .a3 f5
Trying to make use of the advanced position
of the h-pawn.
1 5 .. .f5 1 6.exf5 gxf5 (on the other capture
- 1 6 . . .ixf5 1 7.lDde4 - White controls the
e4-square and remains OK) 1 7 .f4 would lead
to a rather standard double-edged position.
20 1
Chapter 6 - Knight Agility
16.�8
My opponent did not like 1 6.a3?! f5 1 7.exf5
(or 1 7.b4 lLl a6 1 8 . lLl f3 Wffe7) 1 7 . . . i.xf5 , when
Black's pieces are very active.
16.g3 Wffxh3 is impossible, of course.
16...YMe7
b
a
c
d
e
f
g
h
24 .. .f3!; (Yz-Yz, 48) Vaganian - Smirin,
Rostov-on-Don 1 993. Black has obtained a
good and active position.
I7...VNh4 18.�f3 YMe7 19.�d2
White agrees to a draw.
19 �g5!?
..•
a
b
c
d
e
f
g
h
Similar manoeuvres . . . i.c8-g4-d7 and
. .. Wffd 8-h4-e7 (an interesting coincidence)
have improved Black's position a bit, and it
looks quite attractive to me - the black pieces
are well placed and the traditional . . f7-f5 is in
the air.
.
17.�d2
Rafael Vaganian
preferred
1 7.Wffcl!?,
attacking the pawn on h6, in a game
against me. I replied: 1 7 .. . f5 ! 1 8 .exf5 (bad is
1 8 .i.xh6? fxe4 1 9 .i.xg7 �xg7 20.lLld2 lLl g5)
1 8 ... gxf5 1 9 .lLlb5 �ac8 20. lLl h2 (or 20.i.xh6?!
f4 2l.i.xg7 Wffxg7 with an evident initiative for
the pawn) 20 . . . f4 2l.i.xc5 bxc5 22.i.g4 i.xg4
23.hxg4 Wffh4 24.lLlc3 (24.f3 lLl g5 25 . lLl c3 e4!
26.lLlxe4 i.d4t 27.�hl lLl xe4 28.�xe4 i.f2
29.�xf2 Wffx f2 30.Wib l �ce8 3l.�xe8 �xe8
32.Wffg6t �f8 33.Wif6t may have led to a draw
by perpetual)
a
b
c
d
e
f
g
h
But not me! Black's pieces have started to
concentrate around the opponent's king;
. . f7 -f5 is coming and White is on a defensive
footing.
.
20.f3
A kind of prophylactic move.
After the careless 20.a3? lLl xh3t 2 1 .gxh3 i.xh3
Black would get a huge advantage.
202
King's Indian Warfare
20 5
Now the sacrifice on h3 does not work
because of 22.�f2.
stronger than my opponent's queen - the black
pieces work excellently together, and White's
king is under strong pressure.
21 .exf5 gxf5
24... Axb5
The capture with the pawn is much more
interesting and aggressive than 2 1 . .. i.xf5?!
22.ltJ de4 with an equal and dull position
- some exchanges on e4 would soon
follow.
It was a pity to part with one of the
better bishops, but I had great hopes for
the remaining one - the dark squares are
starting to feel vulnerable in White's camp.
By the way, the weak square on g3 was
created because Black provoked the move
h2-h3 earlier. In some variations (as we will see
later in the game, for example) this weakness
might tell.
•••
22.f4
The standard method to stop the black pawn
on f5 physically.
22.�h l !? was, perhaps, more cautious. In
this case a logical continuation would be
22 . . . e4 (I do not like 22 .. .f4?! 23 .ixc5 bxc5
24.ttJde4) 23.f4 ltJh7 with a complex position,
but optically my sympathies lie with Black
here.
25.cxb5 Ae5
8
7
6
5
4
22 ... exf4 23 ..ix£4 gae8!
8
3
7
2
6
1
5
b
d
f
g
h
4
26.�c4?
2
Apparently after this move White's game
is already beyond salvation - Black's pieces
become too dominant.
a
3
a
b
c
d
e
f
g
h
Probably Zilberman underestimated this
strong move. Now the "semi-pin" on the e-file
is rather unpleasant for him.
24)t)b5
If 24.ih5 "!Wxe l 25 .�xe l �xe l t 26.�h2
ttJ ce4 my two rooks would be considerably
c
e
26.ih5 i.d4t 27.�h2 "!Wxe l 28.�xe l �xel
would again be in Black's favour.
The only correct decision would be 26.i.d3!,
creating some counterplay by eyeing the pawn
on f5 .
203
Chapter 6 - Knight Agility
e2-bishop, and this is the major factor which
determines Black's overwhelming advantage.
8
7
6
29.%Vcl
5
8
4
7
3
6
2
a
b
c
d
e
f
g
h
After this I failed to find an advantage for
Black in the following variations:
26 . . . ltJ ge4?! 27.ixh6 is unsound.
26 . . . id4t 27. �h21t's interesting that now the
pawn's position on h3 hugely favours White
- in chess, as in life, the same things under
different circumstances may have completely
different meanings. 27 . . . '\Wxe l 28 .:gxe l :gxe l
29.1xf5 ig l t 30.�g3 :ge2 3 1 .'\Wd l if2t
32.�g4 :ge7 The computer insists strongly on
equality - I will not argue with that.
26 . . . ltJxh3t 27.gxh3 'W'g7t 28.�hl ixf4
29.:gxe8 :gxe8 30.:gxf4 'W'g3 3 l .:gf3 :ge l t
32.ifl :gxfl t ! 33.:gxfl 'W'xh3t 34.�gl 'W'g3t
leads to a draw by perpetual.
26 . . .'W'g7 27.ixe5 :gxe5 28.:gxe5 'W'xe5
29.ixf5 'W'xd5 30.ig4 with an approximately
level position.
5
4
3
2
b
a
d
c
f
e
g
h
29 ... ttlg3!
See the note after Black's 24th move.
30JUl?!
Zilberman misses his last chance. He had to
find a much more stubborn idea:
30.:gf3! 1xf4 3 l .'W'xf4 lDxe2
After 3 1 . .. ltJh5 32.'\Wh4 '\WeSt 33.�gl f4
34.:gb l lD g3 3 5 .'\Wg4t �h8 36.id3 :ge7
Black remains clearly better, but White can
still resist.
32.:gxe2 :gxe2 33 .:gg3
And I would have to find the precise path:
26 ...%Vg7!
The ideal place for the queen - it creates the
threat of . . . ltJ h3t and controls a few important
dark squares - e5, g5 and g3 . At the same time
the e8-rook becomes very active along the
e-file.
27.®h2 ttlge4 28.ttlxe4 ttlxe4!
This is much stronger than 28 . . . fxe4. Now
the knight on e4 is clearly superior to the
a
b
c
d
e
f
g
h
204
King's I ndian Warfare
33 . . . V9xg3t!
On the tempting 33 ... :gf7 34.a3! Black
may easily fall into the trap: 34 . . . V9xg3t
3 5 .V9xg3t :gg7 36.V9f3 :ggxg2t 37.V9xg2t
:gxg2t 38.�xg2 �f7 39.�f3 �f6 40.�f4
�g6 4 1 .h4 �h5 It seems the pawn ending
is won, but . . . 42.�xf5 �xh4 43.�e6 h5
44.�d7 �g3 45.�xc7 h4 46.�xb6 h3
47.�a7 h2 48 . b6 h l =V9 49. b7 leads to a
draw - Black cannot prevent the b-pawn
from queening.
34.V9xg3t �h7 3 5 .V9c3
Here 3 5 .a3 :gg8 36.V9f3 :ggxg2t 37.V9xg2
:gxg2t 38 .�xg2 �g6 39.�g3 �g5 40.h4t
�h5 does not save White - Black has an
extra tempo compared with the previous
line.
3 5 .V9f3 :gxa2 is hopeless, of course.
35 . . . :gg8 36.V9xc7t :gg7 37.V9xb6 :ggxg2t
38.�h l :gc2 39.V9xd6 :gge2
a
b
c
d
e
f
g
h
This manoeuvre gives the black queen access
to e5, with lethal consequences.
32.YMh4 YMe5t 33.<i>gl
All White has left is choosing a way to lose.
After 33.g3 f4! 34.gxf4 ltJxf4 3 5 .:gg l t �h7 he
would have to resign.
8
7
6
5
a
b
c
d
e
f
g
h
40.V9d7t �g6 4 l .V9d6t
4 1 .V9c6t �h5 42.V9xc2 :gxc2 43.b6 �h4
44.b7 �xh3 4 5 .�gl :gg2t 46.�fl :gg8 etc.
4 l . . . �h5
And it's winning in the end. But this would
certainly not be easy to find during the game!
30 ...hf4 3 1 .YMxf4 ttlh5!
4
3
2
a
b
c
d
e
f
g
h
'='
34... �f7!
The last detail; now the bishop is in a deadly
pin and will be lost.
3s.ga gg7 36.YMd2
36.:gf2 V9c3 (not
continuation) 37.:gc l
the
only
winning
205
Chapter 6 - Knight Agility
Evgeny Bareev - Ilya Smirin
Moscow (rapid) 2002
b
a
c
d
e
f
g
h
37 . . . �xe2! 38.�xc3 would lead to checkmate:
38 . . . �e1 t 39 .�h2 �h 1 #
36 ... lbxe2t
This game was played in the Moscow Grand
Prix event in rapid chess with a time control,
if I am not mistaken, of 25 minutes plus 1 0
additional seconds for each move. Th e game
is not free from mistakes, to put it mildly. The
main reason I included it here is the bizarre
and highly unusual final position. Besides, this
encounter stays in my memory as a win over
one of the best players in the world at the time
- Evgeny Bareev.
l .d4 lbf6 2.c4 g6 3.lbc3 .ig7 4.e4 d6 5 ..ie2
0-0 6.lbf3 e5 7.0-0 lb c6 8.d5 lbe7 9.b4 a5
Not 9 . . . tt:J h5 this time. I played the second­
most-popular move, immediately challenging
White on the queenside.
IO ..ia3
8
7
6
a
b
c
d
e
f
g
h
White resigned because of 37.�f2 Wh2.
0-1
The final game of this chapter is as complicated
and chaotic as a novel by Tolstoy. First my
knights terrorize my opponent. Then, in deep
complications, the white king goes running
and the white knights keep him safe, until he
goes astray.
5
4
3
2
a
b
c
d
e
f
g
h
10 ... b6
I have quite often played this move, in
connection with the next one, but in my
opinion this line has mostly experimental
value.
The main line is 1 O . . . axb4 1 I .ixb4 tt:Jd7 1 2.a4.
White, as usual, prepares to push c4-c5, while
206
King's I ndian Warfare
Black usually connects his kingside play with
. . . f7 -f5 . I do not particularly like the scenario
here, although I played this once against San
Segundo in 1 997 (in Gran Canaria, drawn
after 30 moves) .
1 1 .bxa5 � h5!?
Again, more common is 1 1 . . J �xa5 1 2.ib4
�a8 1 3 .a4, but as I already said, that's not to
my liking.
a
12J�e1
Not good is 1 2.axb6 �xa3. This is the tactical
idea behind 1 0 . . . b6. 1 2.lt:Jb5 �xa5 1 3 .ib4
�a6 1 4.lt:Jd2 occurred twice in my games with
Van Wely (in 1 998 and 20 1 1 ) .
1 2 ... £5
b
c
d
e
f
g
h
1 9.h4! @g? 20.exf5 lt:Jxf5 2 l .g3 �f6? 22.ctJe4
ctJe3 23.ctJxf6 ctJxd l 24.�xd 1 �xf6 25 .ctJg5
ixg5 26.hxg5 �f5 27.cxd6 cxd6 28.ixd6 �xg5
29.�c2 �c8 30.�xc8 ixc8 3 1 .�c l ig4 32.�c7t
@g8 33.ie7 �h5 34.d6 @f7 35.f3 ie6 36.g4
1-0 Eljanov - Smirin, Porto Carras 20 1 1 .
I think this game clearly shows the troubles
Black may experience in this line - White was
very quick to achieve a lot on the queenside
and in the centre, while Black had no time to
organize any serious counterplay.
13 ... �£6 14.J.b4
After 1 4.lt:J b3?! I won the following
interesting game against the very strong and
creative player Anton Korobov:
a
b
c
d
e
f
g
h
1 3.�d2
I have faced a lot of trouble after 1 3 .ib4
bxa5 1 4.ia3 . I drew my game against Kramnik
in 1 999 (you will find this game on page 59)
and beat Delchev, but lost convincingly to
Huzman and Eljanov. Here is the last of these
games: 1 4 . . . lt:J f4 1 5 .c5 lt:J xe2t 1 6.�xe2 ia6
1 7.�b2 ih6 1 8 .�ab 1 �f7
a
b
c
d
e
f
g
h
1 4 . . . fxe4 1 5 .ib4 bxa5 1 6.ixa5 ctJ f5 1 7.lt:Jb5
e3 1 8 .fxe3 ih6! 1 9 .id2 ixe3t 20.ixe3 lt:Jxe3
2 l .�d2 ctJ f5 22.a4 id? 23 .id3 �b8 24.@h l
lt:J g4 25 .�fl
207
Chapter 6 - Knight Agility
IS.dxc6 liJxc6 16 ..ia3
8
7
6
5
4
a
b
c
d
e
f
g
h
25 . . . Wd8
25 . . . � fe3! 26.�f3 e4 27.ixe4 �xc4+
26.i.xf5 gxf5 27.h3 � f6 28 .Wg5 t �h8
29.Wh4 f4 30.� d2 ifS 3 1 .a5 Wd7 32.�c3
Wg7 33.a6 �g8 34.�g1 ig6 3 5 .�h2 �h5
36.� de4 ixe4 37.�xe4 Wg6
3
2
a
b
c
d
e
f
g
h
16 ... bxa5?
This was a rapid game and I made this move
without much thought. It is a rather serious
mistake.
Correct was 1 6 . . . � d4!? or the quieter
1 6 . . . �xa5 1 7.exf5 gxf5 1 8.�b5 � b7 1 9. ctJ b3
i.e6 and Black is not doing badly at all.
l?.liJbS
a
b
c
d
e
f
g
h
38.We 1
38.a7!
38 . . . �gb8 39 .We2 �b3 40.a7 �b7 4 1 .�gb 1
�xb 1 42.�xb 1 �xa7 43 .Wg4 �g7! 44.�b8
Wxg4 45.hxg4 lt:J f6; (0- 1 , 5 5) Korobov Smirin , Nakhchivan 20 1 1 .
14... c5
After 1 4 . . . bxa5 1 5 .ia3 i.h6 1 6.c5 ixd2
17.Wxd2 �xe4 1 8 .�xe4 fxe4 1 9 .i.d 1 !? White
would have a pleasant position.
I wanted to give a decent job to the
e7-knight. This piece (or more precisely a lack
of prospects for this piece) often causes trouble
for Black in the Bayonet Attack.
Also strong is 1 7.if3 ! ? �a6 1 8 .c5! and lack
of development is the main cause of Black's
troubles.
17 ... liJ b4 1 8.liJb3 ga6
Not 1 8 . . . �xe4? 1 9.i.xb4 axb4 20.Wd5t.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
208
King's Indian Warfare
19.exf5?
now White should be losing quickly.
Bareev returns the favour.
21. .. �xg4 22.i.xg4
He shouldn't help Black finish his development,
but instead he should open the position as
much as possible with the energetic 1 9 .c5 ! .
After 1 9 . . . dxc5 ( 1 9 . . . l2Jxe4 20.l2Jxd6 �xd6
2 1 .cxd6±; 1 9 . . . d5 20.exd5 l2J fxd5 2 1 .l2Jd6
with a decisive advantage) 20.Wxd8! �xd8
2 1 .�ad l �f8 22.l2Jxc5 White's pieces are very
active, and I would face major difficulties in
this endgame.
1 9 ...Lf5 20.c5
Now this is not as strong, as Black is already
much better prepared for the tactical dash.
b
a
20 ... �c2
This is the difference. Probably Evgeny
missed something in his calculations, because
the move I made is very obvious.
d
c
f
e
g
h
22 ...�g5?
Obvious, but wrong - in such posmons
precise calculation is much more important
than common sense.
22 . . . l2Jxe l was winning.
For instance: 23 .Wd5t �h8 24.ixf5 WgSt
2 5 . � h l �xf5
a
b
c
d
e
f
g
h
21 .g4?
H ow to describe this move: desperation?
If yes, it's way too early. Such a weakening
of the king in an open position is rarely any
good. After the "normal" 2 1 . l2J xd6 �xd6
22.cxd6 l2J xa3 2 3 . � c l White would retain
very decent fighting possibilities due to the
strong pawn on d6 and the unfortunate
position of the knight on a3, whereas
a
b
c
d
e
f
g
h
26.�xe l �xf2
Or: 23 .ixf5 Wg5t 24.�fl Wxf5 25 .Wxe l
Wh3t 26.�e2 e4 27.�d l Wf3t 28 .�d2 Wd3t
23.h3 �xel
This was asking to be played, but better was
the other capture: 23 . . . l2Jxa3! 24.Wd5t �h8
209
Chapter 6 - Knight Agility
25.lDxa3 h5 26.'1Md2 '!Mxd2 27.lDxd2 hxg4
28 . hxg4 .ixg4 29.lDac4 dxc5 30.ltJxe5 .ifS
and the extra pawn and two powerful bishops
allow us to assess this endgame as technically
winning.
24.�d5t �h8 25J�xel hg4 26.hxg4
%Yxg4t 27.�g2
30 . . . '1Mh3t 3 l .@gl Wg4t (3 1 . . J�f5 32.!!e3 and
White seemingly holds after 32 . . . Wh4 33.!!g3
We4 34.@fl Wc2 35 . .ic5) 32.@fl Wh3t=
3 I .Lfs gxf8
8
7
6
8
5
7
4
6
3
5
2
4
1
3
2
b
a
a
b
c
d
e
f
g
h
White has managed to bring his strongest
piece to the defence, and beat off the first wave
of the attack. Now Black switches his attention
to the queenside, where White's pieces (the
knight on b5 and bishop on a3) hang in the
air slightly.
27...�a4 28.�b7 d5!
Cutting off the queen from his king.
d
c
e
f
g
h
32.lbd2!
Someone has to help the king, and the
knight rushes to do it. Just in time!
32... e4
A good alternative was:
32 . . . Wh4!?
This is quite strong, but White has a narrow
path to hold:
33 .'1Mb6 e4!
33 . . . d4 34.@e2 .ih6 3 5 .@d l ! = The only
move (bad is 3 5 .!!fl d3t 36.@xd3 !!d8t ).
29.c6
The a3-bishop comes into play. At the same
time the pawn moves toward future glory.
Of course the rook was untouchable:
29.'1Mxa6? ? Wg4t 30.@fl Wf3 with mate.
29 ... �g4t 30.�fl gaas
We both had little time left. I decided to give
it a try and not to force a perpetual, correctly
sensing that Black might be able to do so later
if necessary.
a
b
c
d
e
f
g
h
210
King's I ndian Warfare
34.c7 d4 3 5 .<tt> e2 Wfxf2t 36.<tt> d l e3 37.lDc4
Wfxa2 38.ltJ ba3 Wfal t 39 .Wfb l Wfxb l t
40.ltJxb l �c8-+
34 . . . Wfh5t 3 5 .<tt> fl d4 36.ltJxd4 Wfh3t 37.<tt> g l
i.e5
33 YNh4
•••
33 . . . Wfg3 was winning too.
34.<ibe2 YNh5t 35.8
8
7
8
6
7
5
6
5
4
4
3
3
2
2
a
b
c
d
e
f
g
h
It looks as though the game is over, but there
is a defence:
38.ltJ 4f3! exf3 39.lDxf3=
Needless to say, all that was impossible to
find over the board, as both Evgeny and I were
down to the last drops of our remaining time
and played according to our intuition.
33.YHd7?
Bareev did not guess correctly.
1
b
a
d
c
f
e
g
h
35 exf3t?
•••
My turn to err.
After 35 . . .!hf3! 36.lDxf3 Wfxf3t 37.<tt> d2 the
white monarch would remain lonely and soon
fall: 37 . . . Wfd3t 38.<tt> c l i.h6t 39.<tt> b 2 Wfxb5t
40.<tt> c2 Wfc4t 4 1 .<tt> b2 Wfd4t 42.<tt> c2 Wfd2t
43.<tt> b 3 Wfb4t 44.<tt> c2 Wfc4t 45.<tt> b l Wfd3t
46. <tt> b2
The calm 33 .Wfb6!, protecting the pawn on f2,
was necessary and would lead to a draw in all
variations.
8
7
6
5
4
a
3
b
c
d
e
f
g
h
46 . . .i.f8! This quiet move is the shortest way
to victory. 47.Wfe8 Wfd2t 48.<tt> b3 (48.<tt> b l
Wfxe l t and so on) 4 8 . . . a4t 49.<tt> xa4 Wfb4#
2
a
b
c
d
e
f
g
h
The simplest would be: 33 . . . Wfh3t 34. <tt> e2
Wfg4t 3 5 .<tt> fl Wfh3t
36.<i>d3 g5?
The second mistake in a row.
Chapter 6
-
Black should have pushed the passer: 36 .. .f2
37.!!e8 Wf3t
211
Knight Agility
8
7
6
5
4
3
2
1
b
a
c
d
e
f
g
a
h
38.�c2 (38.ttJxf3 fl =Wt 39 .�c2 Wxf3
40.!!xf8t ixf8 4 1 .t2Jc3 Wf2t 42.�d3 Wf3t=)
38 . . . f1 =W 39.t2Jxfl Wxfl 40.!!xf8t Wxf8
4 1 . W xd5 W f2 t with perpetual check.
37.�xd5
I remember that after the d5-pawn
disappeared, I started to feel uneasy, but I had
no time to berate myself and made my next
move very quickly.
b
d
f
g
h
��-2������
c
e
41. ..g2
Any player who played the strongest
4 1 . . . h6!! should be rewarded for extraordinary
skills or banned for cheating. The possible
continuation after that could be: 42.ttJf7t
(42.c8=W !!xc8 43.t2Jxc8? ? g2-+) 42 . . . �h7
43.t2Jfg5t hxg5 44.!!h 1 t ih6 4 5 .Wb7 g2
46.c8=Wt !!f7 and the computer evaluates this
position as exactly equal. Sure!
42.gxf2 gl =�
The first promotion.
37... £2
43.gxf8t .txm 44.c8=�
a
b
c
d
e
f
g
h
38.gfl
38.!!h 1 was winning, according to Houdini .
38 ...�g6t 39.tlJe4 g4 40.tlJbd6 g3 4l .c7
a
b
c
d
e
f
g
h
The second promotion. We were already
playing on just the 1 0 second increments.
212
Ki ng's I ndian Warfare
In such a position we could only make
decisions purely by instinct.
Objectively better was: 44.�e5t �g7
45 .�xg7t @xg7 46.c8=� �fl t (46 . . . �b 1 t
47.�c2 �xc2t 48.@xc2 a4 49.@d3 @g6
50.�d4 h5 5 1 .�e5 ie7) 47.�c2 �e2t
48. �b 1 with winning chances.
Here I became desperate for about 1 5 seconds
until I found the right move with 5 - l 0 seconds
left.
46 ...Wgt t!
8
7
6
44 Wdl t 45.®c4 Wa4t
5
...
8
4
7
3
6
2
5
4
3
2
1
a
b
c
d
e
f
g
h
46.®c5??
I should be thankful to Evgeny for this move,
as it leads to a mate of exceptional beauty.
46.�c3 �g7t 47.@d2 �b4t 48.�c2 �bb2t
49.�d3 �b l t 50.�c2 �fl t 5 1 .�e2 �b l t
52.�c2 would have led to a draw.
b
d
f
g
h
0-1
One of the most unusual pictures I have ever
experienced in my games.
a
c
e
Chapter?
Kamikaze Rooks
With Alexander Moiseenko, the 2013 European
Champion; a great player and a great human being
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
... 8
7
6
5
4
3
2
2
a
b
c
d
e
f
g
h
a
3
2
f
g
h
2
a
Spot the weakness
(see page 2 1 5)
b
c
d
e
f
g
h
How do you generate
compensation for the piece?
(see page 220)
e
f
g
h
The bishop looks trapped
(see page 2 1 6)
d
e
f
g
h
4
4
6
5
3
d
c
7
3
2
c
b
Where should Black place the
rook? (see page 23 1 )
... 8
5
b
a
... 8
7
6
a
h
4
3
e
g
5
4
d
f
6
5
c
e
7
6
b
d
... 8
7
a
c
The next one is even worse!
(see page 22 1 )
Positional play. How should
Black react? (see page 2 1 9)
... 8
b
2
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
I missed a beautiful win
The grand finale!
(see page 22 1 )
(see page 232)
h
215
Chapter 7 - Kamikaze Rooks
I n a few o f my games I have managed to use
my rooks almost as if they were minor pieces,
using them with the same sort of energy
and short manoeuvres as a knight or bishop
would use in an attack. I am not sure where I
picked up this idea (I dread to call it a skill, as
I have not been able to show it often enough
for that) of using the rook like this; a logical
place would be from the games of Mikhail Tal,
who was absolutely sensational with all of the
pieces, including the rooks. More about Tal in
the notes to my game against Haba below.
I have gone over a number of recent games to
find one by another player that showed the
same handling of the rooks, and I came up
with the following game, where Black's entire
strategy is centred around a rampant rook.
1 7 . . . ttJh5! Black is already completely winning.
1 8 JHb 1 ( 1 8.h3 ltJc6! 1 9.dxc6 ixh3! also leads
straight to mate) 1 8 . . . ltJg3t 1 9.hxg3 ttJ f5
20.exf5 �g5 2 l .ltJfl �h6t 22.ltJh2 fxg3 0- 1
A. Fedorov - Ozolin, Kazan 20 1 3.
1 3 ... 5 14 ..id3 t!Jf6 15.c5 fxe4 16.t!Jxe4
t!Jxe4 17.Le4 �5
8
7
6
5
4
3
2
Andrey Zontakh Jaroslav Ulko
-
Voronezh 20 1 5
a
l.d4 t!Jf6 2.c4 g6 3.t!Jc3 �g7 4.e4 d6 s.ttJf3
0-0 6..ie2 e5 7.0-0 t!Jc6 8.d5 t!Je7 9.b4 a5
10 ..ia3 axb4 1 1 .�xb4 t!Jd7 12.a4�h6 13.a5
I do not want to go into the theory here,
but I would like to quickly give a game where
a strong grandmaster lost to a well-known
trick that has already been described in this
book: 1 3 .ttJd2 f5 1 4.a5 ttJ f6 1 5 . f3? Inviting
the bishop in is not a good idea. 1 5 . . . ie3t
1 6.�h l f4+ 1 7.�c2?
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
18.cxd6
This reduces White's options, but not Black's,
as . . . dxc5 was not really threatened.
1 8 .�e l is more critical. Compared with our
game, after 1 8 . . . ixe4 1 9 .�xe4 �d7 White has
20.�b3!± as played in two other games. But
Black can play differently too, of course.
18 ... cxd6 19J�e1
21 .'iMd3 gf6
�xe4 20J�xe4 'iMd7
Black plays this game quite aggressively. It is
not without risks, but sometimes you need to
take risks in order to win games.
But what we really need to see here is that
Black has placed the rook on f6 and not f7, as
I am sure some would do intuitively. This is a
fine j udgement. Not only is the d6-pawn in
more danger than the b 7 -pawn, Black also has
the option of swinging the rook to h6 later on,
and it is not improbable that he was dreaming
of doing this here.
216
King's I ndian Warfare
8
way to deal with the questions he has been
posed.
7
6
26.�dl ?
5
26.Wfl was necessary. After 26 . . . Wh5 Black
continues to enjoy a good game, but White's
position is not objectively worse. The most
prudent way to continue is probably 27.We2
Wh3 28.Wfl , with a draw by repetition
coming up.
4
3
2
1
b
a
d
c
f
e
g
h
22.tDd2 �affi 23.tDc4 .i£4!?
Black is preparing to swing the rook to
the h-file. 23 . . . Wb5 would have more or less
equalized, but again Black is going all-in.
24..ic3 g5! 25.g3
Black's attacking potential can be seen in
lines such as this: 2 5 . a6!? bxa6 26J�xa6 ltJ f5
27 . .ib4 El:h6! 28 .h3 g4!
8
7
6
5
4
3
2
1
b
a
8
d
c
f
e
g
h
26... �h6!
7
Black must have been ecstatic to be able to
carry out his plan. This gives him an excellent
chance to show off his agility with the rook.
6
5
4
But it would be a shame not to mention that
Black had an additional win with: 26 . . . .ie3!?
27.Wxe3 El:xf2 28.Wxf2 El:xf2 29.�xf2 Wxh2t
3
2
a
b
c
d
e
f
g
h
The game has sharpened up and White needs
to be very careful not to be worse immediately.
For example: 29.hxg4? .ih2t! 30.�hl ltJg3t!
3 l .fxg3 .ixg3t 32.�gl .if2t 33.�fl .ic5t
and Black wins.
25 ... �h3
Black has played very creatively and is
rewarded for his ingenuity and risk-taking
when his opponent fails to find a reasonable
a
b
c
d
e
f
g
h
Chapter 7
-
217
Kam ikaze Rooks
30.�f3 �h5t 31.g4 �h3t 32.�e2 �xc3, and
although there is some work to do, it should
not be too difficult.
33 dxe5 34.d6 tlJc6 35.VNh3t <i>hs 36.d7
gh3
.••
8
7
27.gxf4 VNxh2t 2s.<it>fl
6
8
5
7
6
4
4
2
3
5
3
2
a
1
a
b
c
d
e
f
g
h
28 gh3!
.••
The rook continues to play a brilliant
supporting role, creating a path for the queen.
29.ge3
Both 29.�c2 �xc3 and 29.ttJe3 gxf4 win
immediately.
29 VNxf4 30.Ael gh2 3 1 .gf3!?
•••
Hoping to escape to an endgame an exchange
down. Although this too was winning for
Black, the game is more convincing.
31. ghl t 32.<i>e2 VNg4!
••
This pin is decisive.
33.tlJxe5
Desperation, but 33.ttJd2 also does not
come close to holding. For example: 33 . . . ttJ f5
34.�e4 ttJd4t 3 5 .�d3 �xe4t 36.�xe4 ttJxf3
37.ttJ xf3 �f4t 38 .�e3 �h3 and Black wins.
b
c
d
e
f
g
h
The rook returns to h3 with devastating
effect.
37.gdd3 ghxf3 38.E!xf3 tlJ d4t 39.<i>fl tljxf3
4o.VNd5 tiJh2#
0-1
The rook's amazing journey on a very small
space is shown here:
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
218
King's I ndian Warfare
6 ... h6
My Games
The following three games are hard to describe,
with no simple theme to be squeezed from
them with the exception of the kamikaze rook:
a rook that runs rampant and is willing to die
for the cause if necessary.
Vyacheslav Dydyshko - Ilya Smirin
Byelorussia 1 986
The popular move in those days. Later I
switched mainly to the more sophisticated
6 . . . 4J a6, examples of which you will find in
this book.
The immediate 6 . . . e5? 7.dxe5 dxe5 8 .'1Wxd8
�xd8 9.4Jd5 leads to material losses - this a
well-known idea behind the Averbakh System.
7.�e3 e5
Now this is allowed.
This was one of my memorable games as a
teenager. It was played in the Championship
of Belarus and my opponent was Vyacheslav
Dydyshko, one of the strongest Byelorussian
masters. Vyacheslav holds the record for the
most tides as Champion of Belarus. He has
been a grandmaster for many years now, but
in the 1 980s the average master in the USSR
was no weaker than the average modern GM
(of course, this is just my personal opinion) .
In this tournament I earned the tide of USSR
National Master - an important step in my
career.
The more flexible (and probably better) move
8 . . . 4J a6 was played in the game Aleksandrov
- Smirin, Minsk 1987, which continued
9.Wfd2 h5 1 0.0-0-0?! (better was 10.f3 - the
light-squared bishop is very important for
White in this structure) 1 0 . . . 4Jg4 11 ..ixg4
.ixg4 12.4Jge2 id7 1 3.c47b1 f5; (0-1, 54) .
l .d4 tLlf6 2.c4 g6 3.tLlc3 �g7 4.e4 d6 5.�e2
0-0 6.�g5
9 ... h5 IO.f3 a6
a
b
c
d
e
f
g
h
The Averbakh System, named after Yuri
Averbakh, at the time of writing the oldest
living grandmaster.
8.d5 c6
9.'l;Yd2
Here we see the main drawback of 6 . . . h6 Black has to protect this pawn.
a
b
c
d
e
f
g
h
Preparing . . . b7-b5 with activity on the
queenside.
219
Chapter 7 - Kam ikaze Rooks
1 1 .0-0-0
Now the king on b 1 starts to feel some
discomfort.
Playing slightly into Black's hands.
More to my taste is 1 l .id3!? with the possible
continuation: 1 l . .. b5 12.tLl ge2 The standard
method of developing the gl -knight in the
Averbakh System. 12 . . . b4 (12 . . . cxd5?! 13.cxd5
tLlh7 14.0-0 tLld7 15 .b4 f5 16.exf5 gxf5
17.a4±; [V2-V2, 32] A. Petrosian - Smirin,
Polanica Zdroj 1989) 13 .tLld1 a5 14.a3
14.cxb5 axb5 1 5.tlJxb5 Y*fb6 16.�c2
A prophylactic measure against 16 . . . id7
17.tLlc3 �xa3 , but isn't it too early for the king
to be involved in the battle?
16.�c1 seems to be a more natural choice. In
this case I would proceed: 16 . . . ttJ a6 (16 . . .id7?
does not work: 17.tLlxd6! Wxd6 18 .ixc5 Wc7
19 .ixf8 +-) 17.�c3!? An interesting resource.
17 . . . tLlc7 18 .�b3 id7! Black has good
compensation for the pawn, at least from a
practical point of view.
8
7
6
b
a
c
d
e
f
g
5
h
would probably prefer White's chances
here - Black's pawns are slightly over-extended
on both wings.
ll. ..Y*fa5 12.�bl c5
Instead 12 . . . cxd5? 13 .ttJxd5 Wxd2 14.tLl e7t
±>h7 15jhd2 ie6 16jhd6 does not look
good for Black.
4
3
2
1
a
b
c
d
e
f
g
h
16 .. J!a4!?
Not a necessary move, and the objective
evaluation is closer to "?".
13.a3 b5!
8
16 . . . id7 17.tLlc3 tL!a6, with very decent play
for the sacrificed pawn, was called for, but I
j ust really liked the idea in the game, and could
not resist temptation!
7
6
5
4
17.b3 tlJxe4!
3
Here it is! The objective evaluation for this
move is also given above.
2
1 8.fxe4 gxe4
a
b
c
d
e
f
g
h
220
King's Indian Warfare
8
7
Adding fuel to the fire; the rook on d4 makes
a nice aesthetic impression. My opponent has
to accept the second sacrifice.
5
2I ..ixd4 cxd4 22. tlJ a4 YMa7
6
4
3
2
1
a
b
c
d
e
f
g
h
So, only one pawn for a knight, I am
underdeveloped and I am attacking paraphrasing Napoleon Bonaparte. The main
purpose of the sacrifi c e was to open lines
for both black bishops. In particular, the
g7 -bishop may turn into a monster.
19.'tt>b2
One possible line here is 1 9 .id3 ifS 20.a4
lLl a6 2l.ixe4 ixe4t 22.�b2 id3 23.lLle2
ixbS 24.axbS e4t 2 S . lLl c3 �b8!, with strong
threats by the rook. Houdini recommends
1 9 .ic4 and promises a big advantage to
White. The computer is probably right, but
Houdini did not exist in 1 986 . . .
Now Black is playing a whole rook down
with no immediate threats. Still he has huge
compensation. The central pawns will move
shortly, opening up the g7 -bishop at the same
time. And this bishop . . . Well, at the moment
it's much stronger than either ofWhite's rooks.
23 ..ic4 d3! 24. 'tt>a2 e4
As was j ust said.
25.tlJh3?!
Not the best choice.
More prudent was: 2S .iWaS iWf2t (less
convincing is 2S . . . iWd4?! 26.lLle2 iWf2 27.'\Wel
'1Wxg2 28.�d2 dxe2 29.�xe2 iWgS 30.�xe4
when White might be better but it's still messy)
26.iWd2 (26.�d2 ? ? iWel-+) 26 . . . '1Wa7
8
7
6
5
4
19. . .tlJ a6 20.tlJc3 �d4!
3
8
2
7
6
a
5
b
c
d
e
f
g
h
27.'\WaS, with a draw.
4
Or 2S .�el ifS 26.lLle2!? though Black still has
lots of play after 26 . . . lLl cS ! .
3
2
25 ... tlJc5?
a
b
c
d
e
f
g
h
A pity. In this case, it was better to attack the
unfortunate knight on a4 rather than exchange
it.
221
Chapter 7 - Kamikaze Rooks
After the natural and logical 25 . . . id7 White
would be on the verge of collapse. For instance:
26.ixa6 (26JWa5 ixa4 27.�xa4 �d4 28.�d2
�c3-+) 26 . . . �xa6 27.�he l ixa4 28.�xe4
29 ...�xd5?
Alas, here I missed a really pretty win:
29 . . .ixd3 30.�xd3 ic3! 3 1 .� b l (3 1 .�c l ?
�xa3t! - the point) 3 1 . . .�b8 32 .�c2 �xa3
33.�xc3 �xb3t 34.�xb3 �xb3t 3 5 .�cl
�c4t 36.� b l �e4t 37.�c l �xg2, with a
completely won endgame. To some degree
such mistakes from both sides can be explained
by the unusual character of the game.
30.tlJb4?
b
a
c
d
e
f
g
Another oversight. White needed to find
30.tLlb2 when the fight continues. After what
he played, the game could immediately be
over.
h
28 . . . ixb3t 29.�xb3 �bSt 30.�b4 �xdSt
3 1 .�c4 �xg2 32.�d2 �xh3 And Black wins.
26.lt�xc5 �xc5 27.�xd3!?
Not a bad practical decision. The pawn
chain e4-d3 was too strong.
27 ... exd3 28.tlJf4 �f5
Trickier was: 28 . . . ic3 29 .ltJxd3 (29.�xd3?
ig4-+ 30.ltJe2? �xa3t 3 1 .�xa3 �a8t - a
cheap but nice trap) 29 . . . �a5 30.�e3 id?
3 1 .tLlb2 (the best defence against . . . �xa3t)
3 1 . .. �e8 32.�f2 ib5 , and the only move here
seems to be 33.�hfl , with a dynamic balance.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
30 ...�e5?
Well, I missed it too. The hard-to-spot killer
was 30 . . . ib 1 t!! when taking with the rook
drops the queen, while 3 1 .�xb l �xb3t is also
a quick mate. It's all obvious to an engine, but
even years after the game, I had no idea there
was a forced mate here.
29.tlJxd3?
Better was 29.�c l �b5 30.�c400•
8
7
6
5
3 1 .�hel
3 1 .ltJc6 could have posed more problems for
me, but Vyacheslav decided to centralize his
rook and decentralize my queen. The price for
that - the pawn on h2.
4
3
2
1
3 1 . ..�xh2
a
b
c
d
e
f
g
h
222
King's Indian Warfare
Now material is about even and so are the
chances. I do not remember exactly, but it
seems to me that somewhere around here
Dydyshko started to get into time trouble.
32.ltJd5 i.e6 33.'i'f2 �b8 34.�d3 h4??
A huge mistake which could have cost the
game. Black is going for too much, playing on
both flanks and moving a pawn away from his
own king.
34 . . . �b5 seems to be the most precise
continuation here. Mter that, all three
outcomes are possible.
8
37....td4
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
38.'i'gl?
7
Just an awful move; White loses composure
in time trouble.
6
5
Mter 38.ttJe7 ie6 39.Wi"c2! ixe3 40.�xe3
Wi"f4 4 1 .�xe6 fxe6 42.Wi"xg6t @h8 43.b4 my
opponent could probably hold, despite his
oversight.
4
3
2
1
38 ... 'i'xgl 39.�xgl i.xe3 40.ltJxe3 i.e6
a
b
c
d
e
f
g
h
35.ltJ f4??
The simple 3 5 .�xe6 fxe6 36.ttJe7t @h?
37.�h3 would put an immediate end to the
game. One (albeit big) mistake could trample
over all my creative play. C'est la vie!
35 ....tf5
Black is on track again.
36.�de3 �h7 37.ltJd5?!
A blunder, though not a terrible one.
The resulting endgame is easily winning
for Black. I was not at my most precise in the
endgame, but the advantage was too big for it
to matter.
4I.ltJc4 hc4 42.bxc4 g5 43.�dl �e8 44.a4
�e2t 45.�b3 �xg2 46.a5 �e2 47.�xd6 h3
48.a6 h2 49.�dl g4
I enjoyed this game very much at the time.
I still love it, and even feel somewhat proud of
its risky and intuitional sacrificial spirit, and
for the imagination I showed.
0-1
223
Chapter 7 - Kamikaze Rooks
Ivery Chighladze - Ilya Smirin
Baku 2009
I .d4 llJf6 2.c4 g6 3.llJc3 .ig7 4.e4 d6 5.f3
8
7
6
9.llJcl
The standard manoeuvre in such positions.
Usually White transfers this knight to
b3, develops the bishop to e2, and castles
(sometimes queenside) . If he manages to do so
without concessions, White usually gets better
chances in the middlegame.
9.Ek 1 b5 1 0.b3 (if 1 0.cxb 5 axb5 1 1 .ctJxb5
:gxa2 the position would open in Black's
favour, who has better development and a
great game) 1 0 . . . bxc4 1 1 .bxc4 :gb8 1 2 . ctJ d 1 a5
happened in the rapid game Postny - Smirin,
Israel 20 1 0, which I managed to win.
5
4
3
2
1
9 ... llJ h5!?
a
b
c
d
e
f
g
h
The Samisch System, named after German
Maestro Fritz Samisch, who first employed it
at a high level. White strengthens the pawn
on e4, building a strong pawn centre. About
the only serious drawback of this setup was
pointed out by a famous admirer of the KI D,
GM Eduard Gufeld: "Do you like the Samisch
System? Ask the knight on g 1 ." Indeed, this
knight is now deprived of its most natural
square f3 and has to be developed somewhere
else: to e2 (in which case the bishop on fl will
be temporarily shut in) or h3.
5 0-0 6 ..ie3 llJc6
...
One of many options for Black. I used to play
virtually every move - 6 . . . c5, 6 . . . e5 , 6 . . . tt:J bd7,
6 . . . a6 and 6 . . . c6 - but I have to confess that
6 . . . tt:Jc6 is my favourite continuation.
7.llJge2 a6 8.'iMd2 .id?
Compared with the usual 8 . . J�b8, this move
makes it less favourable for White to play the
nowadays fashionable 9.Ek1 - see the game
Postny - Smirin below.
a
b
c
d
e
f
g
h
I like this move in similar positions. It opens
up the g7 -bishop, increasing the pressure
against the pawn on d4. At the same time,
Black is ready to play . . . f5 in some lines.
IO.llJb3
An important point is that after 1 O.g4 e5
1 1 .d5? ( l l .dxe5 ctJxe5 1 2 . .ie2 ctJ f6oo) 1 1 . . . ctJ d4
Black would obtain a dream position.
IO ... a5
A change of plan. Now Black is going to
attack the knight on b3 after . . . a5-a4. The
pawn push . . . b7-b5 (the main aim of 7 . . . a6) is
King's Indian Warfare
224
8
not relevant any more. Note that the rook on
a8 is better placed in its initial position rather
than on b8.
7
l l .g4
5
After 1 l .� d 1 a4 1 2 . ttJ c l e5 1 3.d5 lt:J d4
1 4.4J 1 e2 ttJxe2 1 5 .ixe2 4J f4 1 6. 0-0; (0- 1 ,
42) L . Kaufman - Smirin, Rockville 20 1 3 ,
Black should have continued simply 1 6 . . . f5 ,
with a n ongoing initiative on the kingside.
6
4
3
2
8
7
a
6
b
c
d
e
f
g
h
14...VMh4t!?
This was the idea. After the prosaic 1 4 . . . tt:Jd4
1 5 .ttJxd4 exd4 1 6.ig5 �e8 1 7.id3 Black
would simply remain worse - there is not
enough play for the missing pawn.
5
4
3
2
15.1£2 VMxh5 16.dxc6 VMx£3 17.cxd7
a
b
c
d
e
f
g
h
l l . ..a4 12.�xa4
After 1 2. 4J c l the brief tactics after 1 2 . . . e5!
would lead to a dream position for Black.
12 .. J�xa4
Also interesting and perhaps stronger was
1 2 . . . e5! with complications all over the board,
for example: 1 3 .gxh5 (or 1 3 .ttJc3 - the best
move - 1 3 . . . exd4 1 4. 4Jxd4 �h4t 1 5 .if2
4J xd4 1 6.�xd4 ixd4 1 7.ixh4 4J f4 with full
compensation for the pawn in the endgame)
1 3 . . . exd4 1 4.4Jxd4? ( 1 4.ig5 �e8 and White
has to give up the knight on a4) 1 4 . . . �h4t
1 5 .if2 ttJxd4 and Black is close to the win.
1 3.gxh5 e5 14.d5
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
17 ...VMxht
I declined the obvious 1 7 . . . �xe4t 1 8 .�e2
�xh 1 because after 1 9.ih4 �aa8 20.�g2!
�xg2 2 1 .ixg2 White would have a serious
advantage - the d7 -passer, together with the
two bishops, is too much for Black to match.
After the move I made an irrational (from
both positional and material points of view)
situation has arisen.
225
Chapter 7 - Kamikaze Rooks
18.llJc5?!
It was hard for my opponent to make the
right choice. White has a slight material
advantage - two pieces versus rook and pawn,
but even more important is that the d7-pawn
can be awfully dangerous. On the other hand,
White has not finished his piece development
yet and his pieces don't coordinate in the best
way. Black should be careful with his queen ­
it is quite active in the corner, but it may still
be in trouble in some variations.
The best move in the position seems to be
1 8.'l!Md5! centralizing the queen and protecting
the pawn on c4 . Probably the best Black could
do in this case would be: 1 8 .. J�d8 1 9 .'l!Mxb7 c5
( 1 9 . . . ih6!?) 20.ig3 (or 20.'l!Md5 ih6 2 l .ig3
'l!Mg 1 22.ltJ d2 'l!Me3t 23 .ie2 �xd7 24.ttJf3 with
a messy position) 20 . . . �xc4 2 l .@f2 'l!Mxe4
22.'l!Mxe4 �xe4 23.�d 1 White is better in this
complex endgame, even though there is a
lot of struggle ahead after 23 . . . if8 (but not
23 . . . �xd7? ? 24.ttJxc5) .
18 .. J�xc4
It was hard to resist the temptation to take a
pawn, but much stronger was: 1 8 . . . �aa8!
25.�d2 (25 .@xa2? 'l!Mc2-+) 25 . . . 'l!Mh l 26.@xa2
�a8t 27.mb3 'l!Mxfl with a material edge (four
pawns for the knight) and the initiative.
19.YHe2 b5 20.b3
The best move in the position.
20.. Jk3
Not falling into the trap: 20 . . . �xc5? 2 I .ixc5
dxc5 22.0-0-0 �d8 23 .ih3 and the black
queen is gone. At the same time the black rook
starts a remarkable slalom.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
2I ..ih4?
Oops, wrong idea! White had to make a
move with the knight:
a
b
c
d
e
f
g
h
19.ih4 dxc5 20.d8='l!M �fxd8 2 l .ixd8 'l!Mxe4t
22.ie2 'l!Mh 1 t 23.ifl ih6 24.'l!Mxh6 �xd8 and
Black is obviously better.
1 9 .0-0-0 'l!Mxh2 20.ltJd3 ih6 2 l .ie3 ixe3
22.'l!Mxe3 �xa2 23.@b 1 �fa8 24.d8='l!Mt �xd8
2 1 .ltJd3 f5 22.ih4 �c2!? (or 22 . . . fxe4 23.ttJf2
'l!Mxh2 24.d8='l!M �xd8 25 .ixd8 with a very
messy position and a possible perpetual after:
25 . . . ih6 26.'l!Mg4 id2t! 27.@xd2 'l!Mxf2t
28.@xc3 b4t 29.@xb4 'l!Mc5t 30.ma4 'l!Ma7t
3 l .mb5 'l!Mb6t 32.ma4 'l!Ma7t) 23 .'l!Mxc2
'l!Mxe4t 24.m d l 'l!Mxh4 2 5 .�b l (25 .'l!Mxc7? e4)
25 . . . 'l!Md4, and nothing is clear yet.
2 1 .ltJb7 if6 22.@d l �f3 23 .@c2 leads to
quite a wild positon, which I suspect is easier
to handle with Black over the board.
226
King's Indian Warfare
8
The rook slalom has been completed! It is
hard to find another 25-move game in which
a rook has travelled a similar route (a8-a4c4-c3-h3-h2) and, together with a queen on
h 1 (!) , created a mating attack against the white
king. The two extra pieces don't help in this
situation.
7
6
5
4
3
2
24.�xb5
a
b
c
d
e
f
g
h
21 ..J�h3!
This loses immediately, but even after the
most resilient 24.WI'g4 dxc5 2 5 .0-0-0 �xa2
26.ie2 Wfh2 27.�d2 h5 28.Wfd7 Wff4! 29.ixc7
ih6 White is absolutely helpless.
24...�f3 25 ..ie2 �c3t
The slalom continues!
22.d8=�?
Chigladze loses control over the game. After
the only continuation 22.ig3 �xg3 23.hxg3
dxc5 24.�d l �d8 2 5 .Wfg2 Wfxg2 26.ixg2 b4 it
would not be an easy task for Black to convert
the two extra pawns into a win. Indeed, in
the resulting endgame the opposite-coloured
bishops plus the strong passer on d7 give
White reasonable drawing chances.
22 .. J�xd8 23 ..ixd8 gxh2!
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
So although the rook did its job, it survived till
the end. It has no such luck in the next game,
where a runaway rook creates weaknesses in
the white camp, sharpening the game, before
it ends with a surprising little tactic.
Chapter 7
-
227
Kamikaze Rooks
Petr Haba - Ilya Smirin
Again, that's my personal taste - the position
is perfectly playable after 1 O . . .ig4 1 1 .h3 ixf3
1 2 .'1Wxf3 lt:J bd7 1 3 .'1Wd 1 etc.
Polanica Zdroj 1989
This game was played in a tournament in
Polanica Zdroj, Poland in memory of the
outstanding player Akiba Rubinstein. In
this tournament I made my second and final
norm to obtain the tide of International
Grandmaster. My first norm came in the Rilton
Cup tournament in Sweden ( 1 988/89) . The
Rilton Cup was my fi r st tournament outside of
the USSR (at the age of nearly 2 1 ) and my first
possibility to make the norm . Polanica Zdroj
was my second chance.
l.d4 llJf6 2.c4 c5 3.d5 g6 4.lDc3 �g7 5.e4
0-0 6.lDf3 e6
8
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
9.lDd2
The most popular and ambitious move White wants to take on d5 with the c-pawn
after an eventual . . . e6xd5. It has an obvious
drawback though - the bishop on c 1 is
temporarily shut in by the knight.
7
6
5
4
3
2
a
b
c
d
e
f
g
h
7.�e2
I consider 7 .id3 to be a more unpleasant
continuation for Black, but this is a subjective
opinion, of course.
7 ... d6 8.0-0 �e8
In the late 1 980s I loved to play like this: to
postpone the exchange on d5 for a few moves.
I wanted to avoid the position after
8 . . . exd5 9.exd5!? (9.cxd5 El:e8) 9 . . . El:e8 1 O.id3 ,
considering it to be slightly passive for Black.
9.h3 exd5 1 0 .exd5 ifS 1 1 .id3 lt:J e4 1 2.lt:Jxe4
ixe4 has been played a few dozen times and
leads to approximate equality; after exchanges
it's easier for Black to play - the opponent's
space advantage is less significant.
9.dxe6 is another plan: after 9 . . .ixe6 1 0.if4
lt:J c6 1 l .ixd6 lt:Jd4 ( 1 1 . . . '\WaS !? is the riskier
option) 1 2.e5 lt:Jd7 1 3 .lt:Jxd4 cxd4 1 4 .'1Wxd4
lDxe5 1 5 .ixe5 '1Wxd4 1 6.ixd4 ixd4 1 7.El:ac l
El:ad8 Black has good compensation for the
pawn in the endgame. The vast majority of
such games have ended peacefully.
9 ... llJa6 IO.f4
Here we see another reason for Black to
delay the capture . . . exd5: now 1 O.f3 with
the idea of playing lDc4 makes no sense the c4-square is still occupied by the white
pawn.
King's Indian Warfare
228
A well-known game between Smyslov and
Polugaevsky continued: 1 O.�e 1 t:tJc7 1 1 .a4 b6
Polugaevsky: "I got excited, overestimated
my position. I should have continued 3.t:tJf3''
(after l .d4 t:tJf6 2.c4 c5) .
Also possible is: 1 O.dxe6 �xe6!?
a
b
c
d
e
f
g
h
1 2.�b 1 �b8 1 3 .b3 a6 1 4 .ib2 exd5 All the
preparatory moves have been made, and the
time has come for this exchange. 1 5 .cxd5 b5
1 6.axb5 axb5 1 7.ifl
8
7
6
5
a
b
c
d
e
f
g
h
This was tested in one of my games: 1 l .f3
t:tJb4 1 2.t:tJb3 id7 1 3 .ie3 a6 1 4.a4 ic6
The defects of Black's pawn structure are
compensated for by the good position of
his pieces. 1 5 .Wd2 t:tJd7 1 6.�fe l f5 1 7.ig5
Wf8 1 8 .t:tJd5 ixd5 1 9 .exd5 �ee8 20.ifl h6
2 1 .ih4
4
3
8
2
7
6
a
b
c
d
e
f
g
h
1 7 . . . t:tJg4! 1 8 .t:tJf3 f5 1 9 .h3 t:tJe5 20.t:tJxe5 ixe5
2 1 .exf5 ixf5 22.id3 Wf6 and Black won;
(0-1 , 38) Smyslov - Polugaevsky, Riga 1 975.
This game was played in the USSR Team
Championship in the famous match between
the Russian Federation and Moscow, which
ended with the score 8V2-V2 in favour of the
former - an incredible result considering that
the two teams were roughly equal in strength.
After this match ex-World Champion Vassily
Smyslov made his famous statement: "The
most important thing now is to save the
people." And in reference to his game with
5
4
3
2
a
b
c
d
e
f
g
h
2 l . . . g5 22.if2 ie5 23.a5 Wf6 with better
chances for Black; (0- 1 , 5 1 ) Bandza - Smirin,
Dortmund 1 990.
IO ... �c7 l l .a4 exd5
Two years before the game against Haba, I
played the following interesting encounter:
l l . . . �b8 1 2.�e l exd5 1 3.cxd5 a6 1 4 .a5
229
Chapter 7 - Kamikaze Rooks
8
8
7
7
6
6
5
5
4
4
3
3
2
2
a
b
c
d
e
f
g
h
1 4 . . . h5?!
An insolent kind of move, with the idea of
the knight jumping to g4.
The simple 14 . . .id7 1 5 .if3 ibS! ? with
good play was better - the d2-knight is still
deprived of the possibility of moving to c4.
1 5 .h3 b5 1 6.axb6 �xb6 1 7.if3 ?
Correct was: 1 7.lt.'lc4 � b 8 ( 1 7 . . . �b4? 1 8 .e5
dxe5 1 9 .d6 lt.'lb5 [ 1 9 ... exf4 20.dxc7 rffxc7
2 l .Wffd6±] 20.fxe5, with a big advantage)
1 8.e5 dxe5 1 9 .d6 lt.'lb5 20. fxe5 lt.'lxc3
2 1 .bxc3 lt.'ld7 22.if4 lt.'lxe5
a
b
c
d
e
f
g
h
1 8 . . . lt.'lfxd5!
A nice blow. Black wins a pawn because the
white queen is overloaded.
1 9.lt.'lxd5
Or 1 9.exd5 �xe 1 t 20.rffxe 1 ixc3 2 l .�xb4
ixb4 also leaves Black a pawn up.
1 9 . . .lt.'lxd5 20.�xb4 lt.'lxb4 2 1 .lt.'lc4 ie6
22.lt.'le3
Of course 22.lt.'lxd6?? fails to 22 . . . id4t.
22 ... rffh4 23 .�e2 id?!? 24.lDc4 .ib5 2 5 . b3
lt.'lc6 26.�a2 lt.'ld4 27.mh2
8
8
7
7
6
6
5
5
4
4
3
3
2
2
a
a
b
c
d
e
f
g
h
23.rffd 2! White's activity and strong
pawn on d6 provide at least enough
compensation for the minimal material
deficit.
1 7 . . . �b4!
Suddenly White runs out of active ideas.
1 8 .�a4?
b
c
d
e
f
g
h
27 . . . lt.'lxf3t 28.rffx f3 rffe 1 29.�c2
30.�xc4 id4
0- 1 Basin - Smirin, Minsk 1 987.
ixc4
12.cxd5 gbs 1 3.�hl
The position is very double-edged, but I like
Black's chances here - I believe that the white
pieces don't provide sufficient support for his
advanced pawn centre.
King's I ndian Warfare
230
If 1 3 .if3, protecting the e4-pawn and
thus preparing the key move lt:J c4, I would
probably reply with: 1 3 . . . b6 ( 1 3 . . . a6?! 1 4 . lt:J c4
b5 1 5 .axb5 axb5 1 6.lt:Jxd6 �xd6 1 7.e5 �d8
1 8 .d6, with a strong initiative for White)
1 4. lt:Jc4 ia6 1 5 .�b3 ixc4 1 6.�xc4 a6 This
is a kind of dream position for Black in this
structure - all his pieces are active and if
White tries a typical breakthrough with:
1 7.e5
1 3 . . . h5 ?
b
a
c
d
e
f
g
h
1 4 .f5 ! Immediately using the weakening of
the kingside created by Black's previous move.
1 4 . . . a6 1 5 .a5 gxf5?! 1 6.ixh5 lt:Jb5 1 7.exf5
E!:e5 1 8 .if3 E!:xf5 1 9.lt:Jc4 And it's hard to look
at Black's position without crying; ( 1 -0, 29)
M. Gurevich - Smirin, Moscow 1 988.
b
a
c
d
e
f
g
14.a5 .id? 15 ..if3 lL'lb5
h
Black would answer with the intermediate
17 . . . b5! 1 8 .axb5 axb5 1 9.�e2 b4, with an
obvious advantage.
8
8
7
6
5
7
4
5
2
6
3
4
3
2
a
1
a
b
c
d
e
f
g
h
1 3 ...a6
The game with Basin left pleasant memories,
so in a different game I tried the idea with
. . . h7 -h5 once more. But this time my opponent
was from a much higher weight category:
b
c
d
e
f
g
h
Agai n, as we saw in some lines above, all
Black's pieces are actively involved in the
battle, whereas White has not completed his
development yet. His next move is a natural
attempt to change the unfavourable course of
the game. But Black is ready for that.
16.e5 dx:e5 17.fxe5 �xe5 18.lL'lc4 �f5 19.lL'le3
Trying to hunt down the rook on f5 .
231
Chapter 7 - Kamikaze Rooks
19 .lt:Je2, with the same idea, was no better due
to:
b
a
c
d
e
f
g
2 I ..J�e4!?
Also 2 l . . . �b4 22 ..id2 �xb2 23.lt:Jc4 �xd2
24.'\Wxd2 ih3
h
19 . . . lt:J h5 (I was surprised to find that this
position had occurred in a recent game between
two top-ten players: 19 . . . �xf3 Actually the
opening was played slightly differently, and
this was the 21st move. 20.gxf3 ih3 2 l .�e l
�xd5 22.lt:J f4 �xd l 23 .�xd l id7 24.ie3
ic6, with better prospects for Black in the
endgame, though Nakamura - Vachier­
Lagrave, London 2015, was later drawn.)
20.ie3 lt:Jd4! 2 l .lt:Jxd4 cxd4 Black is better,
in particular because 22 . .ixd4? loses material
after: 22 . . . �h4 23.ixg7 lt:Jg3t 24.rngl lt:Jxfl
b
a
c
d
e
f
g
h
with a big advantage, was a good version
of the exchange sacrifice as well. But instead
I used my chance to follow in the footsteps
of my favourite player, Mikhail Tal, at such
an advanced stage of the game. I was not 100
per cent sure about it, but it seemed to me
that I had seen exactly the same position in a
game ofTal's while reading a book I love very
much Vogon Ataki by Tal and Damsky. After
returning home, I discovered I was right!
-
22.be4 lbxe4
19 .. J�f4 20.lbe2 gh4!
Provoking
the
next
move,
which
compromises the white king's pawn shelter.
8
7
6
21 .g3
5
8
4
7
3
6
2
5
1
4
3
a
2
b
c
d
e
f
g
h
23.lbc4
1
a
b
c
d
e
f
g
h
Here my opponent deviated from 23.lt:Jf4,
which happened in the above mentioned game:
232
King's I ndian Warfare
23 . . . ttJd4 24.�g2 We7 (24 . . .ib5 ! 25 J�e 1
Wd7 26.:Ba3 :Be8 was even stronger) 25 .:Be 1
h5 26.:Ba3 :Be8 27.tDe2? (correct was 27.tDc4)
24 ll)f6?
•••
A sloppy move. 24 . . . ttJ ed6 25.lDxd6 Wxe3
26.ttJxb5 ixb5 would win easily.
25.<it>gl YlYe4 26JU4 YlYxd5 27J�xf6 YlYxc4
28J��a4?
Haba did not see the tactical refutation.
The correct 28.:Bf4 We6 29 .ixc5 :Be8 30.E!f2
ic6 3 1 .Wd3! ixb2 32.E!afl would keep
White's hopes alive despite being much worse
- j ust look at the bishop on c6 and the h 1 -a8
diagonal.
b
a
c
d
e
f
g
h
8
27 . . . ih3t! 28 .�xh3 lDg5t 0- 1 ; Ufimtsev Tal, USSR 1 967.
7
23 YlYe7 24 .ie3?
5
•••
•
White's position is bad, no doubt about it:
his king is weak and Black's pieces control
almost the entire board. Still after the best
try 24. ttJ f4 ttJ d4 25 .ie3 he could resist for a
while.
6
4
3
2
8
7
a
6
b
c
d
e
f
g
h
28 ll)c3!
•••
5
Too many white pieces are hanging - a
decisive loss of material is unavoidable.
3
29.YlYxd7ll)xa4 30JU4 YlYxe2 3I .YlY:xf7t <it>h8
32J�e4
4
2
a
b
c
d
e
f
g
h
Many roads lead to Rome - for instance
32 . . . :Bf8 33 .:Be8 Wxe3t. White resigned
without waiting for Black's 32nd move.
0-1
Chapter 8
Fighting for the Initiative
Never end the fight!
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
8
� 8
7
7
6
6
5
5
4
4
3
3
2
2
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
How should Black organize his Traditional King's Indian magic!
pieces? (see page 239)
(see page 26 1 )
8
� 8
5
5
� 8
7
7
7
6
6
6
5
4
4
4
3
3
3
2
2
a
b
c
d
e
f
g
2
h
a
What is the most energetic
continuation?
(see page 236)
8
7
� 8
7
5
4
5
4
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
A bit later: what is the most
energetic continuation?
(see page 236)
d
e
f
g
h
Warrior play such a position?
(see page 246)
(see page 264)
� 8
7
6
5
4
3
2
2
c
Should the queen go to f6 or f7?
3
3
b
How does a King's Indian
6
6
a
2
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
A difficult tactical moment.
How should White have
How does the attack win?
(see page 248)
continued the attack?
(see page 27 1 )
235
Chapter 8 - Fighting for the Initiative
One of the most important aspects of the
King's Indian is the fight for the initiative.
White has more space from the start, and if
nothing happens to combat this, then Black
can easily find himself with his back against
the wall. But as we have already seen, the
King's Indian has great dynamic potential.
Black seeks counterplay with all the themes
we have already examined and more, and is
always looking for a way to grab the initiative.
In the King's Indian this really is a key factor.
The games in this chapter show a brutal fight
for the initiative at all costs. Often pieces are
sacrificed simply to gain time. At other times
it is just an attitude all the way through the
game, as in the following recent game.
Abhijeet Gupta - Kaido Kulaots
example: 1 4 . . . g5 1 5 .tt:J g4 f5 1 6.exf5 i.xf5
1 7.h6 i.f6 1 8.i.e3 i.e7 1 9.�fl '!9e8 20.�g l
'!9g6, and Black was fine in Wimmer - Cill6niz
Razzeto, email 20 1 4 .
14...f5! IS.hS
This feels risky, but is quite interesting.
After the less aggressive 1 5 .gxf5 gxf5 1 6.i.g5
Black is fi n e, but not more.
IS ... fxe4! 16.hxg6 YMf6 17.�f7t?!
After this Black is able to set White a good
deal of problems.
1 7.tLlf5 '!9xg6 1 8 .tt:J h4 '!9f7 1 9.i.e3 with an
unclear game was more prudent.
17 ... <i>g8 1 8.�h6t J.xh6!
Cappelle-la-Grande 20 15
No draw today.
l .d4 �f6 2.c4 g6 3.�c3 J.g7 4.e4 d6 s.�f3
0-0 6.h3 eS 7.dS �hS 8.g3 aS 9.�h2 �a6
IO.J.e2 �f6 l l .�g4 �d7 12.�h6t <i>hs
13.h4 �deS
8
19.gxh7t <i>hs 20.J.xh6 J.xg4!
8
7
6
7
5
6
4
5
3
4
2
3
2
1
a
a
b
c
d
e
f
g
h
14.g4
I am not sure this is the best move here.
1 4.h5 feels a bit more critical, but Black's
position has plenty of resources too. For
b
c
d
e
f
g
h
Threatening a big check on d3.
21 .<tifd2 YMxfl 22.'iMgl?
This is the only serious mistake in the game.
22.i.e3 ! was necessary. White will just about be
able to hold, based on the following variation:
22 . . .i.xe2 23 .'!9xe2 \Wxe2t 24.�xe2 �f7
236
King's I ndian Warfare
25.gag l gxh7 26.gxh7t �xh7 27.gh l t �g7
28.gg l t � f7 29.gfl t �e8 30.gh l ! and Black
cannot make meaningful progress.
8
""'
The final error.
28.�a4 was necessary, leading to a long
vananon:
28 . . . ixc4
29 .gg8t
gxg8
30.hxg8=i t �xg8 3 l .ggl t �f8 32.ih6t
�eB 33.ttJd2 gh3 34.gg8t �f7 35 .gg7t �f6
36.ttJe4t �f5 37.ttJg3t gxg3 38.gxg3 ixa2+
Black is playing for a win with the three pawns
for the exchange, but White has some drawing
chances too.
��
7
27.. JH3 28J!g3?
Y,
6
5
4
3
8
2
a
b
c
d
e
f
g
h
22 ... e3t!
It feels as though Gupta had overlooked this
combination.
23.i.xe3
After 23.<tt> c l �xg l t 24.gxg l ixe2 2 5 .ixf8
(25 .ttJxe2? gf2-+) 2 5 . . . ttJ d3t 26.�b l ih5+
White is also in an unpleasant position.
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
28 ... b5!
23 .. .lDe4t! 24.®c2
Or 24.ttJxe4? �xe2t 25.�c3 ttJ b4 leads
directly to mate.
Breaking down the remains of the white
position.
24 ... tlJ b4t 25.®b3 YNxgl 26J!axgl .ixe2
27.tLlxe4
Mter 28 . . . gaf8 29 .a3 ttJ a6 30.ghgl , or any
other move that failed to fight for the initiative,
White would not be worse.
White is drifting into a lost position, but his
task was already difficult.
29.®c3
27.ttJxe2? gf3 loses on the spot, and after
27.gg8t gxg8 28 .hxg8 =�t �xg8 29.ttJxe2
gf8+ Black will definitely be playing for a win
in the endgame with an extra pawn. Still, this
might have been the best chance.
Black wins in all lines. Also after 29.cxb5
ttJxd5 30.ge l gxg3 3 l .ttJxg3 ixb5 and
29.gxf3 bxc4t.
29 ... bxc4 30.®d2 gxg3 3 1 .tLlxg3 .i8
Black has a winning endgame, but it took
him an additional 52 moves to convert it .
... 0-1
237
Chapter 8 - Fighting for the Initiative
1 1 . .. h6 12 ..ih4 g5 1 3 ..ig3 t[}f6
My Games
Viktor Korchnoi Ilya Smirin
-
After: 1 3 . . . h5 1 4.exf5 ixf5 (14 ... 11J f6 1 5 .h4)
1 5 .11Jde4 l!Mg6 1 6.�e 1 !
Dresden 1998
This game remains very special for me - my
first encounter with the great Viktor Korchnoi,
one of my chess heroes.
I.d4 t[}f6 2.c4 g6 3.tiJc3 .ig7 4.e4 d6 5 ..ie2
0-0 6.tiJf3 e5 7.0-0
b
a
c
d
e
f
g
h
Black would face an unpleasant problem his kingside activity has come to a stop and he
is j ust left with weaknesses.
14.0 f4 15 ..if2
a
b
c
d
e
f
g
h
7... tlJa6
I did not want to play 7 .. /ij c6 against
"Viktor the Terrible". He had won quite a few
very convincing games in the line 8.d5 l1J e7
9 . 11J e 1 CiJd7 1 0.ie3 and I was impressed by his
handling of the positions that arise.
8 ..ie3 tlJg4 9 ..ig5 'lNe8 10.d5
The main line here goes 1 O.dxe5 dxe5 1 1 .h3,
but at the end of the 20th century Korchnoi
followed modern theory less strictly, and often
liked to improvise over the board.
10 ... £5 1 I .tiJd2
Another path is: 1 l .CiJ e 1 11J f6 1 2 .f3 h6
1 3 .id2 g5 1 4.11Jd3 l!Mg6 1 5 . 11J f2 f4 1 6. b4 h5
In G. Nikolic - Antic, Tivat 1 99 5 , Black won
by a direct mating attack.
a
b
c
d
e
f
g
h
So the picture became clear - attacks on
opposite wings. My formidable opponent was
a great master in such positions as White, and
I often like to play them with the black pieces.
1 5 ...VNg6
I had preferred this move as a way to prepare
the . . . g5-g4 push rather than 1 5 . . . h5, which
would also lead to interesting play. For instance:
1 6.b4!? g4
238
King's Indian Warfare
1 6 . . . lLl xb4 1 7.cS , with an overwhelming
position for only a pawn.
1 7.:gb l
The more active option is: 1 7 .cS fi g6
16.�hl !
The first in a chain of deep prophylactic
moves.
16 ...g4
Again, 1 6 . . . hS 1 7.:gb l g4 1 8 .�h4 was not
to my taste. I wanted to leave the hS-square
vacant for a knight or queen, j ust in case.
17.fxg4 tlJxg4 18.-tgl gf7 19.tlJf3
Another option was the pawn sacrifice
1 9 .b4!? lLlxb4 20.cS, but it seems that after
20 . . . �f8 Black does not stand badly at all.
a
b
c
d
e
f
g
h
1 8 .c6! ( 1 8.�h4 lLlxb4 1 9 .cxd6 cxd6 20.lLlc4
lLl e8 and Black looks OK) This may lead
after: 1 8 . . . g3 1 9.�e l figS 20.h3 fih4 to a
typical sharp KID position, in which Black's
attack against the white king looks much
slower than White's activity on the opposite
side of the board. However, after the possible
continuation: 2 1 .cxb7 �xb7 (2 1 . .. �xh3?
22.gxh3 fixh3 23.:gf2 gxf2t 24.�xf2+-)
22.:gb l lLl b 8 23.lLlc4 �c8 24.:gb2 lLl h7
Black transfers the knight to gS and, after
a piece sacrifice on h3 at the appropriate
moment, would have very dangerous threats.
1 7 . . . fff g6 1 8 .�h4
8
7
6
5
4
19 ... .tf6 20.gel ! gg7 2 1 .-tfl
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Korchnoi made the last few moves (and
good ones, I should say) very quickly, and
here he looked at me with a laugh, as if saying:
"Well, young man [this is how he referred to
me immediately after the game] , who are you
going to attack? And how?"
2 1 . .. tlJc5 22.b4
3
2
a
b
c
d
e
f
g
h
1 8 . . . gxf3 1 9.�xf3 lLl g4 20.fie2 �f6 2 1 .�xf6
:gxf6 22.cS lLl e3 23.:gfc l
With a complex position.
Or 22.h3 lLle3 23 .�xe3 fxe3 24.:gxe3 aS !,
and in a practical game I would say that the
strong knight on cS, the initiative against
White's king on the g-file, and the bishop pair
outweigh the missing pawn.
22 ... tlJd7 23.c5
239
Chapter 8 - Fighting for the Initiative
Perhaps more precise was: 23.h3 lt:J e3
24 ..ixe3 fxe3
24.h3
During the game I was very surprised by the
fact that Viktor spent more than an hour on
this move! I was expecting it, and believed this
move to be the most natural - and almost only
- one.
Such variations as 24.!!b l tt:Jdf6 25 .gb3 lt:J h5
26.lt:Jb5 lt:J g3t 27.hxg3 fxg3 28 . .ie3 .ig5
29.@gl l!Nh5-+ cannot inspire White.
24... l!Je3?!
b
a
c
d
e
f
g
h
25.!he3 .ie7 (with the idea . . . lt:J d7-f6-h5)
26.c5! lt:J f6!? (26 ... dxc5 27.lt:Jb5 l!Nb6 28.bxc5
and White has an edge) 27.l!Nc2 lt:Jh5 and
nothing is clear.
8
7
6
5
4
3
I made this move instantly, within a few
seconds (after such lengthy thought by
Korchnoi) . I thought it was forced because of:
24 . . . tt:J gf6 2 5 . lt:J h4
Mter 25 .Ekl lt:J h5 26 ..if2 tt:Jdf6 27.!k2
lt:J g3t 28 . .ixg3 l!Nxg3 Black's chances should
be preferred - the dark squares around
White's king are weak, and White's dark­
squared bishop is gone.
25 . . . l!Ng3 26. lt:J f5 ? !
And I hadn't noticed, that:
26 . . . l!Nxc3
is possible, and after:
27.gb l
2
a
b
c
d
e
f
g
h
23 ....id8!
Probably the hardest move for me to find in
the game. It has two main goals: vacating the
f6-square for the knight on d7, and protecting
the weakest link in the black camp - the pawn
on c7. From f6 the knight will continue to h5
and sacrifice itself on g3, followed by a mating
attack with . . .l!Nh5t (see annotations to Black's
24th move) .
23 . . . .ie7 24.lt:Jb5 is what I was avoiding.
a
b
c
d
e
f
g
h
27 . . . tt:Jxc5! 28.bxc5 .ixf5 29.exf5 b6
Black has an extra pawn and the advantage.
A nice draw could be a result of: 24 . . . tt:Jdf6
2 5 . hxg4 (25 .lt:Jh4 lt:J h 5 ! 26.l!Nxg4 lt:J g3t
27.@h2 l!Nf7 28.l!Nf3 .ixh4+) 25 ... tt:J xg4 26.c6
(26.lt:Jh2? .ih4-+) 26 . . . l!Nh5t 27.lt:Jh2
240
King's Indian Warfare
26... ttJf6 27.cxb7
Again, 27 .!!xe3 was better.
27 ....ixb7 28.�xe3 tlJ hS 29.tlJe2
At this point Viktor Lvovich was already
in serious time trouble, and in such a sharp
position that is a very serious factor.
29 ... c6!?
b
a
c
d
e
f
g
h
27 . . . '!Mfxh2t! 28.ixh2 ltJ f2t 29 .�gl ltJ h3t
30.� hl ltJ f2t So now I understand why
Viktor thought for so long!
25.�xe3
Korchnoi looked at me somewhat angrily
(he had thought for so long and I had replied
immediately and quite weakly) and took my
knight.
Also possible was 29 . . . ia6 30.g4 ixe2
3 1 .ixe2 ltJ f4 with good play (the knight on
f4 is terrific) , but in White's time trouble I
wanted to open up both black bishops. The
nice square b6 is available for the bishop now.
From a practical point of view, I believe the
move I made was the best one.
30.g4!
8
7
6
25 ... fxe3
5
8
4
7
3
6
2
5
1
4
3
a
2
1
a
b
c
d
e
f
g
h
26.c6?!
He played this quickly as I remember.
After the simple 26.!!xe3 ltJ f6 27.'1Mfb3 ltJh5
28.ltJe2 I doubt Black would have sufficient
play for the pawn, although the position
remains tense.
b
c
d
e
f
g
h
Absolutely the correct decision . White
eliminates the pressure on the g-pawn,
simultaneously attacking the knight. This is
more important than the weakening of the
white king's pawn cover.
30 ... cxd5
This sacrifice was asked for, but a subtler
version of it would be:
30 . . . ib6! 3 1 .gxh5
3 1 .dxc6 ixc6 32 .'1Mfc2 ixe3 33 .'!Mfxc6 !!f8
34.ltJh4 '!Mfg5 3 5 .ltJf5 !!xf5 36.exf5 ltJ f6 with
a great position.
24 1
Chapter 8 - Fighting for the Initiative
3 1 . . . \Wxh5 32.dxc6 �xc6 33.\Wb3t c;t;>hs
Black is more than OK. For example:
34.Ek l �b5 3 5 . lt:J fg l �xe3 36.\Wxe3 �ag8
White, with his passive pieces, is in trouble.
Probably the best continuation would be
32 . . . d4 33.�g3 �xe4t 34.�g2 �xg3 3 5 .lt:Jxg3
�xg2t 36.c;t;>xg2 \Wxd l 37.�xd l �c8 with
good chances to draw, although White is the
stronger side here.
3I .gxh5 VMxhs
32 ....ih4!
8
8
7
7
6
6
5
5
4
4
3
3
2
a
b
c
d
e
f
g
h
2
1
a
32.tlJfgl?
The only serious mistake by Korchnoi in
the game, but already the decisive one! The
outstanding defender quickly spotted the right
idea - to prepare �e3-g3 - but in intense zeitnot
he could not find the correct way. Indeed, to
find under any circumstances the only right
move, 32.lt:Jh2!, was almost impossible. It
seems that White would hold the advantage in
this case!
There are variations like:
32 . . . \Wh4 33.iWb3! c;t;>hs (33 . . . �b6 34.�g3
\Wxe4t? 3 5 .�g2 \Wxe2 36.�xd5t +-) 34.�g2
�xg2 3 5 .c;t;>xg2 d4 36.�fl ! �b6 37.�g3 \Wxe4t
38.lt:J f3 \Wxe2t 39.�f2 �xf3t 40.�gxf3 \We4
4 I .c;t;>h2 d5 42.\Wa4!+-
b
c
d
e
f
g
h
33.tlJf3?
After 33.exd5 �f8 34.\Wb3 e4 Black's assault
is irresistible. White does not have the resource
lt:J h2-g4 and this detail proves to be decisive 3 5 .�dl \Wg6.
Even after the most stubborn computer
defence: 33 .\Wb3 c;t;>hs 34.exd5 �ag8 All
Black's pieces are in the attack! 3 5 .�d l �f2
36.�f3 �xg l 37.lt:Jg3 �xg3 3 8.�xg3 �xg3
39.\Wxg3 \Wxd l 40.c;t;>xg l \Wxd5 4 l .a4 \Wd4t
42.\Wf2 \Wxf2t 43.c;t;>xf2 c;t;>g7 44.<j{e3 d5 Black
wins in the endgame. Still, all this would have
had to be found over the board, whereas now
Black's task is easy.
33 .if2
•••
Now it became evident that Black is winning.
Or: 32 . . . �h4 33.exd5 �f8 34.\Wb3 \Wg6 (34 . . .
e4 3 5 . lt:J g4!+- This i s a huge difference - see
the variations after the 33rd move below)
35.�f3
They both do not look good for Black.
34.VMb3 <i>hS!?
Not the only way, but in my opinion the
most logical one. White's pieces are tied up
and completely helpless.
King's I ndian Warfare
242
35.exd5 he3 36.YMxe3 hd5 37.tlJ egl gags
Teimour Radjabov - Ilya Smirin
8
Dos Hermanas 200 1
7
This was my first encounter with Teimour
Radjabov. He was 1 4 years old back then
and was not a really big threat yet. Later, as
we all know, Teimour entered the chess elite.
Radjabov himself has played the KID as Black
all his life, and he is a renowned expert in this
opening.
6
5
4
3
2
a
b
c
d
e
f
g
h
l .d4 tljf6 2.c4 g6 3.tlJc3 i.g7 4.e4 d6 s.tlJf3
0-0 6.i.e2 e5 7.0-0 tlJc6 8.d5 tlJe7 9.b4
tlJ hS IO.gel aS l l .bxa5 f5 12.tlJd2 tlJf6
All Black's forces are participating in the
assault - it's impossible to stop it.
38.®h2 e4 39.gdl exf3 40.gxd5 'iMxdS
4L'iMxh6t gh7 42.YMf6t ggg7 43.tlJxf3
43.�f8t �g8 and there is no perpetual.
43 ...'iMxa2t 44.®hl 'iMf7 45.'iMc3 'iMfS
46.®h2 'iMf4t 47.®hl YMg4 48.'iMc6 gxh3t
49.hh3 'iMxh3t so.tlJ h2 YMe6 5 I .tlJf3 'lMg4
White resigned. A good battle!
0-1
So far as in the game Kramnik - Smirin,
which you will find on page 63.
13.f3!?
Vladimir Kramnik chose 1 3 .c5 in this
position. Radjabov's move is not so active, but
is more fundamental. White strengthens the
pawn on e4 first and only after that is he going
to attack on the queenside.
1 3 ... gxa5 14.tlJb3
The knight is not ideally placed on b3 in
such positions, but it helps White to play
c4-c5 quickly and at some point after that it
can be transferred back to d2 and c4 .
243
Chapter 8 - Fighting for the Initiative
14 .. J�� a8 15.c5 f4
Here it goes! Many players (including
myself) love to have attacks going on on
opposite wings of the board.
16.a4 g5 17 ..ia3
The picture is clear - Black's chances lie
only in the kingside attack, but those chances
should not be underestimated. White may
gain the whole queenside and win a lot of
material, but he will lose if gets mated. This is
a primitive explanation, but, as we will see, it
describes well some of the possible variations,
often leading to really wild situations.
The other way to develop the queenside
attack was: 1 8 .cxd6 cxd6 1 9.ctJb5 ctJ e8 20.ctJd2
Now after 20 . . . h5 White again gets the
upper hand after 2 l . ctJ c4 �a6 22.a5 .
I believe Black should instead continue:
20 . . . Ct:J h4! 2 1 .ctJc4 �f6 22.�c l g4 23.fxg4 �g6
with a strong counterattack - the knight on
h4 makes the difference. There may follow:
24.�h l Ct:J f6 25.ctJ bxd6 ctJxg4 26.�fl
17 ... tlJg6
I believe this move serves the attack better
than 1 7 . . . h5 (too many pawn moves?) which
was played in the game Gelfand - Nisipeanu,
Bazna 2009. In my opinion, if White were to
proceed in this position: 1 8 .cxd6 cxd6 1 9 .ctJb5
ctJe8 20.ctJd2! ctJ g6 2 1 .ctJc4 �f6 22.�h l ! ?
b
a
c
d
e
f
g
h
26 . . . Ct:Jxh2! 27.ctJxc8 �xc8 28.�f2 �h6 With
serious direct threats against the white king.
18 ... tlJh4 19.cxd6 cxd6
8
7
6
5
a
b
c
d
e
f
g
h
He would obtain a considerable advantage
- Black's main forces are tied to the defence
of the pawn on d6 and can hardly join the
kingside attack. For instance: 22 . . . Ct:J h4
23.�gl �g6 24.g3! fxg3 25 .�xg3, and White is
basically better on both wings; his knights are
excellently placed and he has plenty of useful
moves, such as '!Wb3 and �ag l .
1 8.a5
4
3
2
a
b
c
d
e
f
g
h
20 ..ib4!?
Radjabov decides to bring the knight from
b3 to c4, and protects the a-pawn so that he
can do that.
244
King's I ndian Warfare
In the annotations to his victory over
Nisipeanu, Boris Gelfand offered here:
20.tLla4!?
The knight invasion on b6 is very unpleasant,
so Black has to show his hand:
20 . . . g4 2 1 . tLl b6
b
a
c
d
e
f
g
20 ... h5
8
7
6
5
4
3
2
a
b
c
d
e
f
All according to plan.
IfTeimour had changed his mind and played:
2 1 .tLla4
Then after:
2 l . . . g4 22.tLlb6
Black has the same resource:
h
2 l . . . tLlxe4!
Mter the simple 2 l . . . gxf3 22.ixf3 tLlxf3t
23.gxf3 (23.�xf3 ig4 24.�c3 �a600)
23 . . . �a6 24.�h l I would probably prefer
White, mainly because of the poor position
of the rook on a6.
22.fxe4
22.tLlxa8 tLlxg2! 23.fxe4 �h4 - see the
similar annotations to 2 1 .tLla4.
22 . . . f3
Again, see the similar detailed annotations
after White's next move.
1
2I .tiJd2
g
h
a
b
c
d
e
f
g
h
22 . . . tLlxe4! 23.fxe4
Basically the only move.
Bad is: 23.tLlxa8 tLlxg2! 24.fxe4 (24.�xg2
�h4 2 5 .fxg4 �f2t 26.�h l tLlg3t 27.hxg3
fxg3 with mate) 24 . . . �h4! 25.tLld2 (25 .�fl
g3) 25 . . . tLl e3 26.�c l g3 27.tLlf3 gxh2t
28.�hl �g3 29.tLlh4 ih3 30.if3 �xh4
23 . . . f3!
23 . . . tLlxg2? does not work: 24.�xg2 f3t
(24 . . . �h4 25 .tLld2 f3t 26.ixf3 gxf3t)
25.�h l �h4 26.ixf3 �xf3 27.�e2+24.g3
24.tLlxa8 tLlxg2 25.�fl tLl f4 and Black is
better even without the rook and knight:
26.ixf3 gxf3 27.�xf3 �g5t 28.�g3 ig4
24 . . . fxe2 2 5 .�xe2 tLl f3t 26.�h l �b8
The resulting position is better for White
(the knight on b6 is very strong and Black
has problems with the defence of the
d-pawn and the development of the light­
squared bishop) , but Black has counterplay
on the kingside, which gives decent
chances.
A possible continuation is:
245
Chapter 8 - Fighting for the I nitiative
And of course not 23.tt:Jb6?? 8:xb6.
23 ... tlJ h7!
Preparing the next sacrifice by vacating the
road to h4 for the black queen. The knight on
h7 will later join the attack via g5 .
24.fxg4
b
a
c
d
e
f
g
h
27.tt:Jd2 tt:Jd4
27 . . . h4 28.8:g2! hxg3 29.8:xg3±
28 .8:g2 8:f6 29.tt:Jdc4 if8 30.8:b l id7 3 1 .ic3
ie8! 32.ixd4 exd4 33.Wxd4 ig7 34.Wd2
ig6
And the pair of strong bishops provides
compensation for the pawn.
A good illustration of the strength of the
attack would be:
24.tt:Jb6 8:xb6!
24 . . . gxf3? 25 .ixf3!
25.axb6 tt:Jxg2! 26.�xg2 Wh4
8
7
6
5
4
21. .. g4 22.tlJc4
Or 22.�h l !? immediately was an option.
22 ga6
3
2
...
8
a
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
23.\t>hl!
A good defensive move. The king hides in
the corner, and the rook will protect g2 from
gl if needed.
b
c
d
e
f
g
h
The assault outweighs the sacrificed rook.
The only defence here seems to be:
27.8:a2
Not 27.ixd6? Wh3t 28.�gl g3 with a
quick mate.
27.8:h l tt:Jg5+
27 . . . g3 28.8:h l ih3t 29.�gl Wd8!
An excellent move, suddenly targeting the
pawn on b6.
30.ia5 g2
30 . . . Wc8 3 Lib4 Wd8=
3 1 .ib5 Wg5
In a practical game, Black has the upper
hand.
The best move seems to be the retreat: 24.tt:Jd2! ?
246
King's I ndian Warfare
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
b
The knight returns to help His Majesty.
a
24 . . . �a8 2 5 . fxg4 hxg4 26 . .ixg4 �g5 27 . .if3
�f6 28.tDa4 and Black's threats are probably
not strong enough.
24 . . . �g5 25 .�g l g3 (25 . . . �f6 26 . .ixa6 bxa6
27.�c l ) 26.�e l and White is j ust winning ­
the opponent's attack has come to a full stop.
The rook on a6 is under fire and Black should
invest some serious material: 24 . . . ttJxg2!
Anyway! More prudent moves would not do.
2 5 .<tt> xg2 �h4 26 ..ixa6 g3 27.hxg3 The best
choice. 27 . . . �h3t 28.<tt> g l fxg3 29.�e2 lDg5!
d
c
f
e
g
h
24 ... llJxg2!
It seemed to me that this sacrifice came as
an unpleasant surprise to Radjabov, who was
very optimistic until this point. Now Black's
forces get close access to the white monarch.
A knight is not a high price for that.
25.�xg2
The silicon creature prefers 25.�gl !? - a
move that would hardly come to mind over
the board. It is probably right: after 25 . . . f3
26 . .ixf3 ttJ f4 27.gxh5 White would have an
edge in a very complex position.
25 ... hxg4 26.�hl
Teimour had a tough choice between this
move and:
26 ..ixg4 �g5 27.h3
a
b
c
d
e
f
g
h
And suddenly a drawing mechanism is
created, for instance: 30.�g2 �h4 3 1 .�h l
tD h3t 32.<tt> g2 ttJ f4t 33.<tt> g l tD h3t 34.<tt> fl
g2t 35.�xg2 ttJ f4 36.�f2 �h l t 37.�g l �h4
38.�f2=
a
27 . . . f3t!?
b
c
d
e
f
g
h
247
Chapter 8 - Fighting for the I nitiative
Weaker is 27 . . . t2J f6 28J!hl b5!? (another
sudden computer move) and the computer,
as so often, gives equality at the end of a
complicated line.
28.<±>f2 .ixg4 29.hxg4 .ih6!
The King's Indian bishop has joined the
attack.
30.!!h l
The only move.
30 . . . ttJ f6 3 1 . <±>fl
During the post mortem, Alexey Dreev
proposed 3 1 .�g l here, but after 3 1 . .. !!c8
(3 I . .. t2Jxg4t ? ? 32.�xg4 �xg4 33 .!!ag l was
the idea) 32.!!xh6 �xh6 33.�g3 !!xc4 34.g5
�g6 35 . gxf6 �xg3t 36.<±>xg3 !!xb4 Black is
winning.
3 1 . . . t2Jxg4
Even Houdini finds it difficult to repel the
attack - White is in great danger. So I think
Teimour made the best move in the position.
26 ...'1Mh4
8
7
27.!!a2 g3 28 . .if3 ltJg5 was hardly better.
27... �g5!
Of course not 27 ... g3 ? 28.t2J f3 .
28J�gl
The rook could be taken, but not safely:
28 . .ixa6 g3 29.!!e2
8
7
6
5
4
3
2
b
a
c
d
e
f
g
h
After 29 . . .f3 White replies 30.�gl ! g2t
3 1 .!!xg2 fxg2t 32.�xg2 !!f2 33.�g3 �xg3
34.hxg3 bxa6 3 5 .!!a2 then, almost by force,
there arises a dynamically balanced endgame.
But instead after 29 . . . .ig4! 30.�g l !!f6 more
forces join the assault. Despite Black's huge
material losses - rook and knight - White
would be on the verge of collapse; a beautiful
variation.
6
5
4
3
8
2
a
b
c
d
e
f
g
h
Black's pieces are starting to rendezvous with
the white king, and the pawns on g4 and f4 are
leaning dangerously toward it.
27.�d2
This knight quits its aggressive plans
(attacking the d-pawn) and steps back into
defence.
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
248
King's I ndian Warfare
28 .. � h3
.
I could not foresee that after:
28 . . . g3 29.:gg2
29. 4J fl :gf6+
29 . . . ih3 30.�g l :gf6 3 1 .ixa6 ixg2t
32.�xg2 bxa6
The attack would be very strong and
more than enough to compensate for the
opponent's extra knight. The best option for
him here seems to be:
33.lDe2
33.:gfl ? :gh6 34.4Jf3 lDxf3 3 5 .:gxf3 �g4-+
33 . . . :gh6 34.ltJxg3 fxg3 3 5 . 4J fl
Obviously Black is on top. But the move I
made is probably not worse.
attack, but I was hoping that Black's threats
would be very strong. Intuition did not let
me down:
3 1 .�e l
3 l .:ga2 :gh6 32.�e 1 g3 33.lDf3 �h5 34.We2
ig4 35 .Wd3 lDg5 36.lDxg5 Wxg5 37.:gg2
Wh5 38.Wc2 ih3 39.4J d l :gg6 This line
looks pretty forcing and Black is winning at
the end.
3 l . . . g3 32.lDf3 Wh5 33 .We2 ig4
29J�fl !U6!
8
7
6
a
5
b
c
d
e
f
g
h
34 .Wg2 :gh6+ 35 .:ga2 lDg5 36.ltJxg5 Wxg5-+
White is defenceless - the extra rook is small
consolation and Black's excellently coordinated
army prevails.
4
3
2
30...�h5?
1
a
b
c
d
e
f
g
h
The rook calmly joins the attack. Black's
pieces continue to concentrate themselves,
without hurrying, around the white king. I
was really excited by the possibility of making
"quiet" attacking moves while being a piece
(and potentially a whole rook down) .
A pity. This mistake spoils the impression of
the game a little.
Black should have acted resolutely: 30 . . . g3
3 1 .lDf3 Wg4 (32 . . . g2# is the threat!) 32.�g2
:gh6-+, with the very nice point: 33.hxg3
30.�el
Of course,
could not calculate all the
variations after:
30.ixa6 bxa6
Grabbing as much material as possible in the
hope of giving some of it back to repel the
a
b
c
d
e
f
g
h
249
Chapter 8 - Fighting for the Initiative
33 . . . ttJ f2!! 34.Wxf2 Wh3t 3 5 .@gl Wh l # It
would be great to fi n ish the game in such a
wonderful way, but alas I failed to spot this
idea.
37 . . .'\Wxh2t! Pretty - instead of taking the
rook, Black sacrifices the queen. 3 8 .�xh2
�xh2t 39.@gl tD h3t 40.@f1 g2t-+
33 . . ..id7 34.�c6! ?
31.gcl
White has gained some time to organize the
defence.
3 1 . .. gh6
8
7
6
5
a
4
3
2
1
b
a
d
c
e
f
g
h
32.t!Jdl?!
This time it was absolutely "legal" to take
the rook, which remains under fire for several
moves:
32 . .ixa6 bxa6 33.tDe2!
33.tDdl .id7 34.ttJc4 g3-+
33.tDb5 .id7 34.ttJxd6 g3 3 5. ttJ f5 .ixf5
36.exf5 tDg5 37.�f2
b
c
d
e
f
g
h
34 . . . tD g l !
The only move.
3 5 .�f2 ttJ h3
White can take a draw by means of:
36.�fl
Or he may look for something else.
In this variation we see that after my
mistake on move 30, the attack lost a great
deal of speed and sharpness. I hope that these
somewhat numerous variations have not bored
the reader: most of them are really nice and
they are needed to understand what is going
on on the board.
32 .. J�a8
The rook is alive!
33.t!Jfl £3!
Not falling into the small trap set by my
young opponent: 33 . . . tDg5? 34.�xc8t! �xc8
3 5 . tD xg4
34.t!Jxh3 %Vxh3 35.gfl g3!
a
b
c
d
e
f
g
h
250
King's Indian Warfare
tLlg5. In this case, after the logical 38 .. J�f6
(not 38 . . . ig4? 39.tLlg5 �f8 4o:�xf8t ixf8
4 1 .lt:Jxh3 ixe2 42.tLlg5) 39.�gl there arises a
position of dynamic balance - the black king
might become vulnerable.
3S ..J�xd6 39.�g5 '1Hd7 40 ..ig4 �e7!
This was missed by Radjabov. The rest is
easy - Black's material advantage decides the
matter.
a
b
c
d
e
f
g
h
These two attacking pawns cost White a
whole rook.
36.�xf3 gxfl 37.'1Hxfl
The smoke has cleared. Black is left with a
minimal material advantage (the exchange
for a pawn) , but the white pieces are active
enough, and objectively the position is about
equal. We were both in slight time trouble,
and I think I overestimated Black's chances
during the game.
41 .hcS �xg5 42J�gl �f6 43.�£5 gxa5
44.�c2 gas 45.�c7 b5 46.'1Hb7 gadS
47.�xb5 gb6 4S.�c5 gdbS 49.gg6 �xg6
Not a necessary sacrifice. 49 . . . �b l t 50.<i>g2
�h4 5 l .�xg7t <i>xg7 5 2.�c7t <i>f8 53.�d6t
�e7 54.�h6t �g7t would win easily, but I
wanted to finish the game in an attacking
spirit.
50.hg6 gxg6 51.�c2 .ih6 52.h4 grs
53.�c4 g£4
8
7
6
37 .. J�g6
5
4
3
2
1
a
a
b
c
d
e
f
g
h
3S ..ixd6?
A big oversight. If Teimour had realized he
was OK, he would probably have found the
strong continuation 3 8 .id2!, with the idea
b
c
d
e
f
g
h
Two rooks and a bishop create a mating net
around the king.
54.d6t ®hs 55.�eSt ggs 56.�h3 gxe4
0-1
I have to admit that I love this game very
much, and think of it as one of the best in my
career.
25 1
Chapter 8 - Fighting for the Initiative
Jaan Ehlvest Ilya Smirin
-
Khanty-Mansiysk 2009
The game was played in the first round of the
World Cup in Khanty-Mansiysk. This northern
Russian city has become one of the world's
biggest stages for high-calibre chess events
in recent years. In 200 1 Jaan Ehlvest (one of
the strongest player in the world in the late
1 980s) beat me in the equivalent tournament
in Moscow. That third-round defeat was very
painful for me - the level of games was pretty
good, I was leading in the course of the match,
but I lost in an Armageddon blitz game. Now,
after eight years, I was eager to have my revenge.
Shirov and Korchnoi. The main difference
between the two knight moves is that from e8
it will defend the d6-pawn and the important
square c7, after White plays c4-c5 and c5xd6,
whereas on d7 the knight makes it harder for
White to play c4-c5 and it sometimes controls
e5 in the event that White decides to play the
double-edged f2-f4.
I .ltJf3 ltJf6 2.c4 g6 3.ltJc3 ig7 4.e4 d6 5.d4
This was the second of two games in the
classical time control mini-match. In a
knockout tournament, the loser goes home.
I had won the first encounter with White, so
Jaan had to win to stay in the match. Perhaps
the KID is not the most common opening
choice in a situation like that, but I decided
to stay with my favourite opening (anyway, I
am not very good at making a draw "to order"
with Black) .
5 ... 0-0 6.ie2 e5 7.0-0
The Classical System. This time I chose the
most principled continuation.
7 ... ltJc6 8.d5 ltJe7 9.ltJel
I did not expect this move from Ehlvest.
Usually he had played 9.4Jd2 or 9.b4. But
his choice is easy to understand - after 9 . 4J e 1
there usually arises a sharp, double-edged
situation which perfectly suited my opponent,
considering his must-win situation.
b
a
d
c
e
f
g
h
IO.t2Jd3
Ehlvest goes for a more positional approach.
Victor Korchnoi usually played the sharper
line starting with 1 o.�e3 .
For example, the following interesting game
of his was mentioned earlier: 1 o.�e3 f5 1 1 . f3
f4 1 2.�f2 h5 1 3 .c5 g5 1 4.a4 4J g6 1 5 .a5 �h6
1 6.b4 �h7 1 7.cxd6 cxd6 1 8 .4Jb5 g4 1 9 .fxg4
hxg4 20.�xg4 �xg4 2 1 .Wfxg4 4J f6 22.Wff3
Wfd7 23.ltJxa7 ttJ g4 24.b5 tDxf2 25 .Wxf2 f3
26.4Jxf3 Wfg4 27.b6 Wfxe4
9 ... ltJe8
9 . . . 4Jd7 is a more common option here.
The move in the game became popular after
Carry Kasparov played it twice in 1 992 against
a
b
c
d
e
f
g
h
252
King's I ndian Warfare
28.Ei:fe l (28.@h 1 ! would have given White
a big advantage) 28 .. .'�xd5 29.Wic2 Ei:ae8
30.Ei:ad 1 Wffxa5 3 1 .'Wic7t Ei:e7 32.Wixd6 e4
33 . '2Jh4 Ei:g7 34.'2Jxg6 Ei:xg6 3 5 .Wffe7t i.g7
36.Wih4t @g8 37.Wffxe4 Ei:xb6 3 8 .Wid5t
Wffxd5 39.Ei:xd5 Ei:b2 40.Ei:fl i.h6 4 l .Ei:xf8t
@xf8 42.Ei:f5t @e? 43.'2Jb5 i.e3t 44.@fl
Yz-Yz Korchnoi - Kasparov, Debrecen 1 992.
10 ... £5 1 1 .�d2
After l l .f4 (quite a logical move: as we
have already noted, the knight on e8 does not
control e5) I was going to continue l l . . . exf4
1 2.'2Jxf4 c6! ? with a complex struggle in which
Black, in my opinion, has decent counterplay
- he undermines White's not yet very strong
centre from both flanks.
I think that 1 1 . f3 is a slightly more precise move
order compared to the game continuation.
Now 1 1 . . . fxe4 1 2.fxe4 makes no sense, so
Black should proceed l l .. . f4 1 2 .c5 g5 with the
usual attacks on opposite sides of the board.
1 1 . .. fxe4
I decided to make the game simpler from the
strategic point of view. After all, my main goal
was not to lose.
In my optnton, control over the central
square e4 gives White hope for a definite
positional advantage, but Black does not have
much to complain about. Indeed, the bishop
on g7 is slightly passive, but all the other pieces
are doing well. In particular, the knight on e7
has access to f5 and, potentially, to d4 . Black's
pawn structure is solid enough - in other
words, White's advantage is minimal, if any.
1 3.�xf6t � 14.�g4
To my great surprise, I discovered another
game by Ehlvest in which he had played this
position as Black 29 years prior to our game!
1 4.i.c3 't:J f5 1 5 .i.f3 i.g7 1 6.Ei:e 1 @h8 1 7.Ei:c l
Ei:f7 1 8 .ig4 tt:Jh4 1 9 .ixc8 Ei:xc8 20.Ei:e4 g5
2 l .c5 b6 22.cxd6 cxd6 23 .i.d2 h6 24.Ei:ec4
Ei:xc4 25 .Ei:xc4 b5; (Yz-Yz, 34) Dydyshko Ehlvest, Tallinn 1 980. I doubt though that
Jaan remembered that during our encounter.
14 ... �£5
Of course the exchange o f light-squared
bishops would be in White's favour - at least
now, when the black pieces are not yet active.
1 5J�e1 �d7
1 1 . . . '2J f6 1 2.f3 f4 is the main continuation
here, with the usual attacks on opposite wings.
8
7
6
5
4
3
a
2
b
c
d
e
f
g
h
16.�c3
1
a
b
c
d
e
f
g
h
Here or on the next move 1 6.c5 deserved
attention - White is trying to develop activity
253
Chapter 8 - Fighting for the Initiative
on the queenside, but there is the small
drawback that the d5-pawn may become weak
in some variations.
23 ...�g6 24J!el �af8 25.�d3 �g7
Tension is growing. All of Black's pieces are
concentrated on the kingside. They are just
waiting for orders.
16 ... a5 17J!cl b6
It's not easy for White to obtain active play,
so both sides begin manoeuvring, trying to
steadily improve their position.
1 8J!c2 �f7 19.b3 �f8 20.�e4
Ehlvest could have changed the strategic
picture by means of 20.f4!?, but after 20 . . .
exf4 2 l .ctJxf4 ixc3 22.!hc3 Wig? 23 .�d3 �af8
24.ctJ e6 ixe6 25.dxe6 �e7 Black should be
more or less OK - White's main hope - the
pawn on e6 - is firmly blockaded.
8
7
6
26.a3 g4!
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
5
Here we go! The real action starts now.
This push required calculation and, more
importantly, some guts.
3
27.fxg4 .ih4!
4
2
This is the point.
a
b
c
d
e
f
g
h
28.�e4
Or 28.gxh5 Wlxh5 29.�e4 ltJd4 would lead
to the position in the game by transposition.
20 ...�g7!?
Black prepares to start some play on the
kingside. The queen vacates f8 for the a8-rook
and at the same time from g7 it will support
the future advance of the g- and h-pawns.
After 28.g3 hxg4 29 .ig2 (29 .ctJxg4? ctJ xg3
30.Wixg6 �xg6 3 l .hxg3 ixg4) 29 . . . ie7
30.ctJe4 Wih5 a complex, slightly boring, and
approximately equal position would arise.
21.f3 h5 22 ..ih3 g5!
Of course, bad was 28.gxf5 ixf5 and the board
geometry works in Black's favour - 29.ctJe4
ixh3 with a huge advantage.
Following the plan.
23.�fl
A good prophylactic move. White stalls
. . . g5-g4, and from f2 the knight is ready to
jump to e4 at the appropriate moment.
28 ... �d4!?
Black had a good tactical alternative which I
did not see: 28 . . .ixf2t 29.!hf2 ctJ g3 ! 30.�xf8t
�xf8 3 1 .hxg3 hxg4 32.ixg4 ixg4, and White
254
King's I ndian Warfare
can forget about trying to win this game. But
objectively my move was no less strong.
34J�ee2 gf3!?
There was no special need for this spectacular
move.
29.gxh5 YMxh5
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
Black is very active for j ust one small pawn,
and now starts a forcing line.
30 ..ixd7 tiJf3t 3 1 .<i>hl .ixfl 32 ..ie6t <i>hs
33.h3
A little joke - 33.�h3 ?? �xh3 34.gxh3 �gl#
33... tlJg5!
8
After the simple: 34 . . . ttJxe6 3 5 .�xf2 (bad
is 3 5 . dxe6? �xg2! 36.e7 [36.@xg2 �g5t
37.@h2 �g l #] 36 . . . �f6 and Black is winning)
3 5 . . . ttJ f4 36.�f3 �h7 37.�e3 �g6, despite
being a pawn down, Black would have a very
safe (in terms of not losing the game) and
active position - White's king is in constant
danger and all his pieces (especially the bishop
on c3) are rather passive.
But again, the move made in the game is
not objectively weaker - it just leads to a more
complex and unpredictable struggle.
35.'iMxf3!
This queen sacrifice is the only option.
35.gxf3? ttJxh3 36.�g4 �h4 would have led to
immediate disaster.
35 ... ttJxf3 36.gxfl tlJ h4
On h4 the knight fulfils two functions: it
protects the vital f5-square from the white
bishop, and applies pressure to g2 .
8
7
7
6
6
5
5
4
4
3
3
2
2
1
a
b
c
d
e
f
g
h
This strong move clarifies the situation Black is fine.
a
b
c
d
e
f
g
h
Black has a minimal material advantage,
neither king is very comfortable, and both
sides' pieces are fairly active. The verdict: there
255
Chapter 8 - Fighting for the I nitiative
is a dynamic balance on the board. But good
moves are required . . .
37J�ce2
The brave 37.g4 �g5 38 .id2 �g6 39.Ei:f8t
@h7 40.Ei:f5 @h8 4 1 .Ei:f8t @h7 would lead
to a draw, which would not have suited Jaan
anyway.
40 ...'l«e4?
I saw that this move would lead to the
desired draw by force, and went for it, reaching
my goal of advancing into the second round of
the World Cup.
However, 40 . . . �b l 4 1 .Ei:e2 (4 1 .8:fl �e4)
4 1 . . . Ei:xg2 would have led to a simple forced
win. It happens . . .
37 ...V«g6 38 ..iel Wh7 39JU8 'l«d3
41 ..id2 �xg2 42 ..if5t �xf5 43.�xg2 � h4
8
8
7
7
6
6
5
5
4
4
3
3
2
1
a
b
c
d
e
f
g
h
40J�ef'2?
The losing error.
Correct was 40.ixh4! �xe2 4 1 .if5 t Ei:g6
42.8:f7t @h6 43.g4! and the threat of 44.g5t
@h5 45 .Ei:h7t would force Black to give
perpetual check: 43 . . . �f3t 44 .@h2 �e2t
45 .@h l �f3t=
2
1
a
b
c
d
e
f
g
h
44.�f7t whs
And due to the coming perpetual, we agreed
to a draw. Despite the missed win, this game
brought me great creative pleasure.
•!2-•12
256
King's I ndian Warfare
Josh Manion - Ilya Smirin
Las Vegas 1 997
This game is another Las Vegas "masterpiece".
But speaking seriously, it seems that the centre
of the world gambling empire used to inspire
me occasionally to creative and somewhat
risky play.
I .d4 �f6 2.c4 g6 3.�c3 .ig7 4.e4 0-0 5.�8
d6 6..ie2 e5 7.0-0 �c6 8.d5 �e7 9.b4 � h5
10.�d2
This move became well known after the
famous game Petrosian - Gligoric, Rovinj /
Zagreb 1 970. In my opinion, the modern
fashion for 1 O.El:e 1 or 1 O.g3 looks more logical.
It looks like White has changed his mind
and decided to keep this bishop from a
possible exchange for the knight on f4. In
my opinion 1 2.f3 or 1 2 .cS deserved serious
attention.
12 ... g5
1 2 . . . @h8!?, with the rather well-known idea
of transferring the knight from e7 to f6 via g8 ,
has been played a few times here.
13.exf5 �xf5 14.g3
Now there is a choice.
14 ... �h3t
Instead 14 . . . tt:Jd4 was played in the above­
mentioned encounter: 1 S .gxf4 ltJxf3t
1 0 ... � £4 1 1 .a4
The drawback of 1 0.ltJd2 may be seen after
the immediate 1 1 .if3?! ltJ d3 1 2 .ia3 aS! and
Black is fine.
1 1 ... £5
This is clearly better than the possible
continuation 1 1 . .. ltJxe2t 1 2.'\Wxe2 f5 1 3 . f3
f4 1 4.cS - White's queenside attack develops
more quickly compared with the opponent's
play on the other wing.
8
a
b
c
d
e
f
g
h
In one of my games 1 6.ltJxf3 g4 l ?.@h l
was played. I managed to draw it despite being
somewhat worse almost all the way (perhaps
this is a reason why I deviated against Manion?)
- 17 . . . gxf3 1 8.�xf3 ifS 1 9 .El:g 1 ig6 20.id2
�d7 2 1 .ltJe4 El:ae8 22 .ltJgS exf4 23 .ic3
ixc3 24.�xc3 El:eS 2S.ltJe6 El:ffS 26.El:g4 �f7
27.El:ag 1 f3 28.aS c6 29.�d4 cxdS 30.ltJgS
�f6 3 1 .h4 �e7; (Yz-Yz, 46) Kourkounakis Smirin, Gausdal 1 990.
7
6
5
4
3
2
1
a
12 ..i8
b
c
d
e
f
g
h
1 6.�xf3 g4 1 7.�h 1 exf4 1 8 .ib2 ifS 1 9 .El:fe l
f3 20.ltJde4 �h4 2 1 .h3 ieS 22.El:e3 gxh3
23.�xf3
257
Chapter 8 - Fighting for the Initiative
I do not agree with that evaluation.) White's
achievements are excellent piece coordination,
control over the important square e4, and the
currently rather poor bishop on g7.
17.tlJb3 YHe7
The queen changes its location for a more
natural one.
17 . . . \Wf7 was a good alternative in that sense.
b
a
c
d
e
f
g
h
23 . . . .ig4 24.\Wh l h2t 25 .�g2 \Wh5 26.l2J d2
.id4 27.\We l Ei:ae8 28.l2Jce4 .ixb2 29.Ei:g3
.ie5 30.Ei:aa3 �h8 3 1 .�h l Ei:g8 32.\Wfl
.ixg3 33.Ei:xg3 El:xe4 0- 1 Petrosian - Gligoric,
Rovinj /Zagreb 1 970.
1 5.<i>g2 'iHd7!? 16 ..ie4
The idea that 1 5 . . . \Wd7 is connected with
may be seen in the variation 1 6.l2Jde4 l2J d4
1 7 . .ih5 ? g4, intending . . . \Wf5 , and the bishop
on h5 does not feel good.
1 8.'iHd3 h5
1 8 . . . l2J g5 1 9 ..ixg5 \Wxg5 20.c5 was played
in one of my games. I would probably prefer
White here - his play looks very easy, whereas
it's not so clear which plan Black is going to
follow.
19.c5
8
7
6
5
4
3
2
a
a
b
c
d
e
f
g
h
16 ... g4
I did not trust the resulting position after the
forcing line 1 6 . . . l2Jxf2 1 7. �xf2 l2Jxg3t 1 8 .�gl
ltJxfl 1 9. l2Jxfl . Now it seems to me that Black
is not worse - after all he has some material
advantage and the king on gl is slightly open.
(The artificial brain is overoptimistic here,
claiming a big edge for Black in that line -
b
c
d
e
f
g
h
19 ... tlJf4t!?
Not a necessary move, of course, bur I was
in "sacrifice mode". The decision was purely
intuitive, in the style of the young Mikhail Tal.
Just one thing was clear to me: my opponent
would have to solve some complicated
problems and I hoped he would not quire be
able to cope with that. Moreover, the position
after the sacrifice looked very likable to me . . .
20.gxf4 exf4
258
King's Indian Warfare
The compensation for the knight is obvious
- the white king is under attack and the
opponent's well-coordinated forces give him
a serious headache. Black's main threat is
probably . . . f4-f3, after which the white king
may easily fall into a mating net. Manion
decided to put an automatic stop to the
f-pawn.
As I remember, I did not even consider
the strongest and rather simple: 2 l . . . gxf3t
22..ixf3 lDh4t 23.c;!;>h l .ifS!
8
7
6
5
21.f3?
4
Oops! A serious mistake which I failed to
exploit. It was not an easy task to find the best
solution over the board.
3
Obviously 2 l ..ixf4? lDh4t does not work.
The best decision seems to be 2 I .lDd2!,
bringing the knight to help His Majesty.
8
7
2
a
b
c
d
e
f
g
h
After which White would be on the verge
of losing, for instance: 24 . .ie4 .ixc3 25 . .ixf5
�xf5 26.�g l t c;!;>h? and 24.Wfd2 Wfe3! 25 . .ib2
�ae8 26.Wfxe3 �xe3 27.lDd2 .ixc3 28 . .ixc3
�xc3 29 . .ixh5 f3 and so on.
22.ga2!
6
The rook plays a big part, protecting the
whole second rank, especially the pawn on h2.
5
4
22...YMg7
3
2
a
b
c
d
e
f
g
h
After 2 1 . . .'\Wh4 (this continuation is asking
to be played) 22.�g l a5 ! ? 23.bxa5 dxc5 there
would arise a position that is very hard to
evaluate, even with computer help. Still, my
soul belongs to Black here. By the way, this
line shows that the sacrifice on the 1 9th move
was objectively OK.
21. ..�e5?
As I said earlier, I liked the structure of
the black piece and pawn formation, and
subconsciously I did not want to change it.
If 22 . . . ltJ h4t 23.@h l g3 White would
have at his disposal a strong tactical decision:
24.ixf4! ixf4 25.hxg3 ixg3 26.�g l lDf5
27.�h2 h4 28.lDe2 giving a piece back, but
taking over the initiative and in turn creating
strong threats against the king on g8 .
23.tlJdl
Not a bad move, but I think 23 .ixf5
ixf5 24.ltJe4 should have been preferred.
After 24 . . . �ae8 (24 . . . gxf3t? 25.c;!;>h l Wfh7
26.lDbd2+-) 25 .c;!;>h l @h8 26.�gl g3 27.�e2
h4 the position would remain very complex.
23 ... g3 24.ggl gxh2t 25.�xh2 tlJg3
259
Chapter 8 - Fighting for the Initiative
8
the c8-bishop interposes and the attack comes
to a full stop.
7
6
5
8
4
7
3
6
2
5
1
a
b
c
d
e
f
g
h
This knight has survived a possible exchange
and is awarded the right to occupy a strong
outpost very close to the enemy king.
26J3gg2
White must be very careful all the time. If
he were to play the innocent-looking 26.ib2?!
then after 26 ... ixb2 27 Jhb2 �g5 he would
have to find 28.�gg2! (28 . ctJ f2? �h4t 29.�g2
�f7 leads to disaster) 28 . . . �h4t 29.@gl �h l t
30.@f2 ih3 3 l .�xg3t fxg3t 32.@xg3 to stay
in the game.
26 ...�g5 27JtH2 <!>hS!
The king moves off the g-file, freeing the
queen and knight. At the same time, g8 is now
vacated for the black rook.
28.tlJd4
Also now if 28 .ib2 �g8 29 .ixe5t dxe5
30.@gl �h4 (not 30 . . . ctJ e2t? 3 l .@f1 �xg2t
32.@xe2 ih3 33.ctJd2) 3 l .�c3 �g5, with the
idea . . .ih3, Black would be happy.
4
3
2
1
a
c
d
e
f
g
h
30.tlJe6?
Now this does not have a lethal effect, but
instead becomes the main reason for White's
imminent defeat.
Instead, my opponent would be better off
playing the curious variation 30. ctJ f5 ! ctJ xf5
3 l .�xg8t @xg8 32.ixf5 �g5 t 33.ctJg4 hxg4
(33 . . . ixf5 ? ? 34.�xf5) 34.ixg4 ixg4 3 5 . �g2
�h6 36.�xg4t �h8, which would lead to
approximate equality.
30 ....ixe6 3 1 .dxe6 gg7
8
7
6
5
Or 28.cxd6 cxd6 29.�c2 �f7! would lead to a
position with some kind of dynamic balance
(but not 29 . . . �g8 ? 30.�c7+-) .
4
28 ...�h4t
1
It would be a huge mistake to play 28 . . . �g8 ?
29.ctJe6 �h4t 30.ctJh3 immediately -
b
3
2
a
b
c
d
e
f
g
h
260
King's I ndian Warfare
32 ..tb2?
33.he5 dxe5 34.tlJg4
The final and decisive error, but objectively
speaking it was nearly impossible to find the
only defence in time trouble.
For example, 32.cxd6 would lose immediately
after: 32 . . . �ag8! 33.lt:Jg4 '.Wh l t 34.<i>f2 hxg4
3 5 .d7 .id4t 36.'1Wxd4 '.Wfl #
Only 3 2 ..if5 ! dxc5 3 3 . .ig4! would allow
White to continue fighting, even though in
a worse position after 33 . . . �d8 34.�h2 '.We7
3 5 .'1Wb3 <±>g8.
Or 34.e7 lt:J e2t 35.<i>fl �xg2 36.!he2 �g l#.
34.�xg3 �xg3t 3 5 .<i>fl �g l t 36.<±>e2 '.Wxf2t
37.<±>xf2 �8g2# Pretty, isn't it?
Houdini's first suggestion is 34.'1Wd7 �xd7
3 5 .exd7. This basically says everything about
the position after Black's 33rd move.
34 ... �hlt 35.\t>fl hxg4 36.fxg4 �xg4
37..if3
8
7
32 .. J!ag8
6
8
5
7
6
4
4
2
3
5
3
2
a
1
a
b
c
d
e
f
g
h
Now all the black pieces are taking part in
the attack. Their combined effort is irresistible.
b
c
d
e
f
g
h
37... e4 38.�d4t �4g7 39 ..txe4
While making this move, Josh Manion
overstepped the time limit, so the checkmate
39 . . . '\Wfl # was not executed on the board.
0-1
26 1
Chapter 8 - Fighting for the Initiative
Boris Gelfand - Ilya Smirin
Saint Vincent 2005
The chessplayer I have played most often in
my life is Boris Gelfand. Our first game dates
back to 1 979! We have also been teammates
plenty of times - first for Belarus and after that
for the Israeli national squad. Once we even
teamed up in the Russia - Rest of The World
match. This game from the European Club
Championship in Saint Vincent was around
our 30th meeting.
8 . . . a5
Better is 8 . . . ttJ g4 9 .ig5 f6 1 0 .id2 f5 .
9.tlJd2 tlJe8 1 0.0-0
1 O.g4 tlJc5 l l .h4! leads to a very promising
position for White.
IO .. .f5 l l .exf5 gxf5 1 2.f4 ih6!? 1 3 .t2Jdb l
exf4 1 4.ixf4 ixf4 1 5 .�xf4 t2J e5 1 6.�d4
c5 1 7.dxc6 bxc6 1 8 .t2Ja3 t2J g7 1 9.�f2 t2Je6
20.�d2 �a7 2 I .tlJc2 �g7 22.�d l �h4
23.�dfl �f6 24.�h l ttJg5 25 .�d4?
I .d4 �f6 2.c4 g6 3.�c3 �g7 4.e4 d6 5.�£3
0-0 6 ..ie2 e5 7..ie3
Also possible is: 7.0-0 t2J bd7 8 .'\W c2 c6
9.�d l �e7 I O .d5 c5 l l .g3! ? �h8 1 2.ig5!
a6 1 3 .t2Jh4 �e8 1 4.�e l tlJ g8 1 5 .id2 �e7
1 6.t2Jg2 f5 1 7.exf5 gxf5 1 8 .f4 e4 1 9 .g4!?; ( 1 -0,
5 1 ) Korchnoi - Romanishin, Brussels 1 986.
a
7 ... �bd7
b
c
d
e
f
g
h
25 . . . �xh2t!! 26.�xh2 �h6t 27.�g3
27.�g l t2J h3t 28.�h2 t2Jxf2t 29.�gl
t2J h3t 30.�h2 t2J f4t 3 l .�gl �xg2#
27 . . . t2J e6t 28.�g4 t2Jxg4 29.ixg4 �xg4t
30.�f3 �xc4 3 l .�e3 f4t 32.�d2 �g6 33.�el
�c5 34.t2Je4 �e5 3 5 .�cl ia6 36.t2Jxd6
�xe l t 37.t2Jxe l t2Jd4 0- 1 Kantsler - Smirin,
Jerusalem 2005 .
8 '1Ne7
•••
a
b
c
d
e
f
g
h
8.0-0
This position may arise after a transposition
ofWhite's 7th and 8th moves
8.d5
This is also a logical continuation here. It
happened in the following game, in which
there occurred an elegant queen sacrifice:
A relatively flexible continuation - the
central pawn structure is not yet fixed.
A couple of times I have played 8 . . . c6 9.d5
c5 , but I did not particularly like the course of
those games.
Once I tried the rather experimental but
recently popular continuation:
8 . . . �e8! ? 9.d5
According to the usual standards of the KID,
the rook is placed poorly on e8 (it does not
262
King's Indian Warfare
support . . . f7-f5) , but sometimes there are
more flexible rules nowadays . . .
9 . . . tLl h5 1 0.g3 ifS!?
It went well:
1 1 .tLl e 1 tLlg7 1 2 .tLld3 f5 1 3 . f3 ie7 1 4 .\Wd2
lLl f6 1 5 .c5 fxe4 1 6.fxe4 tLlg4 1 7.ixg4 ixg4
1 8.�ac l ih3 1 9.�fe 1 �f8 20. tLl f2 id?
2 1 . tLl d3 tLle8 22.ih6 �f7 23 .�fl tLl f6 24.cxd6
cxd6 2 5 . tLl f2 �c8 26. tLl e2 �xc l 27.tLlxc l \Wb6
28 .ie3 \Wa6 29.�e 1 id8
With excellent play; (0- 1 , 5 1 ) Colovic Smirin, Skopje 20 1 4 .
9 ...c6 IOJ�adl h6
The immediate 10 . . . tLlg4 1 l .ig5 f6 1 2.ic l
did not appeal to me; hence this prophylactic
and somewhat weakening move.
In this game Boris managed to cast serious
doubt on the strategic correctness of this plan,
so I propose playing a different, more open,
type of position instead: 10 . . . exd4!? 1 1 .tLlxd4
�e8 1 2.f3 tLle5 1 3 .�fe 1 a6 1 4.ifl ( 1 4.\Wd2!?)
1 4 . . . c5 1 5 .tLlb3 ie6 1 6.tLld2 \Wc7 1 7.i>h 1
�ab8 1 8 .b3 id? 1 9 .h3 tLlc6; (0- 1 , 60) King
- Kveinys, Germany 1 993.
9.'1Hc2
If 9.d5 I was going to proceed with 9 . . . tLle8,
preparing ... f7-f5 , as happened a few times in
the games of GM Kochyev.
9.dxe5 dxe5 1 0.tLld5 \Wd8 was played three
times against me.
l l .d5 c5 12.g3!
8
7
6
5
4
3
2
1
a
a
b
c
d
e
f
g
h
managed to win all three games. Here
is one of them: 1 1 .\Wc2 tLl g4 1 2.ig5 f6
1 3.id2 c6 1 4.tLle3 lLlxe3 1 5 .ixe3 f5 1 6.exf5
gxf5 1 7.�ad 1 \We7 1 8 .ig5 \We8 1 9.c5 e4
20.tLld2 tLl e5 2 1 .f3 \Wh5 22.if4 ie6 23 .�de 1
\Wg6 24.fxe4 fxe4 25 .\Wxe4 ixa2 26.\Wb4
id5 27. tLl f3 \Wg4 28.id2 lLl xf3t 29.ixf3
id4 t 0- 1 Bercys - Smirin, Philadelphia
20 1 1 .
The move made by Gelfand seems to be the
most principled one.
b
c
d
e
f
g
h
A deep positional approach. White is going
to put his knight on the ideal square h4, after
which it will be hard for Black to obtain his
customary kingside initiative by means of
. . . f7-f5 . Meanwhile White will have a free hand
to develop his usual activity on the opposite
wing. I think this idea, in a similar position,
was introduced to practice by Viktor Korchnoi
in his game against Oleg Romanishin (given in
the note to White's seventh move) . At least this
is one of the first such examples.
12 .. .lbh7!?
I quickly realized the danger of being
positionally outplayed and found this
Chapter 8
-
263
Fighting for the Initiative
manoeuvre. It has its drawback, but apparently
Black had nothing better.
13.�h4?!
Boris follows his plan. During the game I
saw 1 3 .'tWd2!, attacking the pawn on h6.
I was going to sacrifice a pawn with 13 . . . f5 ?!,
but after: 1 4.exf5 gxf5 1 5 .ixh6 ttJdf6 (or
1 5 .. .f4 is met by 1 6.ixg7 'tWxg7 1 7.<±>h l 'tWh6
1 8J�J:g l <±>h8 1 9 .g4) 1 6. ttJ h4 Black's position
would be nearly hopeless - his king is weaker
than White's, and the missing pawn does not
add optimism.
But instead the best move 13 . . . h5 would mean
a kind of strategical failure after 1 4 . ttJ h4 if6
1 5 .ttJg2, and it would be very hard to obtain
counterplay, although the situation is not
tragic yet.
In Chess Informant Gelfand proposed:
1 5 . f3 ttJh5 1 6.ttJg2!?
If 1 6.EJ:fe l ttJ h3t ( 1 6 ... ttJ f4 1 7.ifl ) 1 7.<±>h l
if6! 1 8 .ttJg2 ig5 Black is OK.
1 6 . . . ih3
1 6 . . . f5 ?! is premature due to 1 7.exf5 ixf5
( 1 7 . . . gxf5 ? 1 8. f4±) 1 8.ttJe4 and White has
the better and risk-free position.
1 7.E1be l
After the sample move:
1 7 . . . ttJ f6!?
This knight is heading for a better place.
1 8 .ttJd 1 ttJ fh7 1 9 . ttJ f2 id7
A complex struggle would be ahead.
15 ...cxb4 16.gxb4 �g4 17.i.xg4
An undesirable exchange for White, but
forced.
He could not keep the bishop pair: 1 7.id2
ttJ h3t 1 8.<±>g2 ttJ f4t
13 ... �g5
Now this knight is the basis of my kingside
activity. At an appropriate moment it may
jump to h3, causing a headache for the white
monarch.
17 ...i.xg4
8
7
6
14J�b l �f6
5
8
4
7
3
6
2
5
4
3
a
2
a
b
c
d
e
f
15.b4?!
Playing too straightforwardly.
g
h
b
c
d
e
f
g
h
The plans of both sides are clear now: White
will continue to try to smash Black on the
queenside, whereas the absence of White's
light-squared bishop causes him worries; he
would be glad to take back the move g2-g3 ,
as it considerably weakens his king's residence.
264
King's I n dian Warfare
18.£3 �h3 19JHbl b6 20.a4?
Boris did not see my next move.
He proposed here the prophylactic manoeuvre:
20.'Dd l ! id? (now 20 .. . f5 ? does not work
- 2 l .ixg5 �xg5 22. 'D f2+-) 2 l .'D f2 'D h3t
22.'Dxh3 ixh3 23 .a4 with a double-edged
position after 23 . . . �fc8! ? - the pawn on c4
may be vulnerable in some cases.
b
a
d
c
f
e
g
h
22 ... Y;Yf6?
I became too excited during the game and
lost my cool. Somehow I just forgot about the
existence of the square f7.
a
b
c
d
e
f
g
h
20 . £5!!
..
An excellent resource; Black sacrifices the
exchange and takes over the initiative. This
move was also basically forced - after 20 . . . �ac8
2 l .a5 bxa5 22.�b7 Black's position would j ust
be bad strategically.
21 .lLlxg6
22 . . . �f7!
This was evidently a much stronger move the queen gains access to h5 and in many
variations it would be lethal for the white
king. It seems that White cannot repel the
attack:
23.'D xf8
Or 23.exf5 �fe8 and the pawn on f5 is falling,
just like all the rest of White's position (but
of course not 23 . . . ixf5 ? 24.�xf5 �xf5
25.'De7t) .
23 . . . �xf8
Played after quite lengthy thought.
Probably he had nothing better: 2 l .a5 bxa5
22.�b7 �f6; or 2 l .exf5 gxf5 22 .a5 f4 (not
22 . . . e4 ? 23.f4 'D f3t 24.'Dxf3 exf3 25 .if2)
23.gxf4 exf4 24.if2 �f7! and White will
hardly survive.
a
b
c
d
e
f
g
h
24.exf5
24.a5 is too slow now due to 24 . . .f4
265
Chapter 8 - Fighting for the Initiative
25 .if2 fxg3 26.ixg3 ltJ e l ! 27.Wff2 Wh5.
24.Wlf2 Wh5 or 24J�4b2 f4 are not cures
either.
24 . . . ixf5 25 .Wfe2
25 .Wg2 Wh5 with an irresistible attack;
again the queen goes to h5!
25 ... ltJ d4
24 . . . fxe4 25 .a5 Wfg6! 26.axb6 ltJd4
White would be in trouble - his king is in
real danger. For instance:
27.We l
27.bxa7? loses due to the simple but nice
27 .. Jhf2 28.ixf2 Wf5 29.a8=Wft \t>h7 and
the bunch of white pieces would not help to
avoid checkmate.
27 . . . axb6 28 .�4b2 Wg4 29.�f2
29.ltJb5 �fl t 30.Wxfl ixfl 3 1 .�xfl ltJ f5
32.�e l h5 33.�be2 \t>h7!+
29 . . . �f3 30.�bb2 ltJ f5 ! ?
8
7
6
5
b
a
c
d
e
f
g
h
4
26.ixd4 exd4 27.ltJ b5 d3-+
(Gelfand)
28.Wfg2 d2!
White is helpless.
3
2
a
b
c
d
e
f
g
h
believe that Black's initiative clearly
outweighs the small material disadvantage.
8
24 ...�xf5?!
7
This obvious but insufficiently energetic
move loses the rest of Black's edge. I fully agree
with Gelfand that Black had to proceed with:
6
5
4
24 . . . ltJd4! 25 .ixd4 (25 .Wff2?! ixf5 does not
look promising for White) 25 . . . exd4 26.ltJe4
Wxf5 27.Wd3
3
2
1
a
b
c
d
e
f
g
h
24.exf5
My opponent proposed here:
24.Wf2
As the more precise continuation. However,
after:
4
3
2
a
b
c
d
e
f
g
h
Chapter 8 - Fighting for the I nitiative
27 . . .ifl ! 28.�xfl �xfl t 29 .�xfl �xfl t
30.cj{g2 �el And this endgame is rather
unpleasant for White - at least he would have
to work to save half a point.
267
28 a6 29.tiJa3 tiJd4 30.YMg4 tiJf5
•••
O r 3 0 . . . cj;> h 8 3 l .�d l tLl f5 32.tLlc2 i s about
level.
3 1 .tiJc2 YMe5!?
25.YMg2 Lb1 26J�xbl
Black has regained the exchange but lost
the initiative. Still I was hoping for the best
and somewhere around here I declined a peace
offer.
26 e4 27.tiJb5!
•••
Boris correctly strives for some activity.
If 27.Ci:J e2 �f7!? he would still be on the
defensive.
8
It's Black's turn to take care about
maintaining equality and this centralizing
move is up to the task.
32 .if4
•
An interesting and forced draw could be
the result of 32.�xb6 �c3 33.�xe4 tLlxe3
34.tt:Jxe3 �e l t 3 5 .cj{g2 �e2t 36.cj;>h3 �h5 t
37.�h4 �f3 ! 38 .�b3 (38.tLlg2 �f5) 3 8 . . . �f5!
39.tLl xf5 �xf5t 40.�g4 �fl t 4 1 .cj;>h4 �f6t,
with perpetual check.
32 YMc3 33.YMe2 ges
•••
7
Probably the best solution is to keep
the tension. Mter 33 . . . �f3t 34.�xf3 exf3
3 5 . tLl e3!? Black could still experience some
difficulties in the endgame.
6
5
4
Now 34.�xb6?! �d3 would lead to unnecessary
risk for White.
3
2
a
b
c
d
e
f
g
h
27 YMe7
•••
27 . . . �f7 would pose more practical
problems. Still, after 28.if4 �g6!? 29.tLlxd6!
(29.�e2 is bad due to 29 . . .ie5 30.ixe5 tLlxe5)
29 . . . �xf4 30.gxf4 �xd6 3 l .�g4 cj;>h7 32.a5!
the position would be about balanced: rook
plus pawn is no worse than bishop and knight
due to the insecure position of the black king.
Even so, Black has some chances, I guess.
28.YMh3!
As we used to say in the pre-computer era
"this move leads to a position of dynamic
balance" .
34.tLle3 tLlxe3 3 5 .ixe3 �d3 36.�xd3 exd3
37.if4 with an equal endgame looks rather
logical.
Here I offered a draw, which Boris accepted
after some thought. Right after the game I was
really disappointed at missing such a golden
opportunity on the 22nd move. Besides, a
win would have equalized my lifetime score
of decisive games with Borya, which remains
at 8:7 in his favour now. Still I think it was a
tense battle in which both of us had our ups
and downs.
•!2-•12
268
King's I ndian Warfare
Ioannis Nikolaidis - Ilya Smirin
Panormo 200 1
l .d4 �f6 2.c4 g6 3.� c3 �g7 4.�0 0-0
If Black is going to play the KID it may be
more flexible to start with 4 . . . d6. In this case
after 5 .ig5 and the subsequent exchange of
this bishop for the knight with . . . 'Ll h5 he may
choose a plan with queenside castling. This
happened, for instance, in the following game:
5 . . . h6 6.ih4 g5 7.ig3 'Llh5 8 . e3 e6 9.�c2
�e7 1 0 .ie2 'Ll c6 1 l .ctJ d2 'Llxg3 1 2. hxg3 id7
1 3 .0-0-0 0-0-0 1 4 .�b 1 �b8 And the black
king is relatively safe; (Y2-Y2, 34) Moiseenko Cheparinov, Kemer 2007.
s.�gs d6
Another direction here is 5 . . . c5 6.d5 (after
6.e3 cxd4 7.exd4 d5! Black is OK, as practice
has shown) 6 . . . b5!? in the style of the Benko
Gambit.
6.e3
a
b
c
d
e
f
g
h
White plays the opening in quite a timid
way. He has developed the bishop on g5 and
after that put his pawn on e3, not on e4 as is
usual in the KID . But this scheme is clearly not
without venom, and Black should play actively
and precisely not only to get counterplay
but j ust to try to equalize. Despite White's
setup not being overambitious, he enjoys free
development and his pawn centre has no weak
spots - the pawn on d4 is well defended. OK ­
all these are common words. Now we will look
at some concrete examples.
6... h6
Of course, Black may choose different
setups here, but the main question is whether
to prepare . . . e5 or to play . . . c5 now or
slightly later. I prefer the second option. The
immediate 6 . . . c5 is possible and may often
lead to a transposition of moves. It happened
in the following game:
7.ie2 h6 8.ih4 g5 9.ig3 'Llh5 1 0.dxc5!?
White changes the pawn structure, trying to
exploit the weakening of Black's king.
1 0 . . . 'Llxg3 1 l .hxg3 dxc5 1 2 .�c2 e6
Of course not 1 2 . . . ctJ c6? 1 3 .'Llxg5 hxg5?
1 4 .�h7# - as I said earlier, White's setup
has its venom.
1 3 .�d 1
1 3 .g4!? 'Llc6 1 4.a3 seems to me to be a
more aggressive and better continuation.
14 . . . a6 1 5 .'Lld2 (I think that Alexey
Kuzmin's proposal 1 5 .id3!?, with the idea
of queenside castling and trying to create
threats against the black king, deserves
serious attention) 1 5 . . . �b8 1 6.g3 b5
1 7.'Llce4 b4 1 8 .a4 b3! 1 9 .'Llxb3 'Ll b4 20.�b 1
ib7 2 1 .0-0 ixe4 22.�xe4 �c7; (Y2-Y2,
3 1 ) M . Gurevich - Fedorov, Stockholm
1 997.
13 . . . �e7 1 4.0-0 'Llc6 1 5 .�d2 g4 1 6.'Llh4 h5
1 7.�fd 1 ctJ e5 1 8 .id3 id7 1 9.ie4 ie8
Black has completed his development and
stands well with the two bishops. Control of
the d-fi l e does not promise White anything
concrete, and his knight is badly placed on
h4.
20.�b3 �b8 2 1 .�d6 if6
269
Chapter 8 - Fighting for the Initiative
13 ..ie2 e6 14.dxe6 .ixe6
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
22.lDd5?
This tactical attempt proves to be in Black's
favour.
22 . . . exd5 23.tDf5 Wxd6 24.tD xd6 dxe4
25.tDxe4 ig7 26.�d6 ic6 27. tD f6t ixf6
28.�xf6 �g7 29.�f5 f6 30.�xh5 �h8 3 1 .�xh8
�xh8
(0-1 , 45) Beim - Smirin, Tel Aviv 1 996.
7..ih4 c5 8.d5 g5 9 . .ig3 � h5 IO.�d2
1 0.id3 i s considered t o b e the most
aggressive continuation nowadays - the
bishop eyes the black king. In my opinion,
the following encounter may serve as a good
example of Black's action in this case:
1 O . . .f5!?
To fight for the initiative, Black agrees to
an additional weakening of his already
compromised king.
1 1 .tDd2 tDxg3 1 2.hxg3 ttJd7 1 3 . tD f3 tD f6
The knight arrives to help its monarch.
1 4.Wc2 lD g4 1 5 .tDh2 tDe5 1 6.ie2
( 1 -0, 45) Agrest - Nithander, Copenhagen
2007. Now after 1 6 .. .f4 Black would be
fine - he has active pieces and his king feels
reasonably comfortable.
10 ... �xg3 l l .hxg3 �d7 12.¥Nc2 �f6
I think it was better to act as in the previous
game - 1 2 .. .f5!? 1 3 .id3 (or 1 3 .ie2 tD f6)
1 3 . . . tDe5 1 4.ie2 f4 1 5 .exf4 gxf4 1 6.gxf4 �xf4
1 7.g3 �f8 1 8. f4 ttJg4, and again my sympathies
lie with Black in this complicated position.
a
b
c
d
e
f
g
h
1 5.�de4?!
White could have strengthened his grip over
the centre by means of 1 5 .�d 1 ! ?, wi rh better
prospects - it's hard to see compensation
for the weaknesses on d6 and d5 and the
somewhat exposed king on g8. An important
point is that the freeing attempt 1 5 . . . d5? does
not work due to 1 6.tDb3.
15 ...�xe4 16.�xe4
Now Black has time to try to open the
position for his bishops.
16 ... b5
A tempting but objectively doubtful pawn
sacrifice.
Sometimes one should play simple chess after 1 6 . . . Wa5 t 1 7.Wd2 (stronger is 1 7.�fl
�fd8 1 8. f4 gxf4 1 9.gxf4 d5 20.cxd5 ixd5,
with an unclear position in which both sides
have their trumps) 1 7 . . . Wxd2t 1 8 .�xd2 d5!
1 9 .tD xc5 dxc4 20.tDxe6 fxe6 2 1 .�afl �ad8t
22.�c2 �d6 23.f4 c3! With the idea 24.bxc3
�c8 25.c4 �a6, and Black would enjoy a risk­
free advantage in the endgame.
17.cxb5 ¥Na5t
1 7 . . . �c8, preparing . . . d5, looked like a
reasonable alternative.
270
King's I ndian Warfare
8
7
8
6
7
5
6
4
5
3
4
2
3
1
2
1
a
b
c
d
e
f
g
h
19.a3?
a
b
c
d
e
f
g
h
21 .fxg5!?
A correct sacrifice, although I did not think
so during the game. . .
This continuation is too slow and soft.
Mter the energetic 1 9 .f4! White would have an
edge, though in a rather complicated position.
Mter both 19 . . . f5 or 19 ... �b4 he would react
with 20.lt:Jc3 , with an extra pawn and play
against the black king.
Th e quiet alternative 2 1 .lt:Jc3 axb5 22.lt:Jxb5
was possible, but after 22 . . . !:!ab8 23.a4 �b4
24. <i>f2 d5! Black would obtain a promising
position. Nikolaidis decided to take an active
approach instead.
2 1 . fxe4 22.VNxe4
••
19 a6 20.£4
••.
20.bxa6 understandably did not appeal to
Ioannis - after 20 . . . !hb8 2 1 . lt:J c3 d5 22.f4
d4 his position would be under enormous
pressure for j ust two pawns.
It was possible to sacrifi c e a rook: 20J:hh6!?
.ixh6 2 1 .lt:J f6t <i>f8 (dangerous is 2 1 . . . <i>g7
22.lt:Jh5 t <i>f8 ? [22 . . . <i>g8=] 23 .�h7) 22.�h7
<i>e7 23 .�xh6 !:!h8 24. lt:J h7 !:!ag8 2 5 .�f6t
<i>d7 26 ..id3 axb5 27.<i>gl with an unclear
position - j ust an illustration of a possible
tactical outbreak.
20 £5
•..
White has enough pawns for a bishop, but
the pawn structure fails to impress. Much
more important, however, is that the black
king is facing a surprisingly strong attack mainly because the queen is stuck on aS .
22 �f7
•.•
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
27 1
Chapter 8 - Fighting for the Initiative
23.J.d3
Houdini's suggestion 23.!hh6! indeed seems
to be more worrying for Black. It has j ust one
drawback: it would be almost unreal to make
this move over the board.
Now 23 . . . .ixh6 24.gxh6 @f8 2 5 .�f4 �d2
26.�f6 looks very dangerous for Black.
So the best option seems to be: 23 . . . �a7!
24.�h4 ixb2 25 .�b l .ie5 26.b6 �e7
a
b
c
d
e
f
g
h
25J�xh6?
I recall that I was counting on this move.
a
b
c
d
e
f
g
h
Intuitively I would take Black in this wild
position, even though the silicon brain believes
otherwise.
23 .. J�e8?
In such a sharp and unbalanced position any
mistake may cost dearly. Without any reason,
Black has forced the white queen to go to h7 but it was heading there anyway.
The correct move seems to be 23 . . . c4! ?, striving
for activity and threatening the bishop on d3 .
After 24 ..ic2 d5 (24 . . . hxg5 25. @g l ! and the
a l -rook will join the attack from fl ) 25 .�h7t
@f8 26.gxh6 .ixb2 27.�b l .ie5 28 .�f5 �c3
29. h7 ih8 another crazy position would arise,
and again I would prefer Black. But of course
this is just a sample line.
24.�h7t @f8
Meanwhile the following quiet move would be
very strong:
25.@gl !
The rook will tease Black from fl , while the
white king is in safety. Suddenly Black has
no satisfactory defence. For instance:
25 . . . �d2
Or: 25 . . . hxg5 26.�fl �a7 27.�g6+26.�fl �xe3t 27.@h2 �xg5 28 . .ig6 �a7
29.�f5 ± �g4
Only 29 . . . �xf5 30 ..ixf5 axb5 leaves chances
for a draw.
30.�hfl �ee7 3 l .b6 �ab7 32.� l f4 �d l
33 ..ixf7 �xf7 34.�g6 �b3 3 5 .�xf7t �xf7
36.�xf7t �xf7 37.�xf7t @xf7 38.b7
The pawn sneaks to the prized 8th rank.
25 ...�d8!
Finally, the queen joins the defence. The
white king is stuck on fl and this is enough to
completely turn the tables.
26J�h4
Other moves are hardly better.
The computer's suggestion 26.�g6!? .ixg6
27.�xg6 fails to save White after 27 . . . �a7
28.@gl �f7 29 . .ic4 d5 30.�d l �e5
272
King's Indian Warfare
3 1 . bxa6 El:xg5 . Pawns are pawns, but an extra
rook is an extra rook.
26 ...Wxg5 27JU4 axb5
7
6
5
4
3
Now everything is clear - White is heading
downhill to lose the game.
28. �gl ga7 29.gaf1 gxe3 30 ..ig6 gee?
3 1 ..ih5 gac7 32.g4f3 Wh6 33.Wf5 �g8
34.g4 .ixh5 35.gxh5 ge5 36.Wg4 gg5
0-1
Chapter 9
Materialism
I was very happy when I saw these old photos in a Russian article.
64, the magazine, were happy to allow us to use them for this book.
I would like to thank Maxim Notkin for his help with this.
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering
these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
� 8
7
6
5
4
3
2
2
a
b
c
d
e
f
g
h
a
I missed something simple here
(see page 28 1 )
� 8
3
f
g
h
8
6
b
c
d
e
f
g
h
b
c
d
e
f
g
How can Black refute
the white attack?
(see page 277)
h
e
f
g
h
6
5
4
3
2
a
d
resource? (see page 293/94)
3
2
c
What is White's tactical
4
3
b
� 8
7
5
4
a
Basic King's Indian strategy.
What is the natural move?
(see page 283)
6
5
What is Black's strongest
move? (see page 29 1 )
3
� 8
7
7
h
2
a
Basic tactics. Black wins.
(see page 275)
g
4
2
e
f
5
4
d
e
6
5
c
d
7
6
b
c
� 8
7
a
b
2
a
b
c
d
e
f
g
h
Simple chess. What is natural
here? (see page 284)
a
b
c
d
e
f
g
h
Black's best move? (which
worked wonders!)
(see page 299)
Chapter 9
-
This is a bit of an odds and ends chapter, with
four games that I really wanted to include in
this book, but which did not really fit into
other chapters. In the first two games the
bishops are a factor, but I did not really want
to put them in the chapter about bishops. The
last two games are a mess. One thing that all
four games have in common is that I managed
to win material in them. Hence the theme of
materialism.
However, this topic is not j ust a poor man's
excuse to include four nice games, but also
a chance to make an important point: the
dynamic approach to the King's Indian is about
upsetting the balance of the game, which can
be done by taking material as well as giving it.
Just before finishing this book, I played a
tournament in Sweden where I was able to
come up with a little combination:
Johan-Sebastian Christiansen - Ilya Smirin
Stockholm 20 1 6
8
7
275
Materialism
3o.Y;Yd4 .id? 3L.ids Y;Ygs 32.Lf7t gxf7
33.Y;Yc4 .ifS 34.d7 �c7 35J��xf5 gxf5 36.e6
�xe6 37.Y;Yxe6 .if6 38 ..ie3 f4 39 ..ixf4 Y;YcSt
40.<it>g2 Y;Ye? 4I .Y;Yg4t �g7
0-1
I want to win games. Whether this is achieved
by winning material or sacrificing is less
relevant.
The following game is a wonderful example
of Black accepting material and defending by
astonishing tactics.
Loek van Wely - Teimour Radjabov
Biel 2007
This game is a fine victory for Radjabov, but
I should point out that Van Wely is a great
fighter and he tried very hard to beat his
opponent with ambitious and risky play.
l .d4 �f6 2.c4 g6 3.�c3 �g7 4.e4 d6 5.�f3
0-0 6 ..ie2 e5 7.0-0 �c6 8.d5 �e7 9.b4 �h5
IOJ��el f5 l l .�g5 �f6 12.f3 i>h8 13.�e6
.ixe6 14.dxe6 � h5 15.c5 �f4 16 ..ic4 fxe4
17J�xe4 �f5
6
5
4
5
3
4
2
1
a
b
c
d
e
f
g
h
26 .. J�xe3! 27.fxe3 .ixc4 28.exd4 .ixa2
29.dxe5 .ie6
Black has won a piece. My opponent tried
to create a bit of counterplay, but it was not
enough to compensate for the lack of material.
3
2
1
a
b
c
d
e
f
g
h
The position is roughly balanced after the
opening, but Van Wely plays it as if he is
better. Only because Radjabov reacted with
King's I ndian Warfare
276
great moves was he able show that this is not
the case.
1 8.g3!? � h3t 19.cj{g2 �g5 20.:Sg4!?
Provoking his
material.
opponent
into
winning
20.e7 �xe7 2 1 . 4Jd5 �d8 22.�xg5 �xg5
23.4Jxc7 was the safe continuation, but Black
is also fine after: 23 . . . 4J e3t 24.gxe3 �xe3
2 5 . 4J xa8 e4 26.f4 gxa8 27.cxd6 gd8 , and
White's advantage is almost non-existent.
8
7
6
5
4
3
2
1
b
a
d
c
e
f
g
h
Both players had no doubt predicted this
position at move 20. I guess that Van Wely
had already then seen the very creative idea he
played in the game:
25.:Sf4!?
Fighting for control of the dark squares White tries to sacrifice another exchange.
25 ....ixf4
Radjabov accepts the challenge. 25 . . . gf5!?
was more defensive, but also playable.
26.gxf4?
2 1 . cj{xf3!
2 l .�xf3 ? lt.J h4t 22.gxh4 gxf3 23.c;t>xf3
e4t! 24.4Jxe4 �xa l would only give White
two pieces for the queen. The e6-pawn is not
strong enough to make up for this difference.
But this is wrong. 26.�xf4 was the right
move, when the position is still in dynamic
balance. 26 . . . dxc5 (or 26 . . . gxf4 27.�a l t c;t>g8
28.gxf4 �f8 and White will give a perpetual
with the knight on f6 and h5, as in so many
lines) 27.�a l t c;t>g8 28.bxc5
2 1 . .. e4t 22.�xe4 hal 23.�g5 �e7t
24. cj{g2 �e5
a
b
c
d
e
f
g
h
277
Chapter 9 - Materialism
White is threatening ih6, so Black has no
choice but to play 28 .. J�xf4 29.gxf4 'l!Mf8 and
the game ends in a perpetual after 30 . ltJ f6t
�h8 3 1 .ltJh5t and so on.
28.'i'd4t
I do not know if Van Wely considered
27 . . . h6, but if he had then he could have seen
that 28.ixh6 ltJxf5 29.'l!Mal t c;t>g8 30.ig5 was
the critical line:
26... d5! 27.f5?!
Van Wely continues down the most
ambitious path, but it does not hold up, so
White should have played 27.'l!Ma l t �g8
28.ih6, although after: 28 . . . ltJ f5 29.ixf8
'l!Mxf8 30. ltJ f6t �h8 3 1 .ixd5 'l!Mg7
a
b
c
d
e
f
g
h
5
Now 30 . . . ltJ e7 is the most obvious move.
The variation continues with: 3 1 .ib3 c6
(3 1 . . . dxe4 32.ih6 ltJ f5 33 .e7t) 32.ltJf6t :Bxf6
33 .'l!Mxf6 'l!Mf8 34.'l!Mxe7 'l!Mxe7 3 5 .ixe7 :Be8
36.id6 :Bxe6 37.�f2 and Black will have a
difficult defence in the endgame.
But Radjabov had seen further and found
30 . . . ltJ e3t!!, which turns the position on its
head.
Mter 3 l .ixe3 dxc4 32 .ih6 'l!Me7, White
does not have significant compensation.
And after 3 1 . �g 1 "!M e8 the attack crashes to
earth quite quickly as well. For example: 32.e7
:Bf5 33.ixe3 dxc4 34. ttJ f6t :Bxf6 3 5 .'l!Mxf6 'l!Mf7
36.'l!Me5 :Be8 37.ig5 'l!Mf5 and Black wins.
3
28 ... <i>h7 29.fxg6t <i>xg6 30.Le7 'i'xe7
3 1 .'i'xd5
a
b
c
d
e
f
g
h
32.'l!Me5 c6 33 .ie4 ltJe7 The position is very
close to lost for White.
8
7
6
4
2
Or 3 1 .id3 is met by 3 1 . . . 'l!Mxe6!.
a
b
c
d
e
f
g
h
27 ... h6!!
27 .. Jhf5 28 .'l!Md4t c;t>g8 29 .ih6 looks very
dangerous for Black, but surprisingly he is OK
after 29 ... 'l!Mf8 30.ixf8 :Baxf8, when he has two
rooks for the queen. The position is in balance.
a
b
c
d
e
f
g
h
278
King's Indian Warfare
This looks very dangerous for Black until you
start considering how White should continue
and see that there is nothing. 32.lDg3t @f7
33.if5 WeB 34.'.Wxd5t @g7 3 5 .Wd4t �f6
36.ltJe4 '.Wf7 and the attack is coming to an
end.
3 1 .. J�ad8
My Games
Maxim Rodshtein Ilya Smirin
-
Ashdod 2006
l .d4 tljf6 2.c4 g6 3.tlJc3 i.g7 4.e4 d6 s.tlJf3
0-0 6.i.e2 e5 7.0-0 ttJ a6 s.i.e3
8
The attack is over.
7
6
8
5
7
4
6
3
5
2
4
1
3
2
a
1
b
c
d
e
f
g
h
8 ... h6
a
b
c
d
e
f
g
h
34.�g3t ®h7 35.�xg7t ®xg7 36.®g3 gd4
0-1
The first main game of this chapter is from the
last round of an open tournament in the Israeli
coastal city of Ashdod. My opponent was a
future grandmaster and fellow member of the
national team, Maxim Rodshtein. We both
needed a win to tie for first place.
An experimental move. The idea of this
modest prophylactic move is to prepare . . . ltJg4
without White being able to reply with ig5 .
I played this move a few times with good
results, but now I think that this loss of time is
not fully j ustified.
9.tlJel!?
This original move was invented by Russian
grandmaster Denis Yevseev.
9.dxe5 lDg4 is more natural and was played by
Boris Gelfand against me in the Blitz World
Championship. Here is this interesting game:
1 0.i.d2 dxe5 l l .'.Wc l @h?
279
Chapter 9 - Materialism
Yz-Yz Gelfand - Smirin, Rishon Le Zion
2006.
Playing the "symmetrical prophylactic" move
9.h3 is not a very good idea - after 9 . . . exd4
1 o.lt:Jxd4 ( 1 o.ixd4!?) 1 o . . . El:e8 1 l .f3 (better
is 1 1 .if3 , but it does not bring an opening
advantage either after 1 I . .. lt:Jc5 1 2.�c2
lt:J h7!, with the idea 1 3 . . . lt:Jg5) 1 1 . . . lt:Jh5 the
drawbacks of 9 .h3 become obvious.
a
b
c
d
e
f
g
h
1 2.E!:d 1
In my opinion, better here is 1 2 . h3 lt:J f6
1 3 .lt:Jxe5 and with a slight transposition
of moves a position has arisen from the
game Wantola - Smirin, Ohrid 2009 , in
which I had the worse position, despite
winning in the end. Surprisingly, even in
home analysis I did not find equality after
1 3 .lLlxe5 .
1 2 . . . �e8 1 3 . h3 lt:J f6 1 4.ie3 lt:Jh5 1 5 .c5 f5
1 6.ixa6 bxa6 1 7.lt:Jd5 �f7 1 8 .c6 ie6 1 9 .lLlc3
g5 20 .lt:J d2 lLl f4 2 1 .ixf4 gxf4
9 .. .lDg4!
The correct reply, based on simple tactics.
After 9 . . . exd4?! 1 0.ixd4 c6 1 1 .lLl c2 (this
and the next moves are possible as a result
of 9.lt:J e l ) 1 1 . .. El:e8 1 2. f3 lLl c7 1 3 .�d2 d5
1 4.cxd5 cxd5 1 5 .e5 White has obtained an
indisputable edge, mostly because of the bad
position of the knight on c7; ( 1 -0, 34) Yevseev
- Loginov, Kazan 2005 .
IO.Lg4 Lg4 1 1 .f3
Of course Black would welcome 1 1 .�xg4?!
exd4 .
8
7
6
5
4
3
a
b
c
d
e
f
g
h
22.<i>h2 if6 23.exf5 ixf5 24.lt:J de4 �g6
25 .El:d7t <i>h8 26.f3 El:g8 27.�fl ixd7 28.cxd7
E!:ad8 29.El:d 1 ih4 30.lt:Ja4 ie7 3 1 .El:d2 �c6
32.�d 1 E!:g6 33.E!:d5 id6 34.lt:Jac5 E!:dg8
3 5 .g4 fxg3t 36.<i>g2 E!:d8 37.b4 aS 38.b5 �b6
39.a4 ixc5 40.E!:xc5 El:g7 4 I .E!:d5 �e6 42.lt:Jc5
�f5 43 .�d3 �xh3t 44.<i>xh3 g2 45 .�e3
g 1 =� 46.�xg 1 E!:xg 1 47.lt:Je6 E!:dg8 48 .E!:d2
E!:h 1 t 49.El:h2 E!:d1
2
a
b
c
d
e
f
g
h
l l . ...ic8!
This is stronger than the natural 1 1 . . . id7
1 2.�d2 <i>h7 1 3 .El:d 1 ; ( 1 -0, 25) Golod
- Areshchenko, Port Erin 2007, and the
unfortunate position of the bishop on d7 is
already making itself felt.
280
King's Indian Warfare
12.�d2 <i>h7 13J!d1 �e8
In my opinion Black's position is perfectly
respectable and the bishop pair fully
compensates for the opponent's supremacy in
the centre. As Tarrasch used to say: "The future
belongs to he who has the bishops." In this
particular game he proved to be right!
Starting a rather standard kingside attack,
which gains in strength here because White's
light-squared bishop has already gone.
17.�e2
Slightly more precise was 1 7.b5 ttJb8
1 8 .tLld5 '!Mf7 with complicated play, but over
the board I would definitely choose Black.
14.llJc2
White could close the centre: 1 4.d5 fS
1 5 .tLl d3 f4 1 6 . .if2 gS 1 7.'!Me2, but it would
lead to a dangerous kingside attack after
1 7 . . . '!Mg6 1 8 .c5 g4 1 9.C4ih l .if6 20.b4 �g8 .
17...�g6 18.b5 llJb8
8
7
6
14 .. £5
5
.
8
4
7
3
6
2
5
4
3
a
2
1
a
b
c
d
e
f
g
h
15.b4?!
My young opponent (Maxim was 1 7 when
this game was played) does not sense the
potential danger.
He should have opted to be more active on
the kingside: 1 5 .exf5 gxfS 1 6.f4 e4 1 7.b4
(or 1 7.d5 lLlcS 1 8 ..id4 �g8 1 9 . .ixg7 �xg7)
1 7 . . . c6 and Black, with two bishops and
potential pressure along the g-file, is OK. But
White has his own trumps (good minor pieces,
some space advantage) , so in my opinion the
chances are approximately even.
1 5 ... £4 16 ..ifl g5
b
c
d
e
f
g
h
An interesting picture. In the middlegame
all of Black's queenside pieces are on their
initial positions, yet there is hardly any way for
White to exploit this fact. White's pieces and
pawns look active, but somehow they lack the
energy to deliver a substantial blow. Black, in
contrast, is not very active yet, but he will soon
start a quick action against the white king.
The absence of White's light-squared bishop
is also telling - the pawn on c4 may easily
become a weakness. To put it simply, Black's
undeveloped army is better coordinated than
his counterpart's fully developed one.
19.dxe5
Or 1 9.tLld5 �f7 20.dxe5 dxeS 2 I ..ic5 .ie6
(2 1 . . . tLld7!? is not bad either, after which the
following interesting continuation is possible:
22.tLlxc7 [better is 22 . .ib4] 22 . . . tLlxc5 23.tLlxa8
g4 24.@hl b6 25.�d8 .ib7 26J�b8 g3
Chapter 9
-
28 1
Materialism
8
27.h3 �d6! 28 .El:e8 � d3 and Black dominates
the board) 22.�cb4 g4, with promising
prospects.
7
19 ... dxe5 20 ..lc5 ges 2I .tiJd5 Y;Vf7
5
6
8
4
6
2
3
7
1
5
4
a
3
b
c
d
e
f
g
h
Now it's my turn to attack some targets in
White's camp, and the pawn on c4 is already
doomed!
2
1
a
b
c
d
e
f
g
h
Max has made all the active moves and Black
is still undeveloped, but what should White
do now? Rodshtein could not find the right
answer, but it was not an easy task.
22.a4?
Out of inertia, White makes another active
move. Alas, it proved to be his last active one
in the game.
Perhaps he should have played the defensive
22.ctJa3 instead, protecting in advance the
weakest link in White's camp - the c-pawn. But
even then Black would be fine after 22 . . .ie6
23.El:d2 � d7 (or 23 . . . c6 24. � b4 � d7 25 .id6
� b6) 24.ib4 �b6.
22...c6 23.tiJc3 �e6
24.bxc6?
An oversight.
Houdini recommends parting with the
pawn immediately with 24.�f2 ixc4 25.El:fe l
o r 24 .�a3 'Ll d7 25 .id6 'Ll b6 26.bxc6 bxc6
27.�f2 ixc4. This says a lot about the
problems White is facing.
24... bxc6?
Of course not the immediate 24 . . . ixc4?
25.cxb7.
But neither of us noticed 24 . . . b6! 25.c7
ixc4, with a quick win.
25.tiJa3 tiJd7 26 ..ld6 tiJb6
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
282
King's Indian Warfare
27.c5?
In trying to save a pawn, my opponent loses
even more material.
In the next game I managed to win the
exchange. The way I exploited it was to
continue to play energetically and use the extra
firepower to my advantage.
His prospects were already not bright, but
after the correct 27.�fe l 4Jxc4 28.4Jxc4 .ixc4
29.'1Wf2 I would still have a lot of work to do to
convert the extra pawn.
27 ... tLlc4
Now the loss of the exchange is inevitable.
28.tLlc2 tLlxd6 29.cxd6 .tc4 30.%Yfl hfl
3 1 .%Yxfl E:ab8 32.tLlel E:ed8 33.%Ya6
8
David Anton Guijarro - Ilya Smirin
Yerevan 20 1 4
This game was played i n the 9th round o f the
European Championship in Yerevan. We both
were in the leading group and so the result of
the game was very important from a sporting
point of view.
I .d4 tLlf6 2.c4 g6 3.tLlc3 .tg7 4.e4 d6 5.tLlge2
0-0 6.tLlg3
7
8
6
7
5
6
4
5
3
4
2
1
a
b
c
d
e
f
g
h
3
2
33 %Yb3 34.%Yd3 .t£8
•.•
The d6-pawn will disappear very soon, so
further resistance is pointless. White resigned.
His active position fell apart with surprising
speed!
0-1
a
b
c
d
e
f
g
h
Usually White chooses this variation to avoid
well-trodden paths. The evident drawback is
that the knight moves twice in a row as early
as the sixth move. But the KI D setup allows
White to do that without visible damage. Still,
I do not quite understand the idea behind
this knight manoeuvre, even though I have to
admit that statistically White does reasonably
well here.
6 tLlbd7
.• .
Black is flexible in his choice of continuation.
283
Chapter 9 - Materialism
One of my games went a la the Benko Gambit:
6 . . . a6 7 .ie2 c5 8 .d5 b5 9.cxb5 axb5 1 0.ixb5
ia6
a
a
b
c
d
e
f
g
h
1 1 .0-0 ixb5 1 2 .ttJxb5 ttJe8 1 3 .�c2 ttJa6
1 4.ttJe2 tDec7 1 5 .ttJ bc3 �d7 1 6.b3 ttJ b4
1 7 .�b 1 f5 1 8 .a3 tDba6 1 9Jh2 �ab8 20.�b2
ttJb5 2 1 .�d3 ttJ ac7 22.�b 1 ttJxc3 23.tDxc3
fxe4 24.ttJxe4 �f5 25.ttJc3 ixc3 26.�xc3
ttJxd5 27.�c4 e6 28 .id2 tD c7 29 . b4 �a4
30.�d3 �d5 3 1 .�c3 ttJb5 32.�f6 �xd2
33.�xe6t <±>g7 34.�e7t <±>g8 3 5 .�e6t <±>g7
Y2-Y2 Serper - Smirin, Philadelphia 1 996.
7 .ie2 h5
.
A somewhat risky but logical move, stressing
the disadvantage of the knight's position on
g3.
b
c
d
e
f
g
h
This active and logical move makes Black
worry about his kingside, which was
weakened by the advance of the h-pawn.
In the game Bologan quickly gained an
overwhelming advantage:
8 . . . e5
The other plan is 8 ... c5 9.d5 b5!? trying to
obtain counterplay by acting in the spirit of
the Benko Gambit (well, we have already
seen something like that) . But it does not
seem to equalize after: 1 O.cxb5 a6 1 1 .a4
�a5 ( l l . . . axb5 1 2.ixb5 ttJ b6!?) 1 2.id2
h4 1 3. ttJ fl axb5 1 4 .ttJ xb5 �b6 1 5 .�c2
ia6 1 6.ttJe3 ixb5 1 7.ixb5 �xb5 There
is hardly anything better. 1 8. axb5 �xa 1 t
1 9 . ttJ d 1 The compensation for the queen is
clearly insufficient.
9.fxe5 dxe5 1 0.d5 h4 1 1 . ttJ fl ttJ h7 1 2.g4
s.!L}fl
I am not sure this is a 1 00 per cent healthy
idea - this knight makes too many moves in
the opening.
8 .ie3 e5 9.d5 a5 1 0. a3 ttJ c5 1 1 .h4 tDg4
1 2.ixg4 ixg4 1 3 .f3 id?, with excellent play
for Black, was the game Nabaty - Smirin,
Skopje 20 1 3.
The best continuation was played by Viorel
Bologan at the end of 20 1 5 : 8 . f4!
a
b
c
d
e
f
g
h
1 2 . . . c6?
There was no good alternative to: 1 2 . . . if6!
1 3 .ie3 ig5 1 4.�d2 (or 1 4.if2 ie7!? with
284
King's I ndian Warfare
the idea . . . ic5) 1 4 . . . ixe3 1 5 .lDxe3 lDg5,
with an interesting position. White has an
obvious space advantage, but Black controls
some important dark squares and, thanks to
that, he may count on counterplay.
1 3.h3 �e8 1 4 .ie3 g5 1 5 .lD d2 lD df8 1 6. lD f3
if6 1 7.'!Mb3 lDg6 1 8 . 0-0-0 '!Ma5 1 9.c5 ltJ f4
20.ic4 '!Mc7 2 1 .d6 '!Md7 22.ixf4
1 -0 Bologan - Al Sayed, Qatar Masters
20 1 5 .
9.ig5
This was the move to play. After that Black
may continue:
9 . . . d5!?
9 ... ltJh7 1 0.ie3 e5 1 1 .d5 c5 would be a
more standard approach.
1 O.cxd5 cxd5
a
b
c
d
e
f
g
h
1 1 .e5
1 1 .exd5 would of course be mer by 1 1 . . . ltJ b6.
1 1 . .. ltJ e4 1 2.lDxe4 dxe4 1 3 .lDd2 f6 1 4.exf6
exf6 1 5 .if4 f5 1 6.ltJc4 ltJb6 1 7.lDe5 �h7
All these moves are logical. Black has good
prospects - his pawn chain looks attractive and
his pieces are well placed.
8 . . . a6 9. lDd2 c6 1 0 .0-0 b5 1 1 .a3 had already
happened in one of David's games: 1 1 . . . ib 7?!
As is well known, this bishop usually belongs
on the c8-h3 diagonal in the KID . 1 2.�e 1 e5
1 3 .d5 cxd5 14.cxd5 �c8 1 5 . h3 ih6 1 6.ifl
�g7 1 7.lDb3 etc.; ( 1 -0, 66) Anton Guijarro ­
Naroditsky, Al Ain 20 1 3 .
8 . . . e5?! 9.d5 lD c5 1 0.ig5 '!Me8 1 1 .lD d2 would
play into White's hands: he reaches a very
convenient piece setup after all.
9.liJd2?
But this is already too much. The knight
moves as if it is constantly searching for a
better fate - in chess this rarely pays off.
9 ... e5 10.d5 a5
My opponent probably underestimated this
move - now it became clear that White has
j ust lost one or two tempos compared with the
usual lines (when the knight goes to d2 by the
route g 1 -f3-d2) .
1 1 .0-0 lDc5 12.b3
If 1 2 .'!Mc2 ih6 there would arise a
comfortable position for Black from the game
Petrosian - Geller, Moscow 1 949, with two
"extra" moves, . . . c6 and . . . h5, which makes it
even more pleasant for Black.
12 ....ih6 13 ..if.3 liJd3
Black rarely obtains such a great position out
of the opening in the KID.
285
Chapter 9 - Materialism
14.%Yc2
1 4 . .ia3 would look clumsy, but perhaps
the idea of keeping the dark-squared bishop
deserved some attention. At least I thought
during the game that Anton Guijarro would
play it.
14 ... �xcl
It felt great to eliminate this bishop - a very
important piece in the KID.
15J�axcl cxd5 16.exd5
My opponent did not want to suffer
after: 1 6.cxd5 .id7 1 7.a4 �c8 1 8 .�b l b5!
Understandably he decided to sacrifice the
exchange instead.
16 ...�f5 17.%Yb2
8
7
6
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
White is left with j ust a pawn for the
exchange, but Black will still have to work
to convert his minimal material advantage.
White's problem is the passive bishop on e4. If
we were to imagine this bishop as being on e3,
for example, Black would have no advantage at
all, to say the least.
20 ...%Ye7
After "winning" the opening part of the
game, I relaxed a little - this habit has cost me
dearly during my career.
5
4
3
2
1
a
b
c
d
e
f
g
h
17 ... e4
During the game I hesitated over whether
to win the exchange or to proceed with
1 7 . . . ttJd7!?, which was also strong. But in the
end, greed took over.
18.�dxe4
Or 1 8 . .ie2 e3 1 9 . fxe3 .ixe3t 20.<i>h l tDd7
2 l .�ce l tDc5 was hardly better - Black's
pressure is very strong.
18 ...he4 19.he4 .ixcl 20J�xcl
20 . . . �e8 2 l ..if3 �e5 22.h3 'We7+ would be
more accurate - Black will probably exchange
a pair of rooks, thereby increasing his
domination on the dark squares.
21 .�f3 YNe5 22.%Yd2 �fe8 23.�b5
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
286
King's Indian Warfare
White is trying to create counterplay by
means of c4-c5 . The d6-pawn demands some
attention now.
23 .. J:�e7 24.h3
24.c5 dxc5 25 .d6 El:d7 26.E!:e l �f5 would
lead nowhere.
24... tlJ d7 25.tDc7
A curious, yet at the same time logical,
variation could arise after: 25 .�c3 �xc3 26.El:xc3
El:e l t 27.cj;>h2 E!:a6 28.tLlc7 E!:b6 29.ttJa8
This move cramps White's kingside.
30.tlJd4
30.�d4 E!:ce8 3 l .c5 would be no better after 3 l . .. dxc5 32.bxc5 El:e l 33 .�d2 E!:xc l
34 .�xc l El:e5 35 .�c4 E!:f5! 36.c6 bxc6 37.dxc6
ttJ b6 3 8.�c2 �al White's counterplay attempt
would fail.
30 ...�e4 3I.tlJb5 b6
To avoid having to worry about a possible
c4-c5 .
32.tDc3
The following variation reflects the essence
of the position well:
32.�c2 E!:ce8
Black gives up the pawn on a4, but activates
his pieces and creates concrete threats.
33 .�xa4 E!:e l 34.�c2
8
b
a
c
d
e
f
g
7
h
29 . . . E!:b4! 30.a3 E!:a l ! 3 1 .axb4 axb4 32.El:e3
E!:xa8 It's very unlikely that White can hold this
ending - the difference in the activity of the
pieces is too great.
6
25 .. J;cs 26.tlJb5 a4 27.b4 V9f6!?
2
5
4
3
The queen vacates e5 for the rook.
a
28.�e2 �e5 29.�fl h4!
a
b
c
d
b
c
d
e
f
g
h
32 . . . �f4! 3 5 .E!:d l ttJ f6 36.a4
36.ttJd4 El:8e4 37. ttJ f3 E!:xc4 3 8.ttJxe l E!:xc2
39.tLlxc2 ttJ e4 40.f3 tLlg3 would lead to a
lost position. With such a weak king, White
has no chance of building a fortress, for
instance: 4 1 .a4 �e5 42.a5 bxa5 43.bxa5
� c3 44 .id3 � xa5 etc.
36 . . . ttJh5! 37.�d3 �g5 !
e
f
g
h
287
Chapter 9 - Materialism
8
8
7
7
6
6
5
5
4
4
3
3
2
2
b
a
c
d
e
f
g
h
The decisive regrouping.
38.lDxd6 lD f4 39.'\Wf3 �xd l 40.lDxe8 �e l
White may as well resign.
a
b
c
d
e
f
g
h
36.'=Yc3 V:Yxflt 37.i>hl ges
Or 37 . . . �b8 3 8.c5 lD d3 39.lDc4 (39.�b l
lD e l ) 39 . . . lD xc l 40.ltJxd2 lDxa2 would win
faster according to the silicon brain.
32 .. J�d4 33.'=Yb2 tlJe5!
8
7
38.c5 gxa2
5
39.gel ga3
6
The simplest.
4
Instead 39 . . . <j;>h7 40.cxd6 lD f3 would bring
down the curtain at once, but there was time
pressure . . .
3
2
40.'=Yxa3
b
d
f
g
h
A similar scheme: Black sacrifices a pawn or
two but penetrates into the opponent's camp
and attacks the king.
a
c
e
34.tlJxa4 '=Yf4
Now we can see that the pawn on h4 plays
an important part in the attack, preventing
35 .g3.
35.tlJxb6
There is nothing better left.
35 .. J�d2
40.'I!Mc l was slightly more resilient, but after
40 . . . <j;>g7 4 1 .�e2 �c3! 42.'1!Me l 'I!Md4 White is
still doomed.
40 ... V:Yxel 4I .i>gl '=Yd2!
8
7
6
5
4
3
2
The beginning of the end.
a
b
c
d
e
f
g
h
288
King's I ndian Warfare
This precise move puts the opponent in a
kind of zugzwang.
42.'lMa4
The following game is really chaotic. I
managed to fend off the attack and stack up
extra material. But like many good thrillers,
there is a great twist at the end.
What else?
Evgeny Postny - Ilya Smirin
42 ...'lMe3t 43.<i>hl llJ f3! 44.'lMxe8t
Or 44.gxf3 Wxf3t 4 5 .mgl Wg3t 46.mh l
El:el 47.Wb5 Wxh3t 48.mgl Wg3t 49.mh l
h3, with inevitable checkmate.
Acre 20 13
l .d4 d6 2.llJf3 llJf6 3.c4 g6 4.llJc3 .ig7 5.e4
0-0 6..le2 e5 7..le3
8
44...YMxe8 45.gxf3 'lMel
8
7
6
5
6
7
5
4
3
2
3
4
1
2
1
a
b
c
d
e
f
g
h
White resigned. To the credit of David
Anton Guijarro, after this tough loss he won
his final two games and took second place in
the tournament.
0-1
a
b
c
d
e
f
g
h
A very popular continuation. White delays
castling and tries to arrange his pieces first.
7... llJg4
Starting immediate action against the bishop
on e3.
The main point of 7.ie3 can be seen in the
variation 7 . . . ttJc6?! 8.d5 ctJ e7 9.ctJd2 and
White has obtained an ideal setup - after an
eventual c4-c5 the d2-knight will reach the
very desirable c4-square at once.
I played 7 . . . ttJbd7 against Gelfand in a game
you will find elsewhere in the book (page 26 1 ) .
7 . . . c6, 7 . . . We7 o r 7 . . . h6 (with the idea 8.dxe5
ctJ g4) are other continuations here. The last two
moves have a dubious reputation nowadays.
Chapter 9
-
8 ..ig5 f6 9 ..ih4 g5
The other popular and logical continuation
is 9 . . . 4J c6 1 0.d5 4J e7 1 1 .4J d2 4J h6, managing
without the double-edged . . . g6-g5 . I usually
prefer the move I made in the game.
.
IO .ig3 !iJh6
8
.� • .a.
��-%
7 ., -.,--t'""
-----Y-B
�
��?i
%�F�
-"
r-- - - %�
6
•
� r�. % r�
--- %•.�•.
5
� �r -�
4
"•
�
� �8"R
%
3 � � Btb�
d!i\i:J'· ·
�me?{� -�%-�
�
�
289
Materialism
In my opinion it's the right decision to open
up the centre and to start piece play.
1 1 . . . 4J c6 1 2.d5 4J d4?!
This was played by Kasparov once:
1 3 .4Jxd4 exd4 1 4.�xd4 f5 1 5 .�d2 f4 1 6.ih2
4Jf7 1 7.h4 h6 1 8.hxg5 hxg5 1 9.g3 ?
However, after the correct 1 9 .f3 4Je5
20.0-0-0 White would be better, as future
practice has shown.
!
21
8,1�� ����
'§ � · �
�
a
b
c
d
e
f
g
h
Black's setup (c7-d6-e5-f6-g7-h6) looks
slightly curious, and reminds one of checkers.
But it has a clear idea: to restrict the activity of
the bishop on g3 and to strengthen the pawn
on e5. Black's pawn formation on the kingside
is not fixed and may become quite mobile, as
actually happens in the game.
l l .h3
White wants to keep his knight on f3 hence this prophylactic move against . . . g5-g4.
On a few occasions I have faced the two other
main continuations - l l .d5 and l l .dxe5 which both lead to different types of positions.
a
b
c
d
e
f
g
h
1 9 . . .f3! 20.ixf3 4J e5 2 l .ie2 g4
Very nice: now both white bishops, especially
the dark-squared one, are inactive, to put it
mildly. That more than compensates for two
sacrificed pawns.
22.ig l c5
Black went on to win in great style . . .
2 2 . . . c 5 23.dxc6 bxc6 24.0-0-0 ie6 25 .�xd6
�g5 t 26.�b l �g6 27.�a l �ab8 28.�h5
�xh5 29.�xe6t �h8 30.�e7 4J f3 3 l .ixf3
�xf3 32.4Ja4 �e8 33 .�xa7 �e5 34.�b6
�xe4 3 5 .�b4 �c2 36.�b l �d3 37.�c5 �e2
3 8 .�h 5 t �g8 39.�xg4
Here is one example: l l . .. dxe5 1 2 .�xd8 �xd8
1 3 .4Jd5 4J a6 1 4 . 4J e7t �f7 1 5 .4Jxc8 �axeS
1 6.�c l 4J c5 1 7.4Jd2 aS 1 8 .f3 ifs 1 9.if2 �d7
20.h4 g4 and Black was fine; (Y2-Y2, 58) Roiz
- Smirin, Tel Aviv 2002.
l l . .. exd4
a
b
c
d
e
f
g
h
290
King's Indian Warfare
39 . . .'�Mxb 1 t
0- 1 Speelman - Kasparov, Madrid (rapid)
1 98 8 .
ltJd4 22.ig5 �e8 23Jhe 1 �g6 24.<;t>h 1 E:ae8
0- 1 Gershon - Smirin, Israeli Team
Championship 2007.
12.lLlxd4 tLlc6
13 ... tlJxd4 14.'l;Vxd4 f5
8
This leads to a very dynamic position in
which concrete variations will prevail over
positional considerations.
6
IS.'i;VdSt <i>h8 16.hxg5 'i;Vxg5
7
5
8
4
7
3
6
2
5
b
a
d
c
f
e
g
h
13.h4
The most principled approach - White
wants to make use of the opponent's weakened
king shelter.
Once I won a near miniature after:
1 3.0-0 ltJxd4 1 4 .'t!Mxd4 f5 1 5 .�d5t <;t>h8
1 6.f4 id7!? 1 7.�d2 fxe4
4
3
2
a
b
c
d
e
f
g
h
17J�h5?!
Not the best decision - the rook is kind of
hanging in the air now.
I would suggest 1 7 .c5 id7 1 8 .cxd6 ic6
1 9.�a5 cxd6, with a tense and interesting
struggle. Optically I like Black's chances here
because White's king is still stuck in the centre.
17 ...'l;Vg6 1 8.'i;Vd2
a
b
c
d
e
f
g
h
1 8 .ltJxe4?
Correct was 1 8 .fxg5 ltJ f5 1 9.if4, with a
sharp and unclear situation. Now there
comes severe punishment.
1 8 . . . ttJ f5 1 9 .ih2 gxf4 20 .ixf4 ic6 2 1 .if3
Now queenside castling is possible, but
White has no time for it: 1 8 .0-0-0 fxe4
1 9.ltJxe4 (somewhat better is 1 9.E:dh 1 e3!
20.ltJd 1 , but that too does not solve all White's
problems) 1 9 . . . ie6 20.�d3 ifS+
1 8 ... fxe4 19.tLld5
Evgeny tries to complicate matters, but
objectively his situation is already suspect:
he is a pawn down and the black pieces
29 1
Chapter 9 - Materialism
may parry the threats against the king while
simultaneously counterattacking. But those
are j ust words, which need to be backed up by
strong play.
1 9.0-0-0? was just losing due to 1 9 . . . e3!
20.'1Wxe3 :ge8 2 1 .'1Wd3 :gxe2 22.'1Wxe2 ig4 (the
rook on h5!).
Nevertheless, it seems to leave White facing
defeat - Black prevails in all variations. Such
as:
24.1Wc2 :gf7!
19 ... lLJf5 20.0-0-0 c6
The exchange 20 . . . ttJxg3 2 l . fxg3 seems to be
premature - the f5-knight looks strong at the
moment.
2 I .lLJf4
Houdini briefly suggests 2 1 .ttJc7. Come on!
2 1 . ..%¥£6
A battery against the king on c 1 has been
built. White's chances lie only in the attempt
to create something serious against my own
king.
22J�dhl
8
a
b
c
d
e
f
g
h
2 5 .:gxg7t
Or 2 5 . ttJ h5 '1Wxb2t 26.'1Wxb2 ixb2t
27.<kt>xb2 i>xh7-+ .
Or 25 .:g7h5 ttJd4 26.'1Wh7t i>f8 27.fxe3
tLlxe2t 28.tLlxe2 '1Wxb2t 29.i>d 1 '1Wa 1 t
30. i>d2 '1Wxa2t.
25 ... '1Wxg7 26.ttJh5 ttJd4 27.tLlxg7 tLlxc2
28.tLle8 exf2 29.ttJxd6 ttJd4
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
22 ... h6?!
Too cautious a move for such a sharp
situation. However, it would probably be
made by the vast majority of players. It would
require deep penetration into the position, and
a lot of guts, to play the computer's suggestion:
22 . . . e3! 23 .:gxh7t <kt>g8
30.id3 f1 ='�Wt 3 l .:gxfl :gxfl t 32.ixfl ttJ f5
etc.
a
b
c
d
e
f
g
h
23.�h4?
Now Evgeny in his turn did not guess
correctly. He should have moved the bishop
more modestly: 23.ih2! When Black has a
choice:
292
King's I ndian Warfare
After the prophylactic and wrong 23 . . . @g8?
24.g4 ltJd4 2 5 .El:g l ! White's attack suddenly
becomes too strong: 25 . . . ltJ e6 (or 25 . . . El:f7
26.g5 hxg5 27.El:hxg5) 26.ltJxe6 ixe6 27.g5
hxg5 28.El:hxg5 El:f7 29.ih5 El:d7 30.El:g6+23 . . . e3! ? 24.fxe3 ltJxe3 deserves attention, but
after:
Of course, I have shown these variations for
the sake of seeking the truth. It would be
totally unrealistic to see them during the game.
23 ...Wff7 24.�g5 �g8
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
a
h
2 5 .ig l ! ltJg4 26.g3 @g8 27.El:5h4 �g5
28.id4! White would hold initiative.
23 . . . ltJ d4 24.id l @g8 25.g4 ltJ f3
25 . . . �f7 26.@b l �xc4 27.El:g l , with a mess.
26.ixf3 exf3 27 .El:g 1
b
c
d
e
f
g
h
White cannot break the shield around the
black king with pieces only, and on the very
next move the g-pawn joins the attack - but it
is already too late.
25.g4 hxg5 26J!xg5 e3?
Too optimistic.
Correct was the calm 26 . . . ltJd4 27.ltJh5 ltJ e6
28.El:f5 �e7 29.lbxg7 lbxg7 30.El:xf8t �xf8
3 1 .�h6 ie6 32.g5 (32.�h7t @f7 33 .El:h6
�e7 34.�g6t @f8) 32 ... �f4t winning.
8
7
a
b
c
d
e
f
g
h
White seems to have full compensation for
the pawns. Still, I believe Black may hold his
own here:
27 . . . �d4 28 .g5 �xc4t 29.@b l if5t 30.@a l
ie5 3 1 .gxh6t @h7 32.ltJh3! ixh2 33.lDg5t
@h8 34.El:xh2 �d5
And so on.
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Chapter 9
-
293
Materialism
8
27.'iNc3?
7
White returns the favour.
Correct was the simple 27 .fxe3 when there is
a split:
27 .. Y!le7
28.!hg7t
fixg7
(28 . 4Jxg7?
29.4Jg6) 29.gxf5 I would prefer to be White
here.
.
.
27 . . . 4J g3 28 .�d3 4Jxh l 29.fih2 �d8 30.'Wih4!
Neither of us saw this. 30 . . 4J f2 3 l .�xg7t
fixg7 32.fixd8t fif8 33 .fig5t With a
perpetual.
6
5
4
3
2
1
.
27...¥Ne7!
From now until the 4 1 st move, I suddenly
started to play flawlessly.
a
b
c
d
e
f
g
h
37.¥Nh2 :ghs 38.tl)xc6 .if5t 39.\t>h3 bxc6
4o.YNxn ghst
Perhaps 40 . . . c5 was the simplest solution.
41 .\t>a4 \t>g6?
In an attempt to coordinate Black's forces,
I underestimated the potential of the two
remaining white pieces. Also I have to give
credit to Postny for not losing his fighting
spirit, but setting and exploiting the trap
instead.
28J;hh5 exfl 29.gxf5 :gx5
8
7
6
5
4 1 . . �d8 42.fih2 �f8 43 .fih8t <;t>f7 was one
of the winning options. Black's pieces will soon
develop a deadly attack against the opponent's
king.
.
4
3
2
1
a
b
c
d
e
f
g
h
30.:ghst \t>xhs 3I .tl)g6t \t>h7 32.'iNh3t
�h6 33.tl)xe7 :gxg5 34.%Vd3t \t>g7 35.\t>c2
�e6 36.%Vxd6 \t>£6!
The end of a forced line. Now White has to
give up the knight to neutralize the f-pawn,
and Black gets an awful lot for the queen.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
294
King's Indian Warfare
42 ..ih5t!!
An excellent trick I have to admit.
42.. J�xh5
42 . .. �xh5 43 .Wh2t �g6 44 .Wxb8 or
42 . . . �h7 43 .Wxa7t ig7 44.Wxb8 !hh5
would be no better.
In the final game of this chapter, I again manage
to get the advantage of the two bishops, and
get a lot out of the light-squared bishop.
White tries to gain counterplay through pure
aggression. In the chaos that ensued, I gained
a material advantage and eventually managed
to win the game after a few twists and
turns.
43.1Mg3t <i>h? 44.1Mxb8
Ronen Lev Ilya Smirin
If Black were able to keep both remaining
pawns he would still have winning chances,
but that's impossible.
44....ig7 45.1Mxa7
47.1Mxc6 �xb2
�h2
46.1Mc5
.ig6
-
Ramat Can 1 992
I .d4 tljf6 2.c4 g6 3.tlJc3 .ig7 4.e4 d6 s.tlJf3
0-0 6..ie2 e5 7.d5 tlJ a6
One possible way to react to the Petrosian
System.
8
7
More popular moves here are 7 . . . a5 or 7 . . . ltJ bd7
in connection with 8 .ig5 h6 9.ih4 g5 l O.ig3
ltJ h 5 . Now a possible continuation is: l l .h4
ltJxg3 1 2. fxg3 gxh4? (much better is 12 . . . g4)
1 3 .ltJxh4 Wg5 1 4.ig4! ltJ f6 1 5 .ixc8 Wxg3t
1 6.�fl �fxc8 1 7.ltJf5 Wf4t 1 8 .Wf3 Wxf3t
1 9.gxf3 With an overwhelming advantage; the
knight on f5 j ust paralyses Black.
6
5
4
3
2
a
b
c
d
e
f
g
h
48.1Mhl t .ih6 49.c5 �xa2t 50.<i>b5 �b2t
5 I .<i>a5 �bl 52.1Mh3 �al t
A draw was agreed. I remember my
disappointment after the game, but c'est la
vie. Evgeny, for his part, fully deserved to get
something out of it.
lf2-lf2
8 ..ig5
The more principled continuation seems to
be 8.ltJd2!?, stressing the drawbacks of Black's
7th move ( . . . a7-a5 is impossible now, and
intending to answer 8 . . . ltJc5? with 9.b4) .
a
b
c
d
e
f
g
h
(a) After the logical 8 . . . ih6 White has a strong
295
Chapter 9 - Materialism
and ambitious measure at his disposal: 9 . h4!
'Llc5 1 O.h5 ixd2t 1 1 .Wxd2 'Ll fxe4 1 2.'Llxe4
ltJxe4 And now after both 1 3 .We3 or 1 3 .Wh6
we would probably not find many volunteers
to play as Black here. Indeed, to be under
attack with weak dark squares around the king,
and without counterplay, is too high a price for
a mere pawn.
(b) 8 . . . 'Lle8 9J�b 1
Also now 9 . h4!? is an attractive option: 9 . . .f5
1 O.h5 with a confrontation on the kingside.
After 1 0 . . . 'Ll f6 1 1 .hxg6 hxg6 1 2.exf5 gxf5
1 3 . 'Ll f3 I would probably prefer White's
chances in this complex position. The rook
on h 1 has become very active without
making a single move.
9 .. .f5 1 O.b4
This happened in one of my games:
1 o . . 'Ll f6 1 1 .0-0 c5 1 2.a3
1 2 .dxc6!?, to make the game more open by
means of 1 2 . . . bxc6 1 3 .b5, deserves attention.
1 2 . . . b6
.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Black reacts by analogy with the variation
7 . . . 'Ll bd7 8.ig5 .
l l .tiJd2
If 1 1 .h4 Black may reply 1 1 . . . 'Llxg3 1 2. fxg3
gxh4!? 1 3 .'Llxh4 Wg5 . Now we can see the
difference between the knight positioning on
a6 instead of d7: 1 4.ig4? is impossible and
Black stands well according to practice.
l l . .. tlJf4 12.0-0 f5 13.exf5 tlJxe2t
It's important to exchange the strong bishop
and leave White with the passive one.
14.Wfxe2 .ixf5 1 5.tlJde4
a
b
c
d
e
f
g
h
Black is OK: his position is solid on the
queenside and active on the opposite wing.
1 3 .id3 f4 1 4 .ie2 g5 1 5 .g4!? h5 1 6.h3 E!f7
1 7.f3 'Ll b8 1 8 .bxc5 bxc5 1 9 .Wa4 id7 20.Wc2
ic8 2 1 .Wa4 id7 22 .Wc2 ic8
Yz-Yz Illescas Cordoba - Smirin, Oviedo
(rapid) 1 993.
8 ... h6 9 ..ih4 g5 10 ..ig3 tlJh5
White's knight has occupied the excellent
outpost on e4. So does he have a positional
296
King's Indian Warfare
advantage? I do not think so. White's bishop
is badly placed on g3, and to free it White will
have to play f2-f3 and/or h2-h3. We will soon
find out where this may lead.
1 5 Wfe8 16JUel
...
In the following game of mine we can see the
ideas for both sides:
1 6.f3 Wg6 1 7.h3 h5 1 8 . ltJ f2?! lD c5 1 9.gad l
1 9 .b4? ltJ d3
1 9 . . . a5
Now it's obvious that Black has an awesome
position: all his pieces are well placed and
. . . g5-g4 is in the air.
20.ih2 g4!
As usual, White lacks one tempo - if he
could play g2-g4 now and reoccupy e4 with
the f2-knight he would be fine. " If" is the
key word here.
2 l .fxg4 hxg4 22.hxg4 id? 23.ltJce4 lDxe4
24.Wxe4 Wxe4 25 .ltJ xe4 ixg4 26.gxf8 t gxf8
27.gfl ?
After 27.ge l White could count on drawing
this worse endgame.
Now, without rooks, nothing can prevent
the black bishops from showing their full
strength.
a
b
c
d
e
f
g
h
27 . . . gxfl t 28.Wxfl ifS
That's it: the knight has to retreat.
29.ltJd2 id3t 30.Wel e4 3 l .c5 ixb2 32.cxd6
cxd6
c
d
e
f
g
h
0- 1 lvanov - Smirin, Los Angeles 2000.
White resigned due to 33 .ixd6 e3 34.ltJ f3
ic3t 35.Wdl e2t .
In my opinion better is:
1 8 .a3
Preparing b2-b4. In this case Black may
continue:
1 8 . . . ltJ b8! ?
Premature is: 18 . . . g4?! 1 9. fxg4 hxg4 20.h4
a
a
b
b
c
d
e
f
g
h
This is a very desirable setup for White: his
knight is still on e4, and thanks to the pawn
on h4 it may go to g5 in some cases. Black
has nothing better than to exchange it, but
it damages his pawn structure: 20 . . . ltJ c5
2 1 .ltJxc5 dxc5 22.gad l a6 (22 . . . e4? 23.ltJ b5)
23.gfe l And White will restore the knight's
presence on e4. The attempt 23 . . . e4?! is not
good due to 24.We3! id4 2 5 . gxd4 cxd4
26.Wxd4, with overwhelming compensation
for the exchange.
297
Chapter 9 - Materialism
1 9 .c5 ttJd7
This knight is heading toward f6.
20.cxd6 cxd6 2 l .�b5!?
2 l .if2 ttJ f6 22Jhc l (or 22.ttJ xf6t �xf6
23.ttJe4 �f7 and Black has nothing to worry
about) 22 . . . ttJxe4 23.fxe4 (23.ttJxe4?! g4
is dangerous for White) 23 . . . id7 and I
would even prefer Black here - he has more
dynamic possibilities.
2 l . . . g4 22.fxg4 hxg4 23.h4 ttJ f6
With a double-edged position.
Perhaps Lev should have preferred more static
play, for instance: 1 9.a3 ie7 20.b4 h5 2 1 .<it>h l
(2 l .b5 ttJ b8 22.c5 h4 23.ih2 g4 24. hxg4
ixg4 looks good for Black) 2 l . . . �af8 (or
2l . . . h4 22.ih2 g4 23.hxg4 ixg4 24.�e3!?)
22.�e3!?
16 VHg6 17.YHd2
•••
Stopping . . . h6-h5 for a while.
17 JU7
••
With the idea of bringing the bishop to e7
via f8, and carrying out the desirable advance
of the h-pawn.
1 8.h3 .lf8
8
7
6
5
4
a
b
c
d
e
f
g
h
And despite the menacing look of Black's
army on the kingside, White has his trumps.
His position on the kingside is solid enough,
and the black knight is still inactive.
Meanwhile White is going to start activity
on the queenside, connected first of all with
the usual c4-c5 . In other words, this is a
complex position of approximate dynamic
balance.
19 ... dxc5
3
I wanted to keep the knight - it seemed to
me that its potential j ump to b4 or c5 (after
. . . c5-c4) might be annoying for White.
2
1
a
b
c
d
e
f
g
h
19.c5
This may not be bad, but it is a strategically
risky decision - now the position becomes
much more open and the pair of black bishops
will appreciate that.
1 9.f3 ie7 20.if2 h5 2 l .g4?! This is dubious
here due to the vulnerability of the pawn on f3
(the drawback of 1 7.�d2) . 2 l . . . hxg4 22.hxg4
ixe4 23.fxe4 (a must) 23 . . . �af8 etc.
19 . . . ttJxc5 20.ttJxc5 dxc5 2 l .ixe5 c4 was a
good alternative. White's pieces are centralized,
but the black forces are quite active and his
king feels relatively safe despite the bishop on
e5 owning the key diagonal.
20.he5 ges 21 .lLlg3 .ld3
Black could continue 2 l . . . ttJb4 22.ttJ xf5
�xf5 23 .ig3 �d8 with a balanced position.
But I saw no reason to part with my beloved
pair of bishops.
298
King's Indian Warfare
23.f4?
Tantamount to hara-kiri; this move ruins
White's position.
The only way to proceed was 23.�e3! �xe5
(here 23 . . . c4? is bad: 24.�ae 1 lDc5 25 .ixc7!)
24.Wfxd3 Wfxd3 25.�xd3 c4 26.�d4, with
more or less equal chances in the endgame.
23 ... c4 24.�ce4 �f7
a
b
c
d
e
f
g
h
22.a3!
Probably Ronen underestimated this move.
Now White's minor pieces are hanging in the
air, despite seemingly being the active ones.
25.�hl
A timely prophylactic move.
The possible complications after 22.�ad 1 ?!
lDb4 23 . .ixc7! ? �xe 1 t 24.�xe 1 �xc7 25 .a3
�d7 26.axb4 cxb4 27. lD a2 (what else?) 27 . . . a5
28.�e6 Wff7 would be clearly in Black's favour
- the knight on a2 is misplaced to say the least
(note 29.Wixd3? is bad after 29 . . .Wfxe6) .
25.fxg5 would not save White either:
25 . . . Wfb6t! 26.�h2 (26.4Jf2 �xe5 27.�xe5
�xf2 28.Wfxf2 ic5 29.�e8t �h7) 26 . . . �xe5
27.4J f6t �xf6 28.gxf6 (28 .�xe5 �f2)
28 . . . Wfxf6, with a decisive gain of material.
25 ... gxf4
White has to give up some material.
22 .. J3fe7
After 22 . . . c4 23.id4! �xe 1 t 24.�xe 1 lDc5
25 .�e8 White's active pieces cause Black
trouble, for instance: 25 . . . 4J b3 26.Wie3 lDxd4
27.Wfxd4 b6 28 .�e6 Wig? 29.Wig4, with the
initiative.
a
b
c
d
e
f
g
h
28 ... fxg3?!
a
b
c
d
e
f
g
h
Sloppiness. Instead 28 . . . Wfxg3 29 .ie5 .id6
would win effortlessly.
299
Chapter 9 - Materialism
32.'l;Ycl??
Lev failed to spot my mistake.
This incredibly passive move does not
promise any chances for survival and even
loses on the spot.
He should have played: 29 . .ie7!
Meanwhile the natural 32.We3 Wb l t 33.Wg l
Wxg l t 34.<it>xg l .id3 3 5 . El:xc7 b6 36.El:xa7
would lead to a probably drawn endgame - a
rook plus a pawn may match the opponent's
bishop and knight.
32 tlJd3 33.'l;Yal tlJ £2t 34.®gl
•••
b
a
c
d
e
f
g
h
29 . . . .ixe7 (after 29 . . . .ig7 then 30.'1W f4 is
strong - White has counterplay) 30J�xe7
Wf6 3 1 .We l Wfl t 32.Wxfl .ixfl And in
this ending White has some drawing chances
thanks to the active rook.
29 tlJc5?
..•
A second consecutive error, and a much
more serious one.
After the correct 29 . . . .id6 30 . .ie5 , and only
now 30 . . . tt:Jc5 3 1 ..ixd6 cxd6, Black would
obtain an easily winning position.
30 ..ie7 .ixe7 3 1 .�xe7 .ie2!
The only move, by the way.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
a
0-1
b
c
d
e
34 tlJxh3t 35.gxh3 'l;Yb6t
•••
f
g
h
Chapter 1 0
Endgames
I had many games against the great Vassily Ivanchuk,
although in this book they only made it into the notes
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
8
... 8
7
7
6
6
5
5
4
4
3
3
2
2
a
b
c
d
e
f
g
h
h
5
4
3
2
a
Endgame magic!
(see page 303)
b
c
d
e
f
g
h
g
h
f
g
h
4
3
2
f
e
5
3
e
d
6
4
d
c
... 8
7
5
c
b
here! (see page 326/27)
6
b
a
I missed not one but two wins
Black to play
(see page 3 1 9/20)
... 8
7
a
h
6
3
g
g
7
2
f
f
... 8
4
e
e
(see page 325)
5
d
d
What had White missed?
6
c
c
endgame (see page 307)
7
b
b
Accuracy is important in the
... 8
a
a
2
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
Black has a nice tactic
How can Black prevent the
But later I was lucky thatBrunello
(see page 305)
counterplay (see page 32 1 )
missed the best move on move 40
(see page 328)
Chapter 1 0
-
303
Endgames
The King's Indian is a combative opening in
which the middlegame can go on and on and
on. Frequently, when an ending is reached, it is
so non-standard that it still has many dynamic
traits similar to the middlegame, as you will
see from the four games I have chosen from
my own collection.
But first, here is a great modern example of a
dynamic KID endgame.
Fabiano Caruana - Hikaru Nakamura
Wijk
aan
b
a
c
d
e
f
g
h
Winning directly, as 57.� h 1 �d4 wins a
piece and 57.g3 loses to 57 . . . �f8 5 8 .ih l
�f2t 59.�c3 �e2 60.ltJxc5 bxc5 6 1 .�d2
�xd2 62.�xd2 �f6 63 .�e3 �g5, followed by
. . . ic2 and it's all over.
Zee 20 13
8
7
6
5
57.®d3
57.�c3 would also have lost, but slowly. For
example after: 57 . . . ixe4 (57 . . . hxg2 5 8 .ixg2
ib4t 5 9.�d3 �f2 60.ih l �b2 also wins)
5 8 .ixe4 h2 59.g3 �f2 60.g5 �e2 6 1 .ih l
�e3t 62.�c2 id6 and White is too passive to
hold in the long term.
4
3
2
1
a
b
c
d
e
f
g
h
This is the type of endgame any King's
Indian player would be aiming for. White is
passive, desperately holding on, while Black
enjoys lots of dangerous ideas with his strong
bishop pair. In the game Nakamura found a
wonderful dynamic solution.
54... g4!!
Undermining the white defence on the light
squares.
55.hxg4 h3! 56J�dl
56.gxh3 �f8 57.�d3 �f4! and White loses
a piece.
56.. J�f8?!
The best move was: 56 . . . h2!
57 ... h2 58J�hl .igl
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
59.®e2 .ixe4 60.�xe4 ®f6 6I ..if3 gds
0-1
304
King's I ndian Warfare
8.d5 h6 9 ..ie3 �c5 lO.'iMc2 c6
My Games
Yuri Yakovich - Ilya Smirin
Saint-Vincent 2000
This game was played in the last round of the
very first European Individual Championship
in the Italian resort of Saint-Vincent. Today
such championships have become part of the
annual chess calendar. My opponent in this
game was an experienced Russian grandmaster.
Nowadays Yuri is well known as a chess coach
as well.
The variation 1 0 . . . tiJg4 1 1 .ixg4 ixg4 1 2. f3
id7 1 3 .h5 g5 1 4.g4 shows one of the main
ideas of 7.h4 - to nip in the bud Black's play
on the kingside, and leave him with a half­
dead bishop on g7.
l l .h5
l.d4 �f6 2.c4 g6 3.�c3 .ig7 4.e4 d6 5 ..ie2
0-0 6 ..ig5 � a6
b
a
d
c
f
e
g
h
l l . .. cxd5
1 1 . . . g5 was the choice of Garry Kasparov,
and he managed to draw with creative play:
1 2.f3 a5 1 3 .g4 id7 1 4 .f:iJh3 a4 1 5 .�d2 cxd5
1 6.cxd5 �a5 1 7. f:iJ b 1
a
b
c
d
e
f
g
h
A relatively modern way to fight against the
Averbakh System. It was first employed by lgor
Glek in 1 98 8 .
7.h4
This aggressive move became popular in the
1 990s. Its main proponent was Evgeny Bareev.
a
7 ... e5
7 . . . c5 8.d5 f:iJ c7 is another way to continue,
but I preferred a more typically King's Indian
move.
b
c
d
e
f
g
h
1 7 . . . f:iJ fxe4!? 1 8 .fxe4 CiJxe4 1 9 .�xa5 El:xa5
20.tiJc3 tiJg3 2 1 .El:g1 CiJxe2 22.<j;Jxe2 e4
23.El:ac l f5 24.gxf5 El:xf5 25.tiJf2 ie8 26.El: h 1
ib5t 27.tiJxb5 El:xb5 28 .El:c8t <j;J h 7 29.El:d1
El:xb2t 30.El:d2 a3; (V2-V2, 48) Bareev -
305
Chapter 1 0 - Endgames
Kasparov, Linares 1 992. Despite that game, his
move is not to my taste - I usually do not like
to allow White to seal the queenside in such
a way.
1 5 J�ac8!
..
I prepared this move at home.
1 5 . . . lD h5 ?! is too early:
12.cxd5 VMaS 13.i.d2 i.d7 I4.hxg6 fxg6
15.8
After this move the central pawn on e4 is
secured, but in the future the weakening of the
black squares might tell. Besides, the knight on
f6 now gets access to hS, from where it aims to
go to g3 or f4. I would prefer:
1 5 J�b 1 Wb6 1 6.ie3 ( 1 6.b4? allows a big
punch:
a
b
c
d
e
f
g
h
1 6.b4! Wxb4 1 7.4Jb5 Wa4 1 8 .Wxa4 lDxa4
1 9 .4Jxd6 was played in Yakovich - S. Bekker
Jensen, Gothenburg 2000, which Yuri Yakovich
won convincingly. Indeed, exchanging the
b2-pawn for the pawn on d6 is clearly in
White's favour.
8
7
6
5
4
3
16J�bl
2
b
a
c
d
e
f
g
Mter 1 6.b4? Wxb4 I was intending to
sacrifice my queen:
1 7.4Jb5
h
1 6 . . . 4J cxe4 1 7.lDxe4 E!:ac8! 1 8 .4Jc3 4Jg4-+ The
pawn on f2 is defenceless and with it the white
king.) 1 6 . . . a5 1 7. 4J h3 With complicated play
in Hauchard - Hebden, Cappelle la Grande
1 998.
8
7
6
5
4
a
3
2
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
1 7 . . . Wxb5 ! 1 8 .ixb5 ixbS
It's clear from the very first glance that
Black has overwhelming compensation
and is about to win. Still, I would like to
demonstrate one pretty line:
306
King's I ndian Warfare
1 9.�b l ltJ d3t 20.c;t>e2 ia6 2 1 .c;t>e3
2 l .ixh6 ixh6 22.!hh6 c;t>g? 23 .!�h4 El:c 1-+
More than enough for one small pawn!
19 ... tiJh5�!
Better options were 19 ... El:h8 20.El:xh8 �xh8
2 l .i.fl b5! or the immediate 19 . . . b5!? - "wide"
play was required to disturb the opponent
from both wings.
2o.YNd2 ghs 21.�£1
To develop the knight from gl .
21. ..\'Nb6
b
a
c
d
e
f
g
h
2 1 . . .ltJ g4t! 22.fxg4 ltJ f2 23 .ie l
23 . ltJ f3 ltJ xg4#
23 . . . ltJxg4t 24.c;t>d2 h5!
Despite having a queen for only a bishop,
White is absolutely helpless.
16 ...\'Nds I7.�xh6
Otherwise White would suffer with even
material ( . . . ltJh5 is coming) . Now he has an
extra pawn as consolation.
2 l . . . ltJg3 ? does not work: 22.El:xh8 �xh8
23.�g5
22.tlJge2 tlJf4
8
7
6
5
4
3
2
17 ...�xh6 1 8.gxh6 �g7 19.ghl
8
1
7
6
a
b
c
d
e
f
g
h
5
23.gxhs tlJcd3t 24. �dl gxhs 25.tlJxf4
tlJxf4 26.�c2
3
26 ... gh2 27.gcl �h3
4
Worse is 26.g3 �g l 27.gxf4 ih3!+.
2
The pawn on g2 is under heavy fire, but
White can withstand that.
a
b
c
d
e
f
g
h
So, Black's compensation consists of: a lead
in development; White's king is stuck in the
centre of board; the dark squares around it
are weak; and Black will soon seize the h-file.
28.tlJa4�
This gives Black new hope, whereas 28.ltJe2
ltJxg2 29 .�g5 ltJ e3t 30.c;t>b l ltJxfl (30 . . . ixfl ?
loses nicely to 3 l .�e7t c;t>h6 32.El:c8! with
unavoidable checkmate) 3 l .�e7t c;t>h6
32.�h4t led to perpetual check by force.
Chapter 1 0
-
307
Endgames
8
28 Y!Yd4! 29.Y!Yxd4 exd4
•••
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Black's advantage is obvious in this endgame
- White's minor pieces are passive and the
extra pawn on g2 will fall next move.
30.Wb3
It's true that after 30.e5 !? the g-pawn would
be held, and after 30 . . . id7 (or 30 . . . dxe5
3 1 .<±>b3 ixg2 32.ixg2 E1:xg2 33.d6 <±>f6
34.tt::l c5) 3 l .<±>b3 ixa4t 32.<±>xa4 dxe5 33.d6
<±>f6 34.d7 <±>e7 3 5 .E1:c5 White would have
good drawing chances, but such moves can
only be made by computers.
30 ...hg2
Less good was the immediate 30 . . . b5 ?
3 l .ixb5 ixg2 32.id7! \t>f6 (3 2 . . . ixf3?
33.E1:fl ) 33.ig4 <±>e5 34.E1:c8 and White is OK.
3 1 .bg2 b5
I liked this zwischenzug during the game, but
the simple 3 l . .. E1:xg2 32.E1:c7t mf6 33 .E1:xb7
E1:f2 was probably stronger. I doubt White
can hold: 34.E1:xa7 (34.\t>c4 d3 3 5.\t>b3 E1:xf3
36.ltJc3 g5 37.E1:d7 g4-+) 34 . . . E1:xf3t 35 .\t>c4
d3 36. <±>c3 E1:fl ! and Black should be winning.
32.� b6 axb6 33 ..ifl d3
Or 33 . . . E1:f2!? 34.ixb5 E1:xf3t 35 .\t>c4 d3+.
b
a
d
c
e
f
g
h
In my opinion this endgame is quite
interesting and instructive. Dear reader,
I would like to spend some of your time
analysing it.
34.. J�� h l?
Now the bishop is pinned, but again White
can overcome that.
A more direct approach would do the job:
34 ... E1:f2! 3 5 .\t>d4 E1:xf3 36.e5 dxe5t 37.\t>e4
37.\t>xe5 d2 38.E1:dl E1:xfl 39.E1:xfl ltJd3t
40.<±>e6 ltJ e l and the d-pawn queens.
37 . . . E1:g3 3 8.\t>xe5 g5 39.d6
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
39 . . . <±>f7! 40.E1:d l
40.E1:c7t <±>g6! 4 l .d7 E1:e3t 42.\t>d4 d2-+
40 . . . \t>e8 4 l .b4 <±>d7
The resulting position is easily winning for
Black.
,
King s Indian Warfare
308
35J�dl ?!
Mter 3 5 .�b l lt:Je2t (35 . . . <;t>f6? 36 ..ixd3)
36.<;t>d2 <j{f6 37.�d l lt:J f4 38.<;t>e3 <;t>e5
39 . .ie2! �h2 40 ..ixd3 �xb2 4 l .�d2 White
,
could perhaps hold despite the opponent s
total domination of the dark squares.
35 ... �£6 36.�d4
Probably the sudden counterattack 36.<;t>b4
lt:J h3 37.<;t>xb5 lt:J f2 38.�xd3 �xfl (38 . . . tt:Jxd3
39 ..ixd3 �h3 40.e5t! dxe5 4 l . .ie4 and it looks
like a draw) 39.�d2 could have saved Yuri.
active and the usefulness of the bishop on fl is
not great, to put it mildly.
40.�e3 ltJe2 41 J�dl!
8
7
6
5
4
3
2
36 ... b4 37J�b l ltJe2t! 38.�e3
8
7
6
b
a
d
c
f
e
g
h
41. .. ltJf4?
5
An unfortunate waste of time, whereas:
4 1 . .. lt:J g3 42.<;t>f2 lt:Jxfl 43.�xfl �xfl t
44.<;t>xfl <;t>e5 45.<;t>el
led by force to a queen ending:
45 . . . <;t>f4 46.<;t>d2 <j{xf3 47.e5 dxe5 48 .d6 e4
49.d7 e3t 50.<;t>xd3 e2
4
3
2
a
b
c
d
e
f
g
h
38 ... g5?
Centralization with 38 . . . <;t>e5 was called for,
and after 39.a3 b3 40.�d l tt:J f4 4 l .�c l �h2
42 . .ixd3 �xb2 43 .-ifl g5 the dominance
,
of Black s forces would decide the outcome.
We were both in time trouble and this partly
explains our numerous mistakes in this
endgame. But speaking honestly, there were a
few too many mistakes.
39.�fl ltJf4
39 . . . <;t>e5 40.<;t>e3 b5 leads to a kind of
mutual zugzwang, and after 4 l . b3 (the only
move) 4 1 . .. lt:J c3 42.�e l lt:Jxa2 43.<;t>xd3 g4!
44.fxg4 �g l I would have retained a dose-to­
winning advantage - all the black pieces are
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
5 l .d8=Wi' e l =Wi' 52.Wi'f6t
52.Wi'xg5 ? Wi'e2t 53 .@d4 Wi'e4# A nice picture!
52 . . . <;t>g4 53.Wi'xb6 Wi'b l t 54.<;t>c4 Wi'xa2t
Black has excellent winning chances.
42J�bl d2
42 . . . <;t>e5 allows 43 . .ixd3 .
Chapter 1 0
But still possible was 42 . . . ttJe2 43.�d l tDg3.
-
309
Endgames
The last chance.
49 ..ie2??
43.<i>xd2 gh2t 44.<i>e3 <i>eS
8
Luckily for me, Yuri was the last one to
commit an error in this game.
6
49.ic8! �xf3t was required and the variation
50.�xf3 gxf3 5 1 .ih3 f2 52.a3 bxa3 53.bxa3
tDg6 54.ig2 ttJ f4 5 5 .ifl ttJh5 5 6.ig2 ttJf6
57.@xf2 ttJ xe4t 5 8 .@e3 ttJ f6 59.@d3 ttJ xd5
60.if3 ttJc7 6 1 . @c4 leads to a peaceful
outcome. That's not the case now.
7
5
4
3
2
49 ... tlJg2t 50. <i>d2
1
a
b
c
d
e
f
g
h
Despite being a pawn down, Black is still the
superior side in the endgame that has arisen,
but now his advantage is hardly enough to win.
50.@f2 ttJ f4 5 1 .fxg4 �h2t 52 .@g3 �g2t
53.@f3 !!xe2 was no better.
8
7
6
45 ..ia6 tlJg2t 46.<i>d3
Not 46.@f2? @f4 47.@g l @g3-+ .
5
46... tlJ h4
46 . . . @f4 47.�gl ttJ e3 (47 . . . @xf3 48.�fl t
@g3 49.�f6 with strong counterplay) leads
after 48.�xg5 ! @xg5 49.@xe3 �xb2 50.ic4
�c2 5 1 .f4 t @f6 52.ib3 to a fortress position
- draw.
4
3
2
1
b
d
f
g
h
A triumph of the knight over the bishop no one can protect the dark squares in White's
position.
a
8
7
6
c
e
52 ..ia6 <i>f4 53.e5
5
This is no worse than other moves - the
g-pawn will promote soon.
4
3
53 ... <i>xe5
2
Yakovich lost on time, but his position was
hopeless anyway.
1
a
b
c
d
e
f
g
h
0-1
310
King's I ndian Warfare
Mihail Marin Ilya Smirin
With the better position for White - the
activity of the black pieces is restricted,
Krasenkow - Smirin, Polanica Zdroj 1 99 5 .
-
Ramat Aviv 2000
l .c4 g6 2.tiJf3 .ig7 3.d4 tiJf6 4.g3 0-0 s ..ig2
d6 6.0-0 tlJ c6 7.tiJc3 a6
The most popular move - Black prepares to
play on the queenside by means of . . . b7-b 5 .
Later I mainly switched here to 7 . . . .if5 .
8.. J�b8 9.e4 b5
According to plan.
IO.e5
8
8
7
6
5
6
7
5
4
3
2
3
4
2
b
a
d
c
f
e
g
h
8.h3
A prophylactic move against . . . .ig4.
The immediate 8 .e4 .ig4 9 . .ie3 is also possible.
I was unable to equalize in the following rather
ancient game: 9 . . . ttJd7 l O.ttJe2 e5 1 1 .d5 ttJ e7
1 2. ttJ e 1 ttJ b6 1 3 .Ek1 c5 1 4. ttJ d3 f5 1 5 .h3
.ixe2 1 6.'Mfxe2 fxe4 17 . .ixe4 ttJ d7 1 8 .g4!
8
7
b
a
d
c
f
e
g
h
IO ... tiJd7
This move leads to a complicated position.
I have never liked the simpler 1 O . . . dxe5
1 1 .dxe5 'Mfxd 1 1 2J!xd 1 ttJd7 1 3 .e6 fxe6
14.cxb5 axb5 1 5 . .if4 etc. - even though Black
has decent results in practice.
l l .e6
White can fight for an advantage by means
of the more or less forcing line: 1 1 .cxb5 axb5
1 2.ttJg5 dxe5 1 3 ..ixc6 exd4 1 4.lt:Jxb5 E!:b6
1 5 .ttJxd4
6
8
5
7
4
6
3
5
2
4
3
a
b
c
d
e
f
g
h
2
a
b
c
d
e
f
g
h
31 1
Chapter 1 0 - Endgames
1 5 . . . 4J b8 1 6.lt:lde6 fxe6 1 7.'1Mfxd8 �xd8 1 8 ..ig2
and here the draw was concluded in the game
Nogueiras Santiago - Smirin, Istanbul (ol)
2000. Mter the game I had a feeling that in the
final position White could continue to play I did not particularly like Black's pawn
structure and the potential of the passer on a2.
following encounter:
1 5 . . .id7 1 6.lt:lf4 Wc8 1 7.lt:lde6 �f7 1 8 .lt:lxg7
cj;>xg7 1 9.b3 c5 20 . .ib2 cj;>g8 2 1 .�c l .ifS
22.'1Mfd2 b4 23.g4 c4 24.gxf5 c3 25 . .ixc3 bxc3
26.�xc3 Wxf5 27.lt:le6 lt:lb7 28.f4 lt:ld7 29.�c7
lt:l f6 30.lt:ld4
1 -0 Goldin - Kosanovic, Belgrade 1 988.
l l . .. fxe6 12.d5
13.cxb5 exd5
The position ansmg is very complex and
unbalanced. Black's queenside is seriously
damaged, but he bases his hopes on the pawn
majority in the centre and the potential activity
of the pieces. However, these considerations
are rather abstract and concrete variations will
determine who will prevail.
8
7
6
5
4
3
2
8
1
b
a
d
c
f
e
g
h
12 ... lb a5!?
Worse is:
1 2 . . . exd5 1 3 .cxd5
1 3 .'1Mfxd5t cj;>h8 1 4.Wxc6? is impossible after 1 4 . . . ib7 the queen is lost.
1 3 . . . lt:la5 1 4.lt:ld4 lt:l e5 1 5 .lt:lce2! ?
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
15.bxa6
a
b
c
d
e
f
g
h
Black has an extra pawn, but what about
the numerous holes in his position? He
could not cope with the problems in the
Instead more promising was:
1 5 .lt:lxd5
Eliminating the important pawn on d5
offers more chances to get an opening edge:
1 5 . . . lt:lxd5
Bad is: 1 5 . . . e5? 1 6.bxa6! Nice tactics.
1 6 . . . c6 What else? (bad is 1 6 . . . exd4 1 7.a7 all Black's army cannot stop the single brave
pawn on a7) 1 7.lt:lxf6t ixf6 1 8.a7 �a8
1 9 .lt:lxc6 And so on . . . 1-0 Pigusov - Smirin,
Las Vegas 1 998.
312
King's Indian Warfare
A radical attempt to destroy Black's centre
and to open a diagonal for the g2-bishop.
Still the price (a piece) is high. White would
be better advised to try to undermine the
opponent's central formation from the
queenside:
1 6.ixd5t mhs
b
a
c
d
e
f
g
h
1 7.b4!
In my opinion this is better than 1 7. bxa6
ixh3 1 8 .ge 1 ( 1 8.a7 gas) 1 8 . . . c6! 1 9 .lDxc6
tDxc6 20.ixc6 'I.Wb6 2 1 .if3 , with an unclear
position.
17 . . . c5 1 8 .bxc6 gxb4 1 9 .ie3 ttJ c4
Or 1 9 . . . ixh3 20 .'\Wd2 ga4 2 l .gfc l with a
kind of domination.
20.gc 1 lD xe3 2 l .fxe3
White seems to be better - the c6-pawn is
very strong.
1 5 ... c6
Now Black can be proud of his pawn centre!
16.a7 �aS
1 7.b4 e5 1 8 .t2J b3 lDxb3 ( 1 8 . . . ttJc4 1 9.lDxd5!)
1 9 .axb3 with a rather unusual double-edged
position. The game may continue 1 9 . . . '\We?
20.b5 d4 2 l .ixc6 gxa7 22.lDd5 '\Wf7 23 .gxa7
'I.Wxa7 24.b6 'I.Wb8 and Black still has a strong
centre, but the pawn on b6 may cause him
a headache - the position is unclear in other
words.
17 ... �xc6 18.�xd5
Of course not 1 8 .ixd5t? lDxd5 1 9.'\WxdSt
mh8 20.'1.Wxc6 ixh3 - the g2-bishop is White's
main hope.
18 ....id7 19 ..ie3
8
7
6
5
4
8
3
7
2
6
1
5
4
a
b
c
d
e
f
g
h
3
19 ... �xa7!
1
Worse would be 19 . . . lDxd5 20.'\WxdSt mhs
2 1 .gac l gxa7 (2 1 ... lDxa7? 22.'1.Wxa8 'I.Wxa8
23.ixa8 gxa8 24.gc7) 22.ixa7 lDxa7 23 .'1.Wb7
and White should be OK - the a-pawn may
become very strong.
The right decision.
2
a
17.�xc6?!
b
c
d
e
f
g
h
Chapter 1 0
2o ..ih6 Y;fbs
Of course not 20 . . . �e8 ? 2 J .Cfj c7+-.
2 1 ..ic7
Black would certainly welcome 2 I .Cfjxe7t
<;t>hs 22.ixa8 �xa8 - White's king is under
strong attack and his pieces are poorly
coordinated.
-
313
Endgames
- the black pieces dominate the board. His
bishops are very strong and the knight is on its
way to d4. The white pawn duo at a2 and b2
is more of a weakness than an asset under the
circumstances.
26 . . . ixb2? would be premature: after 27.�ab l
if6 28 .ie5 White activates his forces and is
out of danger.
2 1 . ..Y;fes 22.gel
White's active pieces create unpleasant
threats, but Black manages to give back the
material and liquidate into a favourable
endgame.
Not 22.Cfj b6? ic6.
22 .. .ltJxd5 23.Y;fxd5t e6!
Black must avoid both 23 . . . �f7? 24.�xe7
and 23 . . . �h8? 24.�xa8 �xa8 25 .ixa8 �xa8
26.�xe7.
24.Y;fxa8
24.�xd6 loses to 24 . . . Cfj b5 .
27J�edl
More resilient would be:
27.ic5
Even though after:
27 . . . e5!?
The other option is 27 ... Cfje5 28.�e3 Cfj c4
(in the event of 28 . . . ic6?! 29 .�d l �xa2
30.f4 Cfj f3t 3 l .�xf3 ! ixf3 32.�d8t <j;>f7
33 .�d7t White forces the draw) 29.�e2
Cfjxb2 30.�b l Cfj a4 3 I .ib4 id4, retaining
winning chances.
28.h4
Or 28 .�g2 Cfjd4 29.ixd4 exd4 30.�ed l
if5 3 l .�d2 �c8 .
28 . . . Cfj d4 29.ixd4 exd4
24...Y;Yxa8 25.�xa8 gxa8 26.hd6
8
7
6
5
4
3
2
a
1
a
b
c
d
e
f
g
h
26 .. .ltJc6!
Nominally White even has some material
advantage, but his position is much worse
b
c
d
e
f
g
h
Black would be much better - the passed
pawn on d4, supported by two mighty bishops,
is very strong.
27 ... �d4
This centralized knight looks in all directions.
314
King's I ndian Warfare
Again the impatient 27 . . ..ixb2? 28.�ab l
.id4 (28 . . .�xa2? 29.�d2) 29.�b7 would j ust
help White to activate his rook and to obtain
counterplay.
28.a4
This is better than 28 . .ic5 lLl e2t 29.�fl .ib5
30.a4 .ic4, and loss of material is unavoidable.
30....ih5 3 1 .�g2
Or 3 1 .�e l lLl d4! .
3 1 . ..bb2 32J�abl .ic6t 33.�fl ltlc3
34J�xb2 lbxdl 35J�b6
8
7
6
8
5
7
4
6
3
5
2
4
3
2
a
1
a
b
c
d
e
f
g
h
28 ....ic6!
Played according to the same motto: "activity
is more important than pawns". That's why I
did not consider the capture on a4.
Indeed, 28 . . ..ixa4? 29.�a2 would be clearly in
White's favour and a draw would be the most
probable outcome.
29.a5?!
Losing an exchange by force.
The most stubborn defence was 29.f4. But
even in this case after the precise 29 . . . .if3!
30.�fl lLl b3 3 1 .�xf3 lLl xa l 32.b3 .id4t
33.�fl �d8 Black should overcome the
technical difficulties in converting his material
advantage.
b
c
d
e
f
g
h
35 ....if3?
An inaccurate move in slight time pressure.
35 . . . .id5 36.a6 lLl c3 would finish the game at
once.
36.a6 ltlc3 37.a7!
I overlooked this. White keeps the a-pawn
and even advances it to a7.
37 ....id5
Earlier I did not notice that 37 . . . �xa7?
3 8 .�b8t �f7 (or 38 . . . �g7 39 . .ie5t) 39.�f8t
would cost Black a bishop.
38J�b8t �f7 39 ..ic5
It may seem questionable whether Black can
win - the protected pawn on a7 and bishops of
opposite colour cast such doubts. However, the
aB-square is under firm control and the extra
knight weighs heavily. Black should gradually
win after consolidating his forces.
29 ... ltle2t 30.�fl
30.�h2 is not a remedy: 30 . . ..ixb2 3 l .�ab l
lLl c3 32.�d2 lLl xb l 33.�xb2 �d8! and so on.
39 ... lbe4 40 ..ie3 lbf6 41 .£4 ltld7 42J�b2
�e7 43. �e2 gcs 44.g4 �d6 45.g5 e5!
315
Chapter 1 0 - Endgames
8
53.E:e8t �f5 54JU8t �f7
7
It would be a pity to get checkmated after
the "logical" 54 . . . @e4?? 5 5 .!!f4#.
6
5
55.�e2 �e4 56.E:e8t �e5
3
57.i.c5 E:a2t 58.�el
4
Now everything is OK.
2
Or 5 8 .@d l @d3 and now it's the white king
in a mating net.
1
a
b
c
d
e
f
g
h
The beginning of the end.
58 ... � 6
8
7
6
46.fxe5t �xe5
Black's king gains his freedom.
5
47J�b4 E:c2t 48.�d3 E:a2
The rook has managed to get behind the
a-pawn - it guarantees the win.
49.E:b8 �d6 50.E:d8 E:a3t 5 1 .�d2 �e6
52.E:h8 �e5
8
4
3
2
1
a
7
b
c
d
e
f
g
h
White resigned due to: 59 .ie3
60.@d l lDxg5 6 I .ixg5 (6 I .!!f8t
6 I . . .if3t 62.@c 1 @xg5
6
5
0-1
4
3
2
1
a
b
c
d
e
f
g
h
lt:J f3t
lt:J f7)
316
King's Indian Warfare
Christian Gabriel - Ilya Smirin
Pula 2000
This game was played in a Zonal tournament
in Pula, Croatia. Both of us badly needed to
win to keep our chances alive of qualifying for
the Knockout World Championship.
l .d4 �f6 2.c4 g6 3.�c3 .ig7 4.e4 d6 5.�f3
0-0 6..ie2 e5 7.d5 a5 8.h3 � a6 9 ..ig5 �e8
9 . . . h6 1 0.i.e3 was played in my game with
Kramnik in 1 99 5 , which you will find on page
52.
8
7
6
5
b
a
c
d
e
f
g
h
An important move - the KID bishop
becomes active. 1 4.E:b 1 lD c5 1 5 .b4 axb4
1 6.axb4 lDca4 1 7.lDb5 �d7 1 8 .�b3 @h8
1 9 . lD f3 f5 20.0-0 fxe4 2 1 .lLJh2 i.f4! 22.i.g4
�g7 23 .i.xc8 E:fxc8 24.E:fe 1 g5 25 .i.g3 �g6
With an extra pawn and an excellent position;
Ostenstad - Nakamura, Rhodes 20 1 3 .
1 o... ®hs
4
3
2
1
a
b
c
d
e
f
g
h
10.g4
Not so much an attacking move, but rather
a try to squeeze Black on the kingside. Also,
White is ready now to meet a future . . . f7 -f5
with the exchange gxf5 , opening the g-file and
trying to scare the black king.
Often White prefers l O.lLJd2 here. The
following game by Nakamura may serve as
an example of Black's actions in that case:
l O . . . lDd7 l l .a3 f6 1 2.i.h4 lLJ b6 1 3 .b3 i.h6!
With the idea that I managed to realize in
the game - Black is going to play . . . lD g8 and
carry out . . . f7-f5 and/or exchange the dark­
squared bishops via h6.
Nowadays I would probably prefer a slightly
different plan, starting with 10 . . . lLJd7. For
instance: 1 1 .El:gl lDdc5 1 2.h4 c6 1 3 .h5 cxd5
1 4.h6 i.h8 1 5 .cxd5 i.d7 and Black does not have
much to complain about - the bishop on h8 is
not shut in forever (hopefully) , and meanwh ile
White has problems with the placement of the
king; Piket - J. Polgar, Aruba 1 995.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
317
Chapter 1 0 - Endgames
l l .tlJdl
Twelve years after this game, Anton Korobov
demonstrated a deep and strong strategic idea
in his game with me. This game deserves to be
shown till the end:
1 1 .ie3! ttJg8 1 2 .g5 f5 1 3 . h4!
This is the idea - now Black's kingside is
severely cramped.
1 3 . . . ttJe7 1 4.h5 �g8 1 5 .h6 ih8
This bishop is in complete misery.
1 6.ttJh4 b6 1 7.f3 ttJc5 1 8 .�d2 id7 1 9 .0-0-0
a4
8
33.�xf4! exf4 34.ig7!
A beautiful combination - Black is helpless.
34 . . . ttJ a7 3 5 .�c3 �f8 36.ixf8t �xf8 37.�a5
ixb2t 38.�c2 �a8 39 .�xa4 ie5 40.�b 1
1-0 Korobov - Smirin, Eilat 20 1 2.
1 1 . .. tlJ g8 1 2 ..ie3 f5
8
7
6
5
4
3
7
2
6
5
4
3
a
2
1 3.gxf5
a
b
c
d
e
f
g
h
20.ttJg2 �f7 2 l .ttJ e 1 �a5 22.ttJd3 �fa8
23.�c2 �e8 24.ttJxc5 bxc5 25.ttJb5 �d8
26.id2 �5a6 27.ic3 f4
Trying to close the position as much as
possible.
28 .ifl ttJ c8 29 .ih3 ixb5 30.cxb5 �b6
3 1 .ie6t �f8 32.�h4 �e7
8
7
6
5
4
b
c
d
e
f
g
h
In the event of 1 3 . f3, trying to keep White's
central pawn chain intact, I would proceed
with 1 3 . . . ih6 1 4.ixh6 ( 1 4.g5 ? f4) 1 4 . . . ttJxh6
with good and easy play - some dark squares
in White's camp may easily become weak.
13 ... gxf5 14.exf5 .ih6!?
An interesting alternative would be 14 . . . ttJ e7,
intending to recapture on f5 with the knight.
After the possible continuation 1 5 .ih5 �d8
1 6.ttJde4 (or 1 6.f6 ixf6 1 7.ttJde4 ttJ f5 1 8 .ig4
ih4) 1 6 . . . ttJxf5 1 7 .ig5 �d7 1 8.�d2 b6!?
1 9 .0-0-0 ttJc5 20.ttJxc5 bxc5 a double-edged
situation would arise. Indeed, both kings may
easily find themselves under attack.
1 5 ..ig4 Yffe7 16.Vffe2
3
2
a
b
c
d
e
f
g
h
IfWhite had tried to keep the extra pawn on
f5 by means of 1 6.�f3 , I was going to proceed
with: 1 6 . . . �h4!? (not 1 6 . . . tt:J b4?! 1 7.0-0-0
ttJ d3t 1 8.�b l ) 1 7.0-0-0 ttJ e7 - Black will
take on f5 and remain OK.
318
King's I n dian Warfare
16 ...Lfs
with all his minor pieces. The bishop on g6 is
really strong - it controls the very important
b 1 -h7 diagonal and, j ust in case, covers the
black king.
8
7
6
20 ..if3 ltJd7!? 21 .h4 ltJac5 22.h5 .ifS
23J3dgi ggs
5
4
These manoeuvres are easy to understand.
The next king move looks slightly unnatural,
but it just shows that White has problems
finding a decent positional plan. Indeed,
Black's moves look more natural.
3
2
a
b
c
d
e
f
g
h
17.0-0-0
After 1 7 .ixf5 ixe3 (in the event of 1 7 . . J�xf5
1 8 .0-0-0 ixe3 1 9.fxe3 lt:Jc5 20 . lt:J de4 White
would have some advantage due to the strong
knight on e4 and the slightly vulnerable
position of Black's king) 1 8 .fxe3 '\Wh4t 1 9.'\Wf2
l.Wxf2t 20.<j{xf2 �xf5 t 2 1 .<j;>e2 �af8 22.�afl
�xfl 23.�xfl �xfl 24.<j{xf1 lt:J f6 the numerous
exchanges would lead to a dull draw. But as I
explained earlier, this outcome would hardly
suit either player.
17 ....ig6 1 8.ltJde4 .ixe3t 19.fxe3 ltJf6
24.@d2 gaf8 25.gg2 gxg2 26.�xg2 ggs
27.�fl �fB!?
8
7
6
5
4
3
2
1
a
8
b
c
d
e
f
g
h
Indirectly applying pressure to the knight on
e4.
7
6
28.ltJxc5 ltJxc5 29 ..ie2
5
After relinquishing control over e4, Gabriel
hopes to defend himself in the forthcoming
inferior endgame.
4
3
2
29 ....id7 3o.�xm gxrs 3 1 .gh2 @g?
1
a
b
c
d
e
f
g
h
Black has a harmonious position - he may
effectively challenge White's control over e4
Black's advantage is probably not too big,
but it is obvious. All his pieces are slightly
more active than their counterparts, and the
pawn on h5 has become weak in the ending the black king may easily threaten it.
Chapter 1 0
-
319
Endgames
37 ... tlJc5
32.b3
Sacrificing a pawn for domination.
Or 32.h6t mg6 does not help.
38.tlJxc7 tlJe4t 39.<i>e2 tLlg3t 40.<i>d2 �gl !
32 ... <i>h6 33 ..idl
White prepares to drive away the knight
from c5 by means of a3 and b4, but in doing
so he allows the black rook to penetrate.
33 .. J�fl 34.�c2
34.ie2 :ggl 3 5 .:gf2 mg7 36.:gh2, trying to
preserve status quo, was another way to defend.
34 ...�g4 35.tLl b5
Probably the immediate 35 .a3 was more
accurate.
35 ... tlJa6 36.a3 .ifS!
8
7
6
5
4
3
2
1
a
8
b
c
d
e
f
g
h
The strongest continuation.
7
6
41 .�c2!
5
From a practical point of view this is an
absolutely correct exchange sacrifice.
3
Houdini recommends 4 1 .ie2 :gal 42.ctJb5
lt:Je4t 43 .mc2 (43 .md3? :ga2 with checkmate)
43 . . . :ga2t 44. mc 1 , but I do not believe such a
passive defence could save the game.
4
2
1
a
b
c
d
e
f
g
h
The key move. White faces a tough choice:
to swap the bishops and lose the h-pawn or to
allow considerable activity for the opponent's
forces.
37.�dl
Gabriel picks the second option.
Mter the sample variation 37.:gg2 ixc2
38.mxc2 :gf7 39.ctJc3 mxh5 40.lt:Je4 b6!
4 1 .md3 ctJc5t 42.ctJxc5 bxc5 he would hardly
have real chances of survival - the extra pawn
on h7 is very strong.
4I. .. tiJfl t 42.<i>c3 tlJxh2 43.hf5
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
320
King's I ndian Warfare
The position is sharp - White has strong
counterplay connected with c4-c5 to create
a passed d-pawn. Concrete and precise play
is required from both sides, especially from
Black.
43 .. .ttJg4?
A big mistake - I picked the wrong square
for the knight.
Correct was:
43 . . . ttJ fl !
Attacking the e-pawn from a safe square.
44.c5
If 44.ttJ b 5 ttJxe3 45.ttJxd6 (45 .id3 does
not save White after 45 .. J�g2 and now
46.ie4 �g3 or 46.ttJxd6? ttJ d l #) 45 . . . ttJxf5
46.ttJxf5 t �xh5 47.ttJd6 Black proves to be
j ust faster: 47 . . . �g4 48.ttJxb7 e4! 49.ttJc5 e3
50.ttJd3 �d l etc.
44.�d3? �g3 would lose at once.
44 . . . dxc5 45 .d6 ttJxe3 46.d7
Perhaps 46.ie4 �g7 47.�d3 ttJg4
48.�c4 ttJ f6 49.ixb7 is the most resilient
continuation, but I suppose that after
49 . . . e4 Black is still close to a win.
46 ... �d l 47.id3 e4
Or 53 .ttJd5 �g5 .
53 . . . �g5 54.ttJg7 ttJ e3 5 5 .�b6 ttJ f5 56.ttJe6t
�g4
And so on.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
44 .<ibd3?
White misses the chance to turn the tables.
After the simple 44.ixg4 �xg4 4 5 .ttJe8 �g8
(the best try) 46.ttJxd6 �xh5 47.ttJxb7 �g6
48.c5 h5 49 .c6 he would be on top in the
pawn race: 49 . . . h4 50.c7 h3 5 1 .d6 h2 52.d7
h l =W 53.d8=W Wc6t 54.�d2 '!Mg2t 5 5 .�d3
with serious winning chances. But luckily the
German grandmaster did not spot this.
44 ... <it>xhs 45.hh7?
The second error in a row; the h-pawn was
not too dangerous yet.
Instead of being greedy, White should have
proceeded 45.ttJb5 �g5 46.ttJxd6, attacking
the key pawn on d6. In that case all three
results would still be possible.
a
b
c
d
e
f
g
h
48 .d8=W �xd3t 49 .'!Mxd3 exd3 50.�xd3 ttJg4
5 1 .�c4 �xh5 52.�xc5 �h4
This knight ending should be winning - the
h-pawn will decide the matter. For instance:
53.ltJe8
45 .. J�g3
Now the e3-pawn is doomed.
Chapter 1 0
8
-
32 1
Endgames
lD c6 and the pawn on d7 is fi r mly stopped)
53 . . . �b3 54.if5 lD b8 5 5 .d7 lD c6t 56.c;t>c4
�a3! 57.lDe4 ltJdB 58 .lD c5 �e3 5 9.c;t>ds c;t>gs ,
and again the d-pawn cannot save White.
7
6
5
4
52.llJb6 llJ f7
After 52 . . . a4 53.d7 lDxd7t 54.ltJxd7 �b3
5 5 .id5 �b2 56.c;t>c4 White would be just
in time: 56 . . . b5t 57.c;t>c3 a3 5 8 .ib3 b4t
59.c;t>c4, with a draw.
3
2
a
b
c
d
e
f
g
h
46.c5!
The only way to complicate Black's task.
46 .. J:�xe3t 47.®c4 dxc5 48.d6 e4
This move was begging to be played. At the
same time this is the only way to keep the
advantage.
49.llJds llJest so.<i>xc5 gxb3 s t .Le4
8
8
7
6
5
4
3
2
1
a
7
b
c
d
e
f
g
h
53.Lb7?
The losing mistake. Now White's forces lose
coordination for a moment and this is enough
for Black to prevail. It's curious that three times
in this ending, starting from the 4 5 th move,
the capture of a pawn proved to be wrong!
6
5
4
3
2
a
b
c
d
e
f
g
h
5 1 ... gxa3?
This hasty capture could have thrown away
the victory.
Correct was to consolidate the position first
by inserting a check: 5 1 . . . lD d7t 52.c;t>d4 �xa3
53.lDc3 (53.if5 �a4t 54.c;t>c3 lDe5 5 5 .d7
White should have used his main trump
instead: 53 .d7! �c3t There is nothing better.
54.c;t>d4 �g3 5 5 .ixb7 lD d8 56.id5 And Black
cannot win - his king is too far away.
53 ... gc3t!
The key check.
54.<i>d4
54.c;t>ds �b3 would not change anything.
54... gb3 55.�d5
Or S S .c;t>cs lDxd6 56.id5 �b4.
322
King's I ndian Warfare
8
55 .. J�b4t 56.®c5 �xd6
7
8
6
7
5
6
4
5
3
4
2
3
1
2
1
a
b
c
d
e
f
g
h
In this study-like posmon Black wins,
despite having only one pawn left and his king
temporarily not involved in the battle.
57.�d7 �f5 58.�e5
Black needs:
b
a
d
c
e
f
g
h
69 ... ge5
Here Black had a prettier win: 69 . . . �h5!?
70.ic2 �h6t 7 l .�a7 �h4! 72.�a6 (72.ib3
ltJ d2! 73 .lDxa5 lDxb3 74.ltJxb3 �a4t 75.�b6
�b4t 76.�c5 �xb3 - a real study-like motif)
72 . . . lD e3 73 .ib l a4 and so on.
70 ..id3?!
Gabriel too had little time left.
1 . To preserve his a-pawn.
2. To consolidate his forces.
3. To move his only pawn forward.
He is able to do all this, but precise play is still
required.
58 ... � e3! 59 ..if7t ®h6
59 . . . �g5 was stronger, but I instinctively
wanted to hide from possible knight checks.
Otherwise he would probably have played the
more resilient:
70.ic2! �e6t 7 l .�a7
8
7
6
5
4
60 ..ie8 ge4 61 .�f7t ®g7 62.�d6 ge5t
63.®b6 ge6 64.®c5 ge5t
A repetition of moves in time trouble to gain
additional time on the clock.
3
2
a
65.®b6 gd5! 66.� b7 � c4t 67.®a6 g5
68 ..ic6 ®f8 69 ..ie4
b
c
d
e
f
g
h
7 l . . . �e3!
Apparently this is the only way to victory.
After 7 l . . . �h6? 72.ib3 ltJd2 73.lDxa5
the game would be drawn. Interestingly, if
White had a knight instead of a bishop (i.e.
two knights versus rook and knight) in the
323
Chapter 1 0 - Endgames
final position, then according to "ancient"
computer programs he would lose in a few
hundred moves! I remember seeing the
computer's analysis of such an ending back
in 1 99 5 . This conclusion is hardly relevant
to a practical game though.
72.<±>a6
Or 72.lLJc5 <±>e7.
72 ... <±>e7 73 . lLJ c5 <±>d6
White is left with no hope.
8
7
6
5
Sabino Brunello - Ilya Smirin
Helsingor 20 1 5
l .d4 tiJ f6 2.c4 g6 3.tLlc3 .i.g7 4.e4 d6 5.8
0-0 6..i.e3 c6 7. ttJ ge2
7.id3 a6 8.tt:Jge2 b5 was played in a well­
known game between the K's in their last match
for the crown: 9.0-0 tt:Jbd7 1 0 .Ek1 e5 1 1 .a3
exd4 1 2.tt:Jxd4 ib7 1 3 .cxb5 cxb5 1 4 .�e1 tt:Je5
1 5 .ifl �e8 1 6.if2 d5 1 7.exd5 tt:Jxd5 1 8 .tt:Jxd5
'.Wxd5 1 9.a4 ih6 20.�a 1 tt:Jc4 2 1 .axb5 axb5
22.�xa8 �xa8 23 .'.Wb3 ic6 24.id3 tt:Jd6
25 .'.Wxd5 ixd5 26.tt:Jxb5 tt:Jxb5 27.ixb5 ig7
28 .b4 ic3 29.�d 1 ib3 30.�b 1 ia2 Yz-Yz
Karpov - Kasparov, Lyon/New York 1 990.
4
7 ... a6
2
7
3
8
1
a
b
c
d
e
f
g
h
70 ... tLlb2
This decides the game at once - the a-pawn
will cost White at least a piece, so he resigned.
An interesting endgame!
0-1
6
5
4
3
2
1
a
b
c
d
e
f
g
h
I decided to play this line to surprise my
opponent - I had never used it before. Also
I wanted to test some advice, given to me by
Vlad Tkachiev some 1 Yz years before the game,
which was connected with the encounter
Dreev - Wang Hao (you will find this game on
the next page) .
8.c5
This continuation is by far the most popular
nowadays. The idea is obvious - to use the
weakness of the b6-square and to underline
the main drawback of my 7th move.
324
King's Indian Warfare
8 ... b5 9.cxd6
In my opinion 9.cxb6 ttJ bd7 deserves
attention. And now, let's say 1 0.�d2 lDxb6
1 1 .lDg3 a5 1 2 .Ek 1 ia6 1 3 .ixa6 El:xa6 1 4. ltJ d 1
�d7 1 5 .0-0 and White has no weaknesses and
perhaps may claim some advantage.
9 ... exd6
8
1 3 .ic l ttJb6 1 4.lDxb6 �xb6 1 5 .ie3 d5
1 6.�d3 dxe4 1 7.fxe4 a5 1 8 .El:d 1 ia6 1 9.�c2
ixe2 20.lDxe2 lDd7 2 1 .0-0 lD f6 22.lDg3 lDg4
23.if2 El:ad8 24.e5 c5 25.dxc5 �c7 and Black
was fine despite losing in the end, Carlsen Grischuk, Dubai (rapid) 20 1 4. The move I
made in the game is associated with the same
idea.
l l .�a4
7
6
5
4
3
2
1
b
a
d
c
e
f
g
h
A rather optimistic choice.
1 1 .lDce2
This was probably safer, but even in this case
Wang Hao gained better prospects against
Dreev:
1 1 . . . El:e8 1 2.lDg3?!
It would be fascinating to test the creative
continuation 1 2.�d2 �e7 1 3.�xb4 lDxe4!
1 4.fxe4 �xe4 1 5.id2 ixd4, with an
interesting mess.
10.�£4
The fashionable move. However, I would
probably prefer the less creative "dull"
continuation 1 O.�d2 El:e8 1 1 .lDg3, holding
the pieces together for a while.
10 ... b4
The immediate 1 0 . . . El:e8 was played by
Alexander Grischuk against the World
Champion in the World Rapid Chess
Championship - 1 1 .ie2 b4 1 2. lD a4 lD d5!
a
b
c
d
e
f
g
h
1 2 . . . ltJd5!
This is the tactical j ustification of Black's
opening concept.
1 3 .ic l
This is Dreev - Wang Hao, Troms0 20 1 3
( 1 -0, 53) and now after the simple:
13 . . . lDxf4 1 4.ixf4 c5
White would fall behind in development
and face serious problems.
1 1 ..J�e8 12 ..id3?!
Sabino does not sense the danger.
a
b
c
d
e
f
g
h
325
Chapter 1 0 - Endgames
1 2 .ie2, transposing to the game Carlsen Grischuk, should have been preferred.
8
7
ttJxd 1 25 .ixf7t �xf7 26.�xd 1 a5 27.f4
ih8! 28 .�xd6 a4 29.e5 b3 30.e6t �g8 3 1 .e7
b2 32.�d8t �g7 33.�dl (or 33.h4 b 1 =�t
34.�h2 �e4!) 33 ... �f7 he would be helpless.
16 .. J��b8
6
5
The simplest. Now the pawn on d4 is
impossible to defend, and after that the
difference in piece activity will be j ust too big.
3
17J�adl tlJxd4 18 ..ixd4 .ixd4 19.exd5
4
2
8
1
b
a
d
c
f
e
g
h
12 ... tlJd5!
7
6
5
4
Following the instant classics.
1 3.tiJxd5 cxd5 14.0-0 .ib7
Black's position is a pleasure to see - two
bishops fire at White's centre, and the knight
will j oin them in no time. Meanwhile the
knight on a4 is a silent witness to his army's
opening fiasco.
3
2
1
a
b
c
d
e
f
g
h
19 ...fih4
1 5.fib3 tlJ c6 16. ci> hl
Understandably Brunello did n ot want to
enter the forced path by means of: 1 6.�xd5
ttJ a5 1 7.�g5 �xg5 1 8 .ixg5 ixd4t 1 9.�h 1
Eyeing the white king. The threat of . . . ie5 is
very unpleasant now.
20.g3
Of course, White's position is no cause for
optimism, but it was better to do without this
move, which weakens the pawn shelter near
the king.
a
b
c
d
e
f
g
h
Indeed, after 1 9 . . . ic6 20 .ic2 ttJ c4 2 1 .�fd 1
ie5 22.�ac l ixa4 23 .ixa4 ttJxb2 24.ixe8
More chances for resistance would be offered
by the brave:
20.�c4 �ec8 2 l .�xb4
After that the best continuation for me
seems to be:
2 l . . .ic6
2 1 . .. ixd5 is met by 22.�e l .
22 .�a3 ixd5 23 .�xd6 �d8 24.�c7 �bc8
25 .�g3 �xg3 26.hxg3 ixa2 27.�a 1 ib3
326
King's I ndian Warfare
28 ..ie2
28 .ixa6? �c6! 29.lt:Jc3 g5 would suddenly
lead to checkmate.
28 . . . a5
22JUel
22.ib5 was slightly better, but it would not
change much.
22 .. J�xel 23J�xel Y;Ygs 24.Y;Yc4 .ia7!?
24 . . . V9xd5 25 .V9xd5 ixd5 would lead to
a technically won endgame, but objectively
I made a stronger move - the white king is
under a strong attack and the knight on a4
cannot help.
25 ..ie4
b
a
c
d
e
f
g
Starting from this moment, I missed a few
possibilities, each of which would have led to
a quick win.
h
The resulting endgame is certainly bad for
White, but he still would retain chances to
hold. With queens on the board, his situation
is even worse - various assaults against the
white king are hanging in the air.
25 .. J�e8
25 . . . �c8 26.V9b5 V9d2t 27.V9e2 (27.�e2
V9d l ) 27 . . . V9xe2t 28.�xe2 �c l 29.�c2 �g l t
30.�h3 ia6!-+ and the white king is in a
mating net.
20 ...Y;Yf6 2 1 . �g2 aS
8
8
7
7
5
5
6
6
4
4
2
2
3
3
a
b
c
d
e
f
g
h
Mter I made this move, Sabino became
visibly sad - he has not even a shadow of
counterplay and must await his fate. I was sure
that the game would not last long and partially
lost my concentration. As is often the case, this
could have cost me dearly.
a
b
c
d
e
f
g
h
26.. J�b8?
26 . . . .ic6! 27.V9xc6 V9d2t 28.�h3 �e5, with
an irresistible mating attack, would have won
nicely.
Not so fast, but also a pretty win, would result
from:
Chapter 1 0
26 . . . �e7 27. f4
27.�e2 Wl'c l 28.tDb6 Wl'd 1 29.id3 �c7 and
White is j ust helpless.
27 . . . WI'h5 28 .if3
28.f5 �xe4 29Jhe4 Wl'xf5 30.�e8t <;!;>g?
3 1 .WI'xb7 Wl'f2t 32.<;!;>h3 Wl'fl t 33 .<;!;>h4 Wl'f5!
and the extra rook cannot save the white king.
-
327
Endgames
3 1 .£4 YMe2 32.Wxe2
The position remains lost for White, but at
least he has managed to exchange queens.
32 ... gxe2 33 ..id3 gxal 34..ic4 i.c8?
34 . . . h6 3 5 .�e l g5 36.fxg5 hxg5 37.�e7 ig l
would still have finished the game quickly.
35.gel <i>f7
3 5 ... id7, with the idea 36.�e7?! igl , would
have been much stronger. Before the time
control at move 40, I started to make mistake
after mistake.
36.ge2
8
b
a
c
d
e
f
g
7
h
6
28 . . . ixd5!!
Black wins.
5
27.b3 YMe5 28J�dl f5 29 ..ibl ge8 30.<i>h3
The last line of defence.
4
3
2
8
1
7
6
a
5
b
c
d
e
f
g
h
36 ... gal ?
4
36 . . . �xe2 37.ixe2 id? should still be
enough for a win, but as I said, I was completely
disoriented.
3
2
1
37 ..ib5 .ib7?
a
b
c
d
e
f
g
h
30 .. J�e7
30 .. .f4! 3 1 .g4 ic8 32. <;!;>g2 ixg4 33.fxg4
f3t 34.<;!;>h 1 Wl'e2 3 5 .WI'xe2 fxe2 36.�e1
if2 would again have finished the game
quickly.
37 . . . �cl , preventing 3 8.�c2, was the only
correct move.
38J�c2
Now suddenly White has good counterplay
for the pawn.
38 ... gel 39.�c7t ge7
328
King's I ndian Warfare
8
If 43 . . . <i>f6 44.h4 h6 45.<i>f3 g5 46.hxg5 t
(46.h5!?) 46 . . . hxg5 47.<i>g2 ib7 48.ic4 <±>g6
49.<i>h2 <i>h5 50.ie2t Black would be unable
to storm the fortress. That's why I believed my
43th move was the best practical chance.
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
The critical line, however, comes after an
exchange of bishops:
43 . . . id7 44.ixd7 <i>xd7 45.<i>f3 <±>e7!
And not: 45 . . . h6? 46.h4 <±>e7 47.\t>g2 <i> f6
48. <i>f3 g5
40J�xe7t?
The 40th move . . . Meanwhile 40.id7! ic8
4 1 .:8:xa7 ixd7 42.:8:xa5 ixa4 43.:8:xa4 :8:e4
would have led to a draw.
40 ... <i>xe7 4I ..i.c4
But even now it's very questionable whether
Black can win, even with an extra pawn, the
bishop pair and the knight being cut off:
White has a kind of fortress.
41 ...i.d4 42. <i>g2 .icS 43 ..ih5!
8
7
6
5
a
b
c
d
e
f
g
h
49.h5 The draw is evident; the strangled
knight guards the c5-square - the only place
for the black king to break through. For this
reason we need to keep the h6-square clear
for the king.
46.h4 <i> f6 47. <±>g2 <±>g7 48. <i>f3
48.<i>h2 <i>h6 49.<i>g2 <i>h5 50.<i>h3 h6-+
48 . . . \t>h6 49.g4
4
3
2
1
a
b
c
d
e
f
g
h
43 ... g5!?
During the game, and also for a while
afterwards, I believed that White would be able
to set up a fortress if I did not do something
drastic.
3
2
a
4 9 . . .'�:Yg7!
�
b
c
d
e
f
g
h
Chapter 1 0
At first I analysed a more natural, but weaker,
line: 49 . . . fxg4t 50.<it>xg4 <it>g7 5 I .<it>f3
-
50.g5 h5!
a
a
b
c
d
e
f
g
h
5 1 . . . <it>f6? (Even here Black can win with
5 1 . . . h5! 52.<it>e4 if2 53.l2Jb2 ixh4 54.l2Jc4
ie l 5 5 .l2Jxa5 h4 56.l2Jc4 h3 57.<it>f3 h2
5 8.<it>g2 ig3 59.l2Jxd6 ixf4 60.l2Je4 <it>f7,
as shown by Jacob Aagaard.) 5 2.<it>e4 if2
53.l2Jb2 ixh4 54.l2Jc4 a4 5 5 .bxa4 h 5
(5 5 . . . <it>e7 56.l2Ja5 <it>d7 57.f5 gxf5t 58.<it>xf5
id8 59.l2Jb3 h5 60.l2Jd2 <it>c7 6 I .l2Jb3 <it>b6
62.<it>f4
b
c
d
e
f
g
h
I see no way for Black to win the game. For
example: 62 . . . <it>c7 63 .<it>f5 <it>d7 64.lt:Jd2
h4 6 5 .<it>g4 ie7 66.lt:J b3 <it>e8 67.<it>h3 <it>f7
68.a5 id8 69.a6 ib6 70.<it>xh4 <it>f6 7 I .lt:Ja5
<it>f5 72.<it>g3 <it>e4 73 .lt:Jc6 b3 74.a7 ixa7
75.lt:Jxa7 b2 76.lt:Jb5 <it>xd5 77.<it>f3 with an
obvious draw.) 56.a5 <it>e7 57.a6 if2 5 8 . lt:J e3
b3 59.a7 b2 60.a8='W b 1 ='Wt 6 I .<it>f3 ixe3
62. <it>xe3 Black has winning chances here,
but nothing is guaranteed.
b
c
d
e
f
g
h
5 1 . <it>g2 <it>f7 52. <it>f3 <it>e7 53. <it>g3 <it>d7 54. <it>f3
<it>c7 5 5 .<it>g3 <it>b7 56.<it>f3 <it>a6 57.<it>g3 <it>b5
White is quickly running into zugzwang.
58.<it>f3 ig 1 59.<it>g2 ie3 60.<it>f3 id2 6 I .<it>g3
ie 1 t 62.<it>h3 if2
a
a
329
Endgames
b
c
d
e
f
g
h
Black wins, as he is not afraid of losing the
bishop.
63.l2J b2 <it>c5! 64.lt:Jd3t <it>d4 65.lt:Jxf2 <it>c3
And so on.
44 . fxg5
I gave back my extra pawn, but the fortress
is no longer on the cards without the f4-pawn.
Still, all this is hardly sufficient to win the game.
If 44.h4 gxf4 45 .gxf4 <it> f6 the h-pawns are on
the board, which is in Black's favour: 46. <it>g3
<it>g6 47.<it>f3 ib7 48 .ic4 ic3 49 .l2Jb6 ie 1 ,
with good winning chances.
King's Indian Warfare
330
8
44... ci> f7 45. ci> £3
7
8
6
7
5
6
4
5
3
4
2
3
1
2
1
a
b
c
d
e
f
g
h
45 ... h6
a
b
c
d
e
f
g
h
Black may again seriously count on
converting his extra pawn. Nevertheless, the
reduced number of pawns gives White hope.
An impatient move.
5 1 ...-igl !
More chances would be promised by the
simple 45 . . . mg6 46.h4 h6 47.gxh6 (bad is
47.ie8t mg7 48.g6 ia6) 47 . . . mxh6 48.mf4
mg6 49.ie8t mf6 50.ib5, and Black may try
to win this, even though a draw is the most
probable outcome.
46.gxh6 ci> g6 47.h3
After the natural 47.mf4 mxh6 48 .id3 mg6
49.g4 ie5t so.mf3 the position is j ust equal.
47... ci>xh6 48.g4?
In my opinion this mistake was caused by
nerves.
4s.mf4 mg6 49.iest m f6 so .ibs ib7
5 I .ic4 ia8 52.g4 would still lead to a fairly
easy draw.
4s ... ci> g5 49 ..id3 fxg4t 50.hxg4 Lg4t
5 1 . ci> e4
An exact move.
After 5 I . . . ia7 52.lDb2 if5 t 53 .mf3 ixd3
54.lDxd3 mfs 5 5 .lD b2 mes 56.lDc4t mxd5
57.lDxa5 the outcome probably would be a
draw - after 57 . . . md4 58.lDc6t the bishop on
a7 will be taken by the white knight.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
White is in zugzwang, and my king will
finally penetrate.
33 1
Chapter 1 0 - Endgames
55.t!Jb2 ®e5 56.®e2
56.lt:Jc4t �xd5 57.lt:Jxa5 �d4 58.�g2 ie3
59.�f3 ih6 60.lt:Jc4 if8 6 1 .�e2 �c3 62.lDa5
d5 63.�d l ih6 would not save White either.
65.®e2 ®e4 66.®d2 J.e3t 67.®c2 J.f4
68.t!J c5t ®e3 69.t!Jb7 J.c7 70.t!Jc5 d4
7I.t!Jd3
8
7
56 ... ®xd5 57.®d3
Or 57.lt:Jc4 a4.
6
5
57 ... ®c6 58.t!Jc4 .ih6 59.t!Jb2 d5
8
4
6
2
3
7
1
5
4
a
3
b
c
d
e
f
g
h
71. .. a4! 72.t!Jxb4 a3 73.®dl d3 74.t!Ja2
.ia5 75.b4 .ids 76.b5 ®d4 77.®d2 .ia5t
78.®dl ®c4
2
a
b
c
d
e
f
g
h
White loses because his knight will soon run
out of good squares.
60.t!Ja4 .ia7 6I.t!Jb2 ®d6 62.t!Ja4 ®e6!
63.t!Jb2 ®e5 64.t!J a4 .id4
This knight is strangled again - the bishop
dominates him. The rest is easy.
Not a great game, of course, but the opening
and endgame are quite interesting.
0-1
Chapter 1 1
My Two Favourite Losses
Mikhail Tal playing in Berlin 1986
Test yourself against the book
In this section you get a chance
to train your King's Indian
muscles and measure yourself
against the variations in the
book. Take as long as you like
answering these questions.
Some would want to make
intuitive decisions, others to
practise calculation. Both have
their merits.
8
8
7
7
6
6
5
5
4
4
3
3
2
6.
b
a
c
d
e
f
g
2
6.
h
What is the most accurate
technical move? (see page 337)
8
a
b
c
d
e
f
g
h
How should White organize the
queenside? (see page 34 1 )
8
7
7
4
4
6
6
5
5
3
3
2
2
a
b
c
d
e
f
g
h
What simple move had I
overlooked, giving Tal an
advantage? (see page 336)
b
a
c
d
e
f
g
h
a
b
c
d
e
f
g
h
White wins the exchange with
20 years passed before the
ie6 but how can he improve
the position first? (see page 337)
you find it? (see page 342/43)
defence was discovered. Can
,
8
8
7
7
6
6
5
5
4
4
3
3
2
2
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
How can White take the
Which great positional move
I did not find the defence, giving
initiative? (see page 336)
did Kasparov make?
Kasparov an option to do what?
(see page 339/40)
(see page 343)
Chapter 1 1 - My Two Favourite Losses
335
In this book I have given a collection of my
best games in the King's Indian. This can give
a somewhat lopsided impression, which is by
no means my intention. When you live by the
sword, you will at times die by the sword. I
have played the King's Indian for decades and
have of course suffered my share of undeserved
and unlucky defeats. Of these, two are more
memorable than any others.
Mikhail Tal - Ilya Smirin
Podolsk 1990
This was my only game with my chess hero
Mikhail Tal. One of the first chess books I read
was " Vogon Ataki" by Tal and Damsky. I read
this book tens of times and I think it influenced
my chess style. By 1 990 Tal was already very
ill: he looked much older than his age. Still he
remained Tal - a true genius. In the USSR Team
Championship in which we played this game
he performed very well, and won three games
with no losses, on Board 1 . Alas, it was one
of his last triumphs, and about two years later
Mikhail passed away. The name of the Magician
from Riga will remain in chess forever. As for
me, this game occupies a very special place in
my memory.
Most of the main games in this book are games
I have won, but there are a few exceptions. The
two in this chapter are the kindof games you
want to lose; if you absolutely have to!
b
a
c
d
e
f
g
h
One example is the following game, which
I played in a match against four computer
programs (which I won 5-3). Perhaps this was
one of the last occasions on which a human
prevailed over the silicon monster in chess.
14.a4 id7 1 5 .ttJd2 bxa4 1 6.ixa4 cxd5 1 7.exd5
ixa4 1 8 .!ha4 ttJd7 1 9 .ie3 ifS 20.ttJc4 ib4t
2 l .@e2 !!ec8 22.b3 !!ab8 23.ttJxa5 ixa5
24.!!xa5 !!xb3 25 .!!d l @f8 26.d6 !!b7 27.!!c l
!!xc l 28 .!!a8t !!b8 29.!!xb8t ttJxb8 30.ixc l
@e8 3 l .@d3 @d7 32 .ia3 f5 33.g3 g5 34.h3
h5 35.@c4 @c6 36.@d3 ttJd7 37.f3 ttJ f6 38.g4
fxg4 39.fxg4 hxg4 40.hxg4 Y2-Y2 Hiarcs 8
- Smirin, Internet 2002.
lO.liJxe5 ge8 1 1 .0-0-0 liJa6
Not l l . . . !!xe5 ? 1 2 .!!d8t !!e8 1 3 .ixf6 !!xd8
1 4.ixd8.
l .liJf3 liJf6 2.c4 g6 3.liJc3 .ig7 4.e4 d6 5.d4
0-0 6.ie2 e5 7.dxe5
The first surprise. Very often this move is
made by players who have nothing against a
draw. But Tal? Soon I realized that the matter
is not so simple.
7 ... dxe5 s.Y;Yxd8 gxd8 9.ig5 c6
A few times I have played: 9 . . J!e8 l O.ttJd5
ttJxd5 l l .cxd5 c6 1 2.ic4 b5 1 3 .ib3 a5
a
12.gd6!?
b
c
d
e
f
g
h
336
King's Indian Warfare
I knew the continuation 1 2.f4 h6 1 3 ..ih4
g5 and so on, with good play for the pawn,
which Black usually wins back. The move
made by the ex-World Champion caught
me completely by surprise. Mter the game I
found out that it had been played for the first
time in 1 98 8 . Mikhail Nekhem' evich noticed
the move during a tournament in Tel Aviv in
which he was participating, shortly before our
encounter.
I could not cope with the unexpected move
and answered with one of the worst possible
options.
12 .. J!xe5?
The game that inspired Tal to play this variation
continued:
1 2 . . ..ie6 1 3 . f3
1 3 .f4 h6 1 4 . .ixf6 .ixf6 1 5 .�hd 1 .ixe5
1 6. fxe5 �ac8 happened in a game by two
legendary players: 1 7.�d2 �e7 1 8 .�e3
�ce8 1 9 .�d8 �f8 20.h4 ltJ c5 2 l .�xe8t
�xe8 22.b4 ltJ d7 23.�f4 �e7 24.ltJd5t
cxd5 25.cxd5 �c8 26.dxe6 fxe6 27.�d2 �f8t
Vz-Vz Korchnoi - Kasparov, Tilburg 1 99 1 .
1 3 . . . ttJc5 1 4 ..ie3 ltJa6 1 5 . f4 .if8 1 6.�dd 1 .ib4
17 ..if3 ltJ c5 1 8 . .ixc5 .ixc5 1 9 .�he 1
(Vz-Vz, 39) Bruk - Lev, Tel Aviv 1 990, and
White had a healthy extra pawn which he was
unable to convert later.
c
d
e
f
g
h
15 ... �xe4
Playing a pawn down would bring no greater
joy.
16J!d8t <it>g7 17.�xe4 gxe4
Still 1 2 . . . .ie6 was the right choice, as was
shown by Kasparov in the game above.
But 1 2 . . . ttJ h 5 ! ? deserves attention as well.
b
I missed this second rook trip to d6. Now
Black's position quickly becomes critical.
a
a
b
c
d
e
f
g
h
18 ..if3!
An important finesse. The ex-World
Champion sacrifices a pawn to keep the
c8-bishop firmly tied to the last rank. I was
naive to count on 1 8 . .id3 �e7 1 9.�d 1 b6,
when Black is more or less OK.
18 ... gxc4t 19.<it>d2 c5?
337
Chapter 1 1 - My Two Favourite Losses
A second and decisive mistake. I wanted to
create the possibility of exchanging a pair of
rooks via d4, but this does not work.
More stubborn was:
1 9 . . . b6 20.b3
20.!k 1 �xc l 2 1 .mxc 1 ib7 22.�d7 �b8
23.�c7± was the line Tal proposed in Chess
Slightly better was 20 . . . �d4 t 2 1 .�xd4 cxd4,
but even in this case after 22.�c7 ie6 23 .ixb7
�d8 24.b3 Black will hardly be able to keep
both weaknesses on d4 and a7 alive.
21. ci>xcl
It is already more or less clear that the pinned
bishop will cost Black at least an exchange.
Informant.
20.�e 1 also promises White better prospects.
20 . . . �f4!?
Worse is 20 ... �c5 2 1 .�e 1 ib7 22.�xa8
ixa8 23 .�e8 ib7 24.b4 �c4 25.b5 cxb5
26.ixb7 �a4 27.�e7 �xa2t 28.me3 and
here the bishop is much stronger than three
pawns - I think White should win.
2 1 .ixc6 �xf2t 22.me3 �e2t 23.mf4
Not 23.mxe2? ig4t.
23 . . . �b8
Black retains reasonable practical chances to
hold this.
21. .. h5 22.h4 gbs 23.ges f6 24 ..id5 g5
25.g3 gxh4 26.gxh4 ci> g6 27. ci> d2 b5
27 . . . ie6 would lose at once after 28.ie4t!.
28. ci> e3 b4
8
7
6
5
4
3
20J!cl !
2
8
1
7
6
a
5
b
c
d
e
f
g
h
29.a3! bxa3 30.bxa3 a5 3 1 . ci> f4 a4 32.£3!
Zugzwang.
4
3
8
2
7
1
a
b
c
d
e
f
g
h
I underestimated the strength of this move.
White simply exchanges the opponent's only
piece that is able to do something. After that
Black is left with an army of invalids.
6
5
4
3
2
20 .. J!xcl
a
b
c
d
e
f
g
h
King's I nd ian Warfare
338
32 ... ®g7 33 ..ie6 .ixe6 34J�xb8 .ib3
35J�b5
8
Moscow 1988
This game remains one of my most memorable
encounters, even though I happened to be
on the receiving end. The game was deeply
annotated by Garry Kasparov twice: together
with A. Nikitin in Chess Informant back in
1 988, and in his recent book Carry Kasparov
on Carry Kasparov Part 11: 1985-1993. I have
used some of his annotations here.
7
6
5
4
3
2
a
b
c
d
e
f
g
h
resigned due to the loss of either the
c-pawn or the h-pawn. A clean win by Mikhail
(or Misha, as he was called by his numerous
friends) Tal.
1-0
Garry Kasparov - Ilya Smirin
l .tiJf3 tiJf6 2.c4 g6 3.tiJ c3 .ig7 4.e4 d6 5.d4
0-0 6 ..ie2 e5 7.0-0 tlJc6 8.d5 tlJe7 9.tiJd2
a5 10.a3 tiJd7 l l .E:bl f5 12.b4
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
12 ... b6
As I remember, my preparation for this game
lasted about 1 0 minutes. When I returned
after the game I discovered the position after
Black's 1 2th move was still on the board in
my hotel room. Great preparation for my first
encounter with the champion of the world!
Actually the last move has been considered
second-best for about 25 years. The reason is
clear: Black prevents c4-c5 for a while, but
after White does carry out this advance, it will
gain in strength because of the pawn being on
b6.
339
Chapter 1 1 - My Two Favourite Losses
The main continuation here became:
1 2 . . . @h8 1 3 .f3
Or instead 1 3.'1Nc2 l2Jg8 ( 1 3 . . . t2J f6! ? 1 4 . f3
axb4 1 5 .axb4 c6 is an interesting alternative)
1 3 . . . b6, which I played later in 1 988 against
Gavrikov, see page 1 32.
13 . . . t2Jg8
This was played several times by Kasparov
himself. The most famous of these games
you may find in the annotations to the game
Kasimdzhanov - Smirin in this book, as well
as those to Khalifman - Kasparov, Paris 1 99 1
(see page 1 7 1 ) .
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
In the post mortem Carry Kimovich told me
that he had thought that Black could not hold
both of the weak pawns on b6 and d6, and
that the game was basically over, but . . .
13.f3
1 8. . .i.d7 19.�c4 �c8
Eureka! Everything is defended and the
battle continues.
20.i.a3 �e8
8
a
b
c
d
e
f
g
h
13 ... £4
Black unambiguously reveals his plan attack on the kingside by means of . . . g5 etc.
The exchange of the dark-squared bishops
deserved some attention: 1 3 . . ..ih6 1 4 . t2J b3
.ixc l Still, after 1 5 .Wxc l axb4 1 6.axb4 f4
1 7.c5 bxc5 1 8 . bxc5 t2J f6 1 9 .!h1 gb8 20.'1Na3
White is ahead in creating concrete threats.
I4.ltJa4 axb4 1 5.axb4 g5 16.c5 �f6 17.cxd6
cxd6 18.b5
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Black's pieces are quite passive at the
moment, but they successfully withstand
the opponent's pressure on the kingside.
Meanwhile he is ready to play . . . h7-h5,
... gf6-g6 and ... g5-g4, which looks threatening.
The World Champion decides to nip that in
the bud.
340
King's Indian Warfare
2 1 .g4!
23 ... gxf3 24.ha tl)f6
Trying to seal (or almost seal) the kingside.
Now I faced a complicated choice: to take en
passant or to leave it as it is.
Kasparov recommends 24 ... �c7 as a more
precise continuation.
25 ..ig5
2 1 . .. fxg3?!
I agree with Kasparov - this is a dubious
decision, even though he (and Nikitin) marked
it as "the only move" in Informant back in
1 98 8 .
Correct is hi s latter recommendation: 2 1 . . . h5
22.h3 �f7 (22 ... �f6!?) 23.<j;>g2 .if8, with an
unclear position - Black will play . . . �h7 and
speculate, with the intention of opening the
h-file at the appropriate moment.
An unpleasant pin - White has more space,
and that allows him to set up his pieces very
actively.
25 .. J�a7 26J:U2 �b7!?
Overprotecting the pawn on b6, with the
idea of freeing the queen from this function
and playing . . . �e8.
27.�b3
8
7
22.hxg3 g4!
6
8
5
7
6
4
4
2
3
5
3
2
a
a
b
c
d
e
f
g
h
The only chance to create some counterplay
- otherwise White would proceed tt:J e3 , g4,
tt:J f5 and so on.
23 ..icl !
Controlling g5.
23.fxg4 �xfl t 24 .mxfl (if 24 ..ixfl then
24 . . . tt:J f6, attacking the pawns on e4 and g4)
24 . . . �g5 25 . .ic l �g6 26 ..if3 tt:J f6! would be
welcomed by Black.
b
c
d
e
f
g
h
27 ... �a7
I do not remember why I did not fulfil
my intention to play 27 . . . �e8 : it was quite
possible, attacking the pawn on b5 and
trying to activate the queen via g6. After the
probable 28 ..ixf6 �xf6 29 . .ih5 �e7 30.�xf6
.ixf6 3 1 ..ig4 .ig5 Black would still be worse,
but his defence is hard to crack - White has
problems activating the knight on a4. But the
move made in the game is hardly worse - at
least from a practical standpoint.
28.�bl
Kasparov writes that he was in doubt as to
whether he would win after 28.tt:Jaxb6!? tt:J xb6
Chapter 1 1
-
34 1
My Two Favourite Losses
8
29.lt:Jxb6 ltWxb6 30.ie3 ltWa5 3 l .ixa7 ltWxa7
32.b6, and he decided to repeat moves to save
some time on the clock.
7
28 .. J�b7 29J�b3 �a7 30.�b4!
5
6
8
4
6
2
3
7
5
4
b
d
f
g
h
A very strong move, indirectly applying even
more pressure against the knight on f6.
a
3
2
1
a
b
c
d
e
f
g
h
This rook cements the queenside, protecting
virtually everything.
A move made JUSt in case". True, Black
has a limited number of ideas, but 30 .. J�b7
(why not continue to move the rook back and
forth?) with the same intention to play . . . ltWe8
would be stronger - I am still not sure if White
has more than a "pleasant position" after that.
30... ®h8?!
c
e
3 t . ..Lbs
As I remember I still had some optimism
at this point, hoping to somehow repel the
champion's assault.
32.�xb5 �xa4 33 ..ig2
33.id l ! El:al 34.lt:Je3 would be stronger Black has no satisfactory defence after that,
despite the pin on the first rank. But honestly
speaking, to play this during a game is not very
realistic, even for the World Champion.
33 ... h6 34..ih4 �e8!
8
I did not play the immediate 30 . . .ltWe8? due to
3 l .lt:J axb6 lt:Jxb6 32.lt:Jxb6 ixb5 33.ixf6 El:xf6
34.ih5, and Black loses the piece.
7
3 1 .�fl !
5
6
4
3
2
a
b
c
d
e
f
g
h
342
King's I ndian Warfare
8
The only move which retains hope - passive
defence would be doomed.
7
6
35.hf6 gxf6
5
Not 3 5 . . . \Wxb S ? 36.ixg7t <i>xg7 37.�xf8 .
4
36.gxf6 'lMxb5
3
8
2
7
6
5
a
4
b
c
d
e
f
g
h
38 ..ih3!!
3
2
a
b
c
d
e
f
g
h
At this moment we both were in time
trouble. Unfortunately, I do not remember
exactly when Kasparov took his time and
thought deeply about the position - now or
before his 35 th move? Anyway, after that he
played very quickly and without hesitation.
37.ge6!
Again I do not remember exactly, but
perhaps I did not see this.
37.�f8t? ixf8 38.\Wxf8t <i>h7 39.\Wf7t would
lead only to a draw (39 .ih3? is bad due to
39 . . . \WcSt 40. <i>g2 �a2t 4 L<i> h l �al t) .
37 ... ®g8!
The only move - the king covers f7. The
tension has reached its peak now. No one
among the numerous spectators (several strong
grandmasters among them) saw how White
could continue the attack. Neither did I .
Of course, 37 . . . �xc4 38 .\Wf7 would lose o n the
spot.
Just an amazing idea! Kasparov was visibly
nervous and he made this move very fast.
I think it's irrelevant that Gary Kimovich
himself awarded the move a "?"in his
book.
Yes, the computer's vanauon 3 8.if3 �b4
39.id l , "with an irresistible attack on the
light squares" (Kasparov) is objectively much
stronger - White is winning. Kasparov adds
quite a few variations which you may find in
his book. But first of all, it would be absolutely
unreal to play this during a game, especially
in time trouble. And even more importantly
in my opinion: why replace the real beauty of
chess with a computer's mathematics? Even for
the sake of the truth . . .
38 ... gxc4?
I was kind of disoriented, and in time trouble
I made the most natural move - not seeing the
beautiful refutation.
But I am not sure if I would have found the
defence even with a couple of hours left on the
clock (perhaps I would have done it just by a
process of elimination) . Indeed, it's absolutely
extraordinary to see the only saving move
IS:
343
Chapter 1 1 - My Two Favourite Losses
3S . . . h5!! 39.lt:Je3
39 .�h6? does not work - 39 . . ..ixh6 40 . .ie6t
<i>hS 4 1 .�f6t .ig7 42.�dSt <i>h7 43 ..if5t
@ h6 - the pawn on h5 covers the black king
from a lethal check from h4.
39 . . . �xfl t 40 ..ixfl
40. <i>xfl @f7 4 1 . ttJ f5 .ifS and everything is
protected.
40 . . . �xe4 4 1 .lt:J f5 .ifS 42.ib5
Or 42.�eS lt:Je7! (another finesse) 43.lt:Jxe7t
@f7 44.�xfSt <i>xfS 4 5 . lt:J f5 h4 with equal
chances.
42 . . . �e l t 43. <i>g2 �c l 44.�eS <i>f7
Black successfully defends (Kasparov) .
Needless to say, 3S . . . h5!! was discovered in the
computer era - the modern Houdini "sees"
this move in a fraction of a second.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
Just beautiful, isn't it?
40.�e6t ®hs 41.�f6t
I resigned due to the forced checkmate:
4 l . .. <i>h7 (4 1 . .. .ig7 42 .�h4t)
8
7
39J�xh6!!
8
6
5
7
4
6
3
5
2
4
3
a
b
c
d
e
f
g
h
42.�f7t .ig7 43 ..if5 t <i>hs 44.�h5t <i>gS
45 . .ie6t <i>fS 46.�f7#
2
1
1-0
b
d
f
g
h
A bolt from the blue. I was astonished, but
even if I hadn't been, it's too late for Black to
do anything.
a
c
e
A picturesque checkmate with a pinned
white queen would be the result of: 39 . . . �c5t
40.<i>h l �cl 4 l ..ie6#
39 ...1xh6
A real masterpiece by the then World
Champion. Alexander Nikitin, who coached
Kasparov for many years, wrote in his book
that he considers this game one of the best in
Kasparov's outstanding career. (By the way, I
had a conversation with Nikitin a few days after
this game, and he told me not to quit playing
the KI D, because I play it well. I have followed
his advice until now!) I have to confess that I
never gained such a creative impression from a
lost game as I did from this one.
Game Index
Chapter 1
Nimrod Veinberg - Ilya Smirin, Jerusalem 20 1 5
Yannick Pelletier - Andreas Skytte Hagen, Legnica 20 1 3
Lembit 011 - Ilya Smirin, Rostov-on-Don 1 993
Portisch - Kasparov, Linares 1 990
Reshevsky - Najdorf, Zurich 1 9 53
Van der Sterren - Smirin, Tilburg 1 992
Alexey Kuzmin - Ilya Smirin, Nabereznye Chelny 1 988
Baron - Smirin, Israeli Championship 20 1 6
Vitiugov - Ganguly, Khanty-Mansiysk (ol) 20 1 0
Lautier - Smirin, Rethymnon 2003
Kuzmin - Badea, Moscow 1 99 1
Alexander Beliavsky - Ilya Smirin, Belgrade 1 998
Kazhgaleyev - Shchekachev, Livry Gargan 200 5
Pinter - Tkachiev, Porec 1 998
A. Mikhalevski - Smirin, Rishon Le Zion 1 997
Boris Avrukh - Ilya Smirin, Israel 1 999
Alexander Grischuk - Ilya Smirin, Beersheba 2005
Sokolov - Smirin, Groningen 1 993
Chapter 2
Vladimir Kramnik - Garry Kasparov, Linares 1 994
Vladimir Kramnik - Garry Kasparov, Novgorod 1 997
Goncalves - Sender, corr. 200 1
Szeberenyi - Zimmerman, Budapest 1 999
Baeckstroem - Petraitis, corr. 2002
Vladimir Kramnik - Ilya Smirin, Moscow 1 992
Beliavsky - Ilya Smirin, Moscow 1 98 8
Zaid - Smirin, Israel 1 99 1
Vladimir Kramnik - Ilya Smirin, Moscow 1 99 5
Cramling - Smirin, Stockholm 20 1 2
Kramnik - Ivanchuk, Moscow 1 995
Vladimir Kramnik - Ilya Smirin, Belgrade 1 999
Pelletier - Nakamura, Skopje 20 1 5
Vladimir Kramnik - Ilya Smirin Russia - ROW (rapid) 2002
Chapter 3
Wang Yue - Teimour Radjabov, Wijk aan Zee 2009
Florin Gheorghiu - Garry Kasparov, Thessaloniki Olympiad 1 988
Teimour Radjabov - Ding Liren, Wijk aan Zee 20 1 5
Jan Gustafsson - Vladimir Kramnik, Dortmund 20 1 2
Loek van Wely - Daniel Stellwagen, Amsterdam 2009
Van Wely - Degraeve, Mondariz 2000
11
12
14
15
15
15
18
18
19
19
19
22
23
23
23
27
33
33
41
43
44
45
45
46
47
47
52
53
53
59
59
63
69
70
71
72
73
73
Game Index
Van Wely - Radjabov, Dresden (ol) 2008
Giorgi Kacheishvili - Ilya Smirin, Minneapolis 2005
Macieja - Smirin, Maalot-Tarshiha 2008
Wendt - Smirin, Ohrid 2009
Mchedlishvili - Avrukh, Beersheba 2005
Mchedlishvili - Hillarp Persson, Gothenburg 200 5
Denis Khismatullin - Ilya Smirin, Poikovsky 20 1 5
Avrukh - Smirin, Tel Aviv 1 996
Alterman - Xu, Cap d'Agde 1 994
Ubilava - Grigoriants, Linares 2003
Christopher Lutz - Ilya Smirin, Groningen 1 993
Bykhovsky - Smirin, Haifa 2008
Boris Alterman - Ilya Smirin, Israel 1 99 1
Dreev - Smirin, Dos Hermanas 200 1
Bobotsov - Tal, Varna 1 9 58
Igor Naumkin - Ilya Smirin, Ischia 1 99 5
Ivanchuk - Smirin, Pinsk 1 986
I. Ivanov - Smirin, Philadelphia 1 990
Savchenko - Smirin, St Petersburg 20 1 2
Chapter 4
Mark Taimanov - Miguel Najdorf, Zurich 1 9 53
Tatiana Shumiakina - Narine Karakashian, Volzhsky 1 989
Igor Bjelobrk - Alexander Grischuk, Troms0 ( 1 . 1 ) 20 1 3
Giovanni Vescovi - Teimour Radjabov, Bursa 20 1 0
Levon Aronian - Magnus Carlsen, Nice (blindfold) 2 0 1 0
Wesley So - Hikaru Nakamura, Sinquefield 20 1 5
Per Auchenberg - Brian Elgaard, Copenhagen 1 992
Alexander Huzman - Ilya Smirin, Sverdlovsk 1 987
Korchnoi - Hulak, Zagreb 1 987
Korchnoi - Ye Jiangchuan, Novi Sad
Yuri Dokhoian - Ilya Smirin, Sverdlovsk 1 987
Doroshkievich - Tal, Yerevan 1 975
Christiansen - Kasparov, Moscow 1 982
Rafael Leitao - Ilya Smirin, New York 2000
Yakovich - Smirin, Munich 1 993
Yusupov - Gelfand, Dortmund 1 997
Yusupov - Smirin, Yerevan 1 996
Viktor Gavrikov - Ilya Smirin, Klaipeda 1 988
Ivanchuk - Kasparov, Linares 1 992
Gavrikov - Kasparov, Moscow 1 98 8
Lputian - Frolov, Helsinki 1 992
Igor Novikov - Ilya Smirin, Las Vegas 1 999
Kaidanov - Smirin, Mallorca 2004
345
73
76
76
76
77
77
81
81
83
83
87
88
93
93
94
101
1 02
1 03
1 03
1 09
111
1 12
1 13
1 14
115
1 17
118
1 18
1 19
1 23
1 24
125
1 28
1 28
1 29
1 29
1 32
1 33
1 33
1 34
141
1 42
346
King's Indian Warfare
Dreev - Smirin, Biel 1 993
Evgeny Postny - Ilya Smirin, Maalot-Tarshiha 2008
Aronian - Radjabov, Sofia 2008
Bareev - Smirin, Moscow 2002
Mikhalevski - Smirin, Philadelphia 2008
Lputian - Smirin, Kemer 2007
Chapter 5
Aleksej Aleksandrov - Alexei Fedorov, Minsk 20 1 6
Eesha Karavade - Pierre Bailet, Nancy 20 1 6
Alexei Barsov - Ilya Smirin, Khanty-Mansiysk Olympiad 20 1 0
Cvitan - Smirin, Tilburg 1 993
Durarbayli - Smirin, Plovdiv 20 1 2
Mamedyarov - Smirin, Khanty-Mansiysk (blitz) 20 1 3
Mikhail Nedobora - Ilya Smirin, Israel 2008
Milov - Smirin, Haifa 1 99 5
Aleksandrov - Smirin, New York 1 998
Kiril Georgiev - Ilya Smirin, Elenite 1 994
Chapter 6
Alexander Khalifman - Garry Kasparov, Paris (rapid) 1 99 1
Mesgen Amanov - Ilya Smirin, King of Prussia 20 1 0
Greenfeld - Smirin, Haifa 1 993
lvanchuk - Smirin, Odessa 2007
Michalik - Smirin, Jerusalem 20 1 5
Avrukh - Smirin, Netanya 2009
Boris Chatalbashev - Ilya Smirin, Athens 2007
Khalifman - Smirin, Sverdlovsk 1 987
Malakhov - Grischuk, Moscow 20 1 0
Lev - Smirin, Haifa 1 993
Moiseenko - Smirin, Maalot-Tarshiha 2008
Rustam Kasimdzhanov - Ilya Smirin, Elista Olympiad 1 998
Larsen - Fischer, Denver (4) 1 97 1
Karpov - Kasparov, Skelleftea 1 989
M. Gurevich - Smirin, Haifa 1 99 5
M . Gurevich - Smirin, Elenite 1 994
Tigran Kotanjian - Ilya Smirin, European Championship, Yerevan 20 1 4
Boris Alterman - Ilya Smirin, Israeli League 2004
Dydyshko - Vokac, Ostrava 200 5
Razuvaev - Zagrebelny, Moscow 1 983
Parligras - Smirin, Skopje 20 1 4
Yaacov Zilberman - Ilya Smirin, Tel Aviv 1 999
Petrosian - Geller, Moscow 1 949
Zilberman - Smirin, Israel 20 1 1
Gligoric - Fischer, Siegen (ol) 1 970
1 42
1 46
1 46
1 46
1 46
1 46
1 53
1 53
1 54
155
1 55
1 56
1 60
1 60
161
1 64
171
1 73
1 73
1 74
1 74
1 75
1 78
1 79
1 79
1 79
1 79
1 82
1 82
1 83
1 84
1 84
1 87
1 92
1 92
1 92
1 93
1 99
1 99
1 99
1 99
Game I ndex
Zilberman - Smirin, Israel 2007
Zilberman - Smirin, Panormo 200 1
Vaganian - Smirin, Rostov-on-Don 1 993
Evgeny Bareev - Ilya Smirin, Moscow (rapid) 2002
Eljanov - Smirin, Porto Carras 20 1 1
Korobov - Smirin, Nakhchivan 20 1 1
Chapter ?
Andrey Zontakh - Jaroslav Ulko, Voronezh 20 1 5
A. Fedorov - Ozolin, Kazan 20 1 3
Vyacheslav Dydyshko - Ilya Smirin, Byelorussia 1 986
Aleksandrov - Smirin, Minsk 1 987
A. Petrosian - Smirin, Polanica Zdroj 1 989
lvery Chighladze - Ilya Smirin, Baku 2009
Postny - Smirin, Israel 20 1 0
L . Kaufman - Smirin, Rockville 20 1 3
Petr Haba - Ilya Smirin, Polanica Zdroj 1 989
Smyslov - Polugaevsky, Riga 1 975
Bandza - Smirin, Dortmund 1 990
Basin - Smirin, Minsk 1 987
M . Gurevich - Smirin, Moscow 1 988
Nakamura - Vachier-Lagrave, London 20 1 5
Ufimtsev - Tal, USSR 1 967
Chapter 8
Abhijeet Gupta - Kaido Kulaots, Cappelle-la-Grande 20 1 5
Wimmer - Cill6niz Razzeto, email 20 1 4
Viktor Korchnoi - Ilya Smirin, Dresden 1 998
Teimour Radjabov - Ilya Smirin, Dos Hermanas 200 1
Jaan Ehlvest - Ilya Smirin, Khanty-Mansiysk 2009
Korchnoi - Kasparov, Debrecen 1 992
Dydyshko - Ehlvest, Tallinn 1 980
Josh Manion - Ilya Smirin, Las Vegas 1 997
Kourkounakis - Smirin, Gausdal 1 990
Petrosian - Gligoric, Rovinj/Zagreb 1 970
Boris Gelfand - Ilya Smirin, Saint Vincent 2005
Kantsler - Smirin, Jerusalem 2005
Colovic - Smirin, Skopje 20 1 4
Bercys - Smirin, Philadelphia 20 1 1
King - Kveinys, Germany 1 993
loannis Nikolaidis - Ilya Smirin, Panormo 200 1
Moiseenko - Cheparinov, Kemer 2007
Beim - Smirin, Tel Aviv 1 996
Agrest - Nithander, Copenhagen 2007
347
200
200
20 1
205
206
207
215
215
218
218
219
223
223
224
227
228
228
229
230
23 1
232
235
235
237
242
25 1
252
25 2
256
256
25 7
26 1
26 1
262
262
262
268
268
269
269
348
King's Indian Warfare
Chapter 9
Johan-Sebastian Christiansen - Ilya Smirin, Stockholm 20 1 6
Loek van Wely - Teimour Radjabov, Biel 2007
Maxim Rodshtein - Ilya Smirin, Ashdod 2006
Gelfand - Smirin, Rishon Le Zion 2006
Yevseev - Loginov, Kazan 200 5
Golod - Areshchenko, Port Erin 2007
David Anton Guijarro - Ilya Smirin, Yerevan 20 1 4
Serper - Smirin, Philadelphia 1 996
Nabaty - Smirin, Skopje 20 1 3
Bologan - Al Sayed, Qatar Masters 20 1 5
Anton Guijarro - Naroditsky, Al Ain 20 1 3
Petrosian - Geller, Moscow 1 949
Evgeny Postny - Ilya Smirin, Acre 20 1 3
Roiz - Smirin , Tel Aviv 2002
Speelman - Kasparov, Madrid (rapid) 1 98 8
Gershon - Smirin, Israeli Team Championship 2007
Ronen Lev - Ilya Smirin, Ramat Gan 1 992
Illescas Cordoba - Smirin, Oviedo (rapid) 1 993
Ivanov - Smirin, Los Angeles 2000
Chapter 10
Fabiano Caruana - Hikaru Nakamura, Wij k aan Zee 20 1 3
Yuri Yakovich - Ilya Smirin, Saint-Vincent 2000
Bareev - Kasparov, Linares 1 992
Hauchard - Hebden, Cappelle la Grande 1 998
Yakovich - S. Bekker Jensen, Gothenburg 2000
Mihail Marin - Ilya Smirin, Ramat Aviv 2000
Krasenkow - Smirin, Polanica Zdroj 1 99 5
Goldin - Kosanovic, Belgrade 1 98 8
Pigusov - Smirin, Las Vegas 1 998
Christian Gabriel - Ilya Smirin, Pula 2000
Ostenstad - Nakamura, Rhodes 20 1 3
Piket - J . Polgar, Aruba 1 99 5
Korobov - Smirin, Eilat 20 1 2
Sabino Brunello - Ilya Smirin, Helsingor 20 1 5
Karpov - Kasparov, Lyon/New York 1 990
Carlsen - Grischuk, Dubai (rapid) 20 1 4
Dreev - Wang Hao, Troms0 20 1 3
Chapter 1 1
Mikhail Tal - Ilya Smirin, Podolsk 1 990
Hiarcs 8 - Smirin, Internet 2002
Korchnoi - Kasparov, Tilburg 1 99 1
Garry Kasparov - Ilya Smirin, Moscow 1 988
275
275
278
279
279
279
282
283
283
284
284
284
288
289
290
290
294
295
296
303
304
304
305
305
3 10
310
31 1
31 1
316
316
316
317
323
323
324
324
335
335
336
338
Name Index
A
Aagaard
Agrest
Aleksandrov
AI Sayed
Alterman
Amanov
An and
A. Petrosian
Areshchenko
Aronian
Auchenberg
Averbakh
Avrukh
c
8, 329
269
1 5 3, 1 6 1 , 2 1 8
284
83, 93, 9 5 , 98, 1 00,
1 82, 1 92, 1 93, 1 9 5 , 1 96
1 73, 1 75 , 1 77
52, 63, 1 1 5
219
279
1 1 4, 1 1 5 , 1 46
1 17
218
1 1 , 23, 27, 29, 30, 32,
77, 8 1 , 82, 1 7 5 , 1 82
B
Badea
Baeckstroem
Bailer
Balcazar Novoa
Bareev
Baron
Barsov
Basin
Beim
Bekker Jensen
Beliavsky
Benjamin
Bercys
Bjelobrk
Bobotsov
Bologan
Bronstein
Bruk
Brunello
Bykhovsky
59, 1 46, 20 5 , 208, 2 1 0,
1 54, 1 56,
229,
22, 24, 25, 26, 27, 47,
283,
7,
323,
19
45
1 53
44
304
18
1 73
230
269
305
1 23
1 79
262
1 12
94
284
1 09
336
325
88
Carlsen
Caruana
Chatalbashev
Cheparinov
Chighladze
Chigladze
Christiansen
Cill6niz Razzeto
Colovic
Craciunescu
Cramling
Cvitan
24, 1 1 4, 324, 325
303
1 78, 1 79, 1 80
268
223
1 8 , 226
1 25 , 275
235
262
44
53
1 55
D
Dam sky
Degraeve
Delchev
Oing Liren
Dokhoian
Doroshkievich
Dreev
Durarbayli
Dydyshko
23 1 , 335
73
206
71
1 23, 1 24
1 24
93, 1 42, 247, 323 , 324
155
1 92, 2 1 8, 222, 252
E
Ehlvest
Elgaard
Eljanov
25 1 , 252, 253
1 17
206
F
Fedorov
Fischer
Frolov
1 53, 2 1 5 , 268
5, 7, 33, 1 82, 1 99
1 34
King's Indian Warfare
350
K
G
Gabriel
Ganguly
Gavrikov
Gelfand
Geller
Georgiev
Gershon
Gheorghiu
Gligoric
Goldin
Golod
Goncalves
Greenfeld
Grigoriants
Grischuk
Gufeld
Guijarro
Gupta
Gurevich
Gustafsson
3 1 6, 3 1 8 , 3 1 9 , 322
19
1 32, 1 33 , 1 34,
1 37, 1 4 1 , 1 83 , 339
3 , 6, 63, 1 1 5 , 1 1 8 , 1 23 ,
1 29, 1 60, 1 82, 243, 244, 26 1 ,
262, 263, 265 , 278, 279, 288
7, 1 83 , 1 84, 1 99, 284
1 64, 1 67
290
70, 7 1 , 72, 87
1 09, 1 99, 256, 257
31 1
279
45
1 73
83
7, 33, 3 5 , 37, 63,
1 1 2, 1 28 , 1 79, 324, 325
223
282, 284, 285, 288
235, 236
1 84, 230, 268
72
H
Haba
Hagen
Hammer
Hauchard
Hebden
Hillarp Persson
Hulak
Huzman
2 1 5 , 227, 228, 232
1 2, 1 3
1 54
305
305
77
1 18
1 1 8 , 1 22, 1 23 , 206
I
Illescas Cordoba
lvanchuk
lvanov
295
52, 53, 63, 1 02, 1 23 , 1 33 ,
1 74, 30 1
1 03 , 296
Kacheishvili
76
Kaidanov
1 42
26 1
Kantsler
111
Karakashian
1 53
Karavade
Karpov
63, 1 23, 1 7 1 , 1 73 , 1 83 , 323
Kasimdzhanov
1 82, 1 83 , 339
Kasparov
7, 1 5 , 4 1 , 43, 46, 52, 63,
70, 72, 87, 1 1 0, 1 23 , 1 24, 1 25 , 1 32,
1 33, 1 7 1 , 1 72, 1 73 , 1 82, 1 83 , 25 1 ,
252, 289, 290, 304 , 305, 323, 334,
336, 338, 339, 340, 342, 343, 344
224
Kaufman
Kazhgaleyev
23
Khalifman
67, 1 23 , 1 64, 1 7 1 , 1 79,
339
8 1 , 84
Khismatullin
262
King
Korchnoi
1 1 8, 1 1 9, 237, 238, 239,
240, 24 1 , 25 1 , 252, 26 1 , 262, 336
206, 207, 3 1 7
Korobov
Kosanovic
31 1
Kotanj ian
1 87, 1 89
256
Kourkounakis
3, 5 , 39, 40, 4 1 , 42, 43, 44,
Kramnik
4 5 , 46, 5 1 , 52, 53, 59, 62, 63, 6 5 ,
72, 73, 1 1 5 , 1 32, 206, 242, 3 1 6
235
Kulaots
Kuzmin
1 8 , 1 9 , 2 1 , 268
L
Larsen
Lautier
Leitao
Lev
Loginov
Lputian
Lutz
1 82
19
82, 1 28, 1 3 1 , 1 42
1 79, 294
279
1 34, 1 46
87, 88
Name Index
M
Macieja
Malakhov
Malaniuk
Manion
Marin
Mchedlishvili
Michalik
Mikhalevski
Milov
Moiseenko
Morozevich
N
R
76
1 79
1 23
256, 258, 260
310
77
1 74
23, 1 46, 1 47
1 60
1 79, 2 1 3 , 268
1 8 , 63
1 5 , 1 09, 1 1 0, 1 1 8 , 1 1 9
Najdorf
Nakamura
7, 59, 1 1 5 , 23 1 , 303, 3 1 6
284
Naroditsky
1 0 1 , 1 04, 1 06, 200
Naumkin
1 60, 1 6 1
Nedobora
338, 340, 344
Nikitin
268, 270
Nikolaidis
243, 244
Nisipeanu
269
Nithander
31 1
Nogueiras Santiago
273
Notkin
1 4 1 , 1 43 , 1 44
Novikov
0
011
Ostenstad
Ozolin
p
14, 1 5, 17
316
215
1 92, 1 93
1 2, 1 3 , 59
45
1 8 1 , 1 99, 256,
257, 284, 294
31 1
Pigusov
Piket
316
Pi mer
23
316
Polgar
228
Polugaevsky
15
Portisch
Postny
87, 88, 1 46, 1 47, 223, 288, 293
Parligras
Pelletier
Petraitis
Petrosian
35 1
Radjabov
Razuvaev
Reshevsky
Rodshtein
Roiz
Romanishin
Rubinstein
s
Sakaev
Salov
Samisch
San Segundo
Savchenko
Seirawan
Sender
Serper
Shchekachev
Shirov
Shumiakina
Smyslov
So
Sokolov
Spassky
Speelman
Stein
Stellwagen
Svidler
Szeberenyi
T
7, 64, 69, 7 1 , 74,
1 1 3 , 1 32, 1 46, 242, 243,
246, 250, 275 , 276, 277
1 92
15
1 60, 278, 28 1
289
26 1 , 262
227
41
1 23
223
206
1 03
1 79
45
283
23
63, 25 1
Ill
228
1 15
33, 43
181
290
5, 7
73, 74, 75
63
45
1 09, 1 1 8, 1 1 9
Taimanov
5, 7, 93, 94, 1 1 0, 1 24, 2 1 5, 23 1 , 232,
Tal
257, 333, 334, 335, 336, 337, 338
49, 280
Tarrasch
23, 323
Tkachiev
205
Tolstoy
1 64
Topalov
1 23
Tseshkovsky
1 23
Tukmakov
352
King's Indian Warfare
u
Ubilava
Ufimtsev
Ulibin
Ulko
X
83
232
41
215
y
1 7, 1 23 , 20 1
15
73, 74, 206, 275 , 276, 277
11
1 13
19
1 92
z
V
Vaganian
Van der Sterren
Van Wely
Veinberg
Vescovi
Vitiugov
Vokac
w
Wang Hao
Wang Yue
Wantola
Wendt
Wimmer
Xu
Yakovich
Ye Jiangchuan
Yevseev
Yusupov
Zagrebelny
Zaid
Zilberman
Zimmerman
Zontakh
323, 324
69
279
76
235
83
1 28 , 304, 305, 309
1 19
278 , 279
1 23, 1 29
192
47
1 02 , 1 8 5 , 199,
200, 202, 203
45
215