beating king's indian and benoni

148

Upload: martiniano-saldivar-zuniga

Post on 10-Oct-2014

960 views

Category:

Documents


56 download

TRANSCRIPT

Page 1: Beating King's Indian and Benoni
Page 2: Beating King's Indian and Benoni
Page 3: Beating King's Indian and Benoni

Beating the King's Indian and Benoni

Anatoli Vaisser

B. T. Batsford Ltd, London

Page 4: Beating King's Indian and Benoni

First published 1997 © Anatoli Vaisser 1997

ISBN 071348022 X

British Library Cataloguing-in-Publication Data. A catalogue record for this book is available from the British Library.

All rights reserved. No part of this book may be reproduced, by any means, without prior permission of the publisher.

Typeset and edited by First Rank Publishing, Brighton and printed in Great Britain by Redwood Books, Trowbridge, Wilts for the publishers, B. T. Batsford Ltd, 583 Fulham Road, London SW6 5BY

I am grateful to Mr. Destrebecq, a lover of the Four Pawns Attack, for making available his large collection of thematic material.

A BA TSFORD CHESS BOOK Editorial Panel: Mark Dvoretsky, Jon Speelman Commissioning Editor: Paul Lamford General Manager: David Cummings

Page 5: Beating King's Indian and Benoni

Contents

Preface

Part 1. Systems common to the King's Indian and Modern Benoni

(1 d4 ttJf6 2 c4 g6 3 ttJc3 i.g7 4 e4 d6 5 f4 0-0 6 ttJf3 c5 7 d5 e6 8 .te2 exd5 9 cxd5 or 1 d4 ttJf6 2 c4 c5 3 d5 e6 4 lLlc3 exd5 5 cxd5 d6 6 e4 g6 7 f4 .tg7 8 ttJf3 0-0 9 i.e2)

5

1 Main Line with 9 ... lXe8 7 Game 1: Vaisser-Berthelot, French Team Ch. 1992 8 Game 2: Vaisser-Degraeve, Cappelle la Grande 1987 21 Game 3: Cranbourne-Crespo, corr. 1988-89 28 Game 4: Lagontrie-Boulard, French Corr. Ch. 1987 32 Game 5: Vaisser-Ibragimov, Bern 1992 39

2 Main Line with 9 ... i.g4 48 Game 6: Vaisser-Yrjola, Helsinki 1991 48 Game 7: Vaisser-Berelovich, Groningen 1993 52 Game 8: Vaisser-Smirin, PCA (rapidplay) Moscow 1996 57 Game 9: Monin-Shchekachev, St Petersburg 1994 61

3 Main Line: 9 ... b5 and rare moves 69 Game 10: Komarov-Frolov, Kiev 1995 69 Game 11: Blokh-Kitchev, USSR (corr.) 1991 73

Page 6: Beating King's Indian and Benoni

Part 2. Systems specific to the King's Indian

(1 d4 tbf6 2 c4 g6 3 tbc3 ~g7 4 e4 d6 5 f4)

4 Black plays 6 ... tba6 79 Game 12: Zsu. Polgar-Chiburdanidze, St Petersburg 1995 80 Game 13: Vaisser-Golubev, Biel1995 86

5 The Pseudo Benko Gambit 7 ... b5 95 Game 14: Nogueiras-Sax, Graz 1984 95

6 Other Systems for Black 106 Game 15: Vaisser-Krasenkov, Paris 1990 106 Game 16: Vaisser-Kr. Georgiev, French Team Ch. 1996 111

Part 3. Systems specific to the Modern Benoni

(1 d4 tbf6 2 c4 c5 3 d5 e6 4 tbc3 exd5 5 cxd5 d6 6 e4 g6 7 f4)

7 Taimanov System 8 ..tb5+ 119 Game 17: Sokolov-Topalov, Wijk aan Zee 1996 120 Game 18: Yuneev-Kostometov, St. Petersburg 1995 128

8 Mikenas Attack 8 e5 135 Game 19: Meszaros-Stefanov, Satu Mare 1987 135

Index of Variations

Symbols

;!;;

± +-+ +

White is slightly better White is clearly better White is winning Black is slightly better Black is clearly better

144

-+ Black is winning = The position is equal Ch. Championship corr. Correspondence game

Page 7: Beating King's Indian and Benoni

Preface

The King's Indian Defence currently enjoys a very high standing amongst queen's pawn openings: it figures in the repertoire of many of the world's leading grandmasters, with Garry Kasparov at their head. Black goes all out for complex play and will often launch an attack on the white king.

How is White to oppose this dynamic opening? The Four Pawns Attack offers a fitting answer: immediate and total occupation of the centre with pawns, seizure of the initiative, and sometimes a 'Romantic' attack against the king without shirking from sacrifices. At the same time it gives White a formidable weapon against the Modem Benoni, since the main lines of the Four Pawns Attack can also arise from the Benoni by transposition. It is worth noting, however, that most Modem Benoni specialists - such as Psakhis for example - no longer play that opening in its 'pure' form but only reach it via the move order 1 d4 4Jf6 2 c4 e6 3 4Jf3 c5. Among other things, this move order enables Black to avoid the dangerous Taimanov system (1 d4 4Jf6 2 c4 c5 3 d5 e6 4 4Jc3 exd5 5 cxd5 d6 6 e4 g6 7 f4 ~g7 8 .tb5+).

Often the name 'Four Pawns Attack' is reserved for systems arising after the King's Indian order of moves (1 d44Jf6 2 c4 g6 3 4Jc3 iLg7 4 e4 d6 5 f4). Here, however, I have taken the more natu­ral view that any system in which White plays d2-d4, c2-c4, e2-e4 and f2-f4, including the Taimanov and Mikenas systems in the Modem Benoni, is considered to belong to the Four Pawns Attack.

In the mid-1920s the Four Pawns Attack was employed by many of the 'greats' - such as Alekhine, Euwe and Bogolyubov - but it then went somewhat out of fashion. Over the past few decades, many grandmasters have played it; I can name Szabo, Sosonko, Knezevic, Kouatly, Nogueiras, Arencibia, Kozul, Glek, Piskov, Moskalenko, Komarov and Zsuzsa Polgar, recently joined by Ci­fuentes, Gabriel, Topalov and Lautier. However, its present level of

Page 8: Beating King's Indian and Benoni

6 Preface

popularity does scant justice to the variation's importance. It is sig­nificant that the Austrian Attack, which embodies ideas akin to the Four Pawns, is considered to be one of the most important replies to the Pirc Defence (1 e4 d6 2 d4lbf6 3lbc3 g6 4 f4).

It was exactly thirty years ago - in 1967 - that I started playing the Four Pawns Attack. I have remained true to it ever since, and it has repaid me handsomely: in the last 15 years (1982-1996) my score with it has been +31=10-2, for a rating of over 2700! I have now decided to draw up the balance sheet and publish my analyses, including prepared novelties - a rarity in any openings book. The format I have adopted is 'subjective', so it is natural that most of the principal games should be taken from my own practice. The book is basically designed to set out a White repertoire. Nonethe­less, by examining all White's important moves, I have also pro­vided King's Indian and Modem Benoni players - who have good reason to be afraid of the Four Pawns - with enough information to find their way quickly and effectively through the myriad compli­cated variations and to prepare against it. With a few exceptions this book is based on games played before 1 January 1997.

Before proceeding with my examination of the Four Pawns At­tack, let me give King's Indian players a suggestion for a radical antidote to this attack: avoid it! This can be done as early as move one, with 1 d4 d6. Now 2 c4 is met by 2 ... e5, and 2 e4 by 2 ... lbf6, transposing to the Pirc. Black also has the following move order at his disposal: 1 d4lbf6 2 c4 d6 3 lbc3 e5, or first 3 ... lbbd7, answer­ing 4 e4 with 4 ... e5. Unfortunately, however, the knight on d7 and pawn on e5 do not always constitute the best set-up in the other King's Indian variations that may arise.

A word of advice to White: to play this opening successfully you must not only study the variations but also acquire a 'feel' for the resulting middlegame positions, for example those with a passed pawn on d5. Playing through the games in this book will help you to develop this 'feel'.

The author would be pleased to receive (c/o the publishers) any proposed analytical improvements that readers may have.

Anatoli Vaisser Paris

August 1997

Note: Bold italic has been used throughout the book to denote the author's own novelties.

Page 9: Beating King's Indian and Benoni

1 Main Line with 9 .. . Me8

Here we analyse the starting point of the main line of the Four Pawns Attack.

This position can be reached from the King's Indian Defence with the move order 1 d4 lZJf6 2 c4 g6 3 lZJc3 iLg7 4 e4 d6 5 f4 0-0 6 lZJf3 cS 7 dS e6 8 iLe2 exdS 9 cxdS, and from the Modem Benoni after 1 d4 lZJf6 2 c4 cS 3 dS e6 4 lZJc3 exdS 5 cxdS d6 6 e4 g6 7 f4 ~g7 8 lZJf3 0-0 9 ~e2.

The material in this section is arranged as follows: this chap­ter deals with 9 ... .:e8, while 9 ... ~g4 is covered in Chapter 2 and 9 ... b5 and rare moves are

dealt with in Chapter 3. The common strategic goal

for White in all this lines is the break e4-e5, either immediately or after due preparation; and sometimes involving the sacri­fice of a pawn. As a result White gets a strong passed d­pawn and an initiative, which often results in a powerful at­tack. Black for his part needs to find a way of countering White's initiative. It is not un­common for him to offer a counter-sacrifice of the ex­change in order to stabilise the position. After 9 ... .:e8 we reach the following position:

Page 10: Beating King's Indian and Benoni

8 Main Line with 9 ... l:!.e8

Now the main line 10 eS dxeS 11 fxeS ctJg4 12 j"gS 'i'b6 13 0-0 ctJxeS 14 ctJxeS is studied in Game 1 (Vaisser-Berthelot). The following deviations from the main line are also analysed: 14 d6 is considered in Game 2 (Vaisser-Degraeve) and 12 .. .f6 in Game 3 (Cranbourne-Cres­po). Less popular but interesting plans for White after 10 eS dxeS 11 fxeS ctJg4 are presented in Game 4 (Lagontrie-Boulard). Finally, the positional line 10 ctJd2 is covered in Game S (V aisser -Ibragimov).

Game 1 Vaisser-Berthelot

French Team Championship 1992

1 d4 ctJf6 2 c4 g6 3 ctJc3 i.g7 4 e4 d6 5 f4 0-0 6 ctJf3 c5 7 d5 e6 8 i.e2 exd5 9 cxd5 :Ie8

10 e5 For the positional continua­

tion 10 ctJd2 see Game S, Vais­ser-Ibragimov.

White's other possibilities are not well founded:

a) 10 'i'c2? is bad because of 1Q ... ctJxe4! 11 ctJxe4 i.fS 12 i.d3 j"xe4 13 i.xe4 'i'e7 14 ctJd2 fS IS 0-0 fxe4 16 :Ie 1 e3 17 ctJf3 e2+ Loeber-Hoerst­mann, Soest 1996.

b) 10 O-O?! ctJxe4 11 ctJxe4 :Ixe4 12 i.d3 :Ie8 13 fS ctJd7 (13 ... c4!? 14 i.c2!?) 14 ctJgS ctJeS! (Not 14 ... ctJf6?! IS fxg6,

followed by 16 'i'a4! with an initiative) IS fxg6 (IS ctJxf7? is just losing after IS ... ctJxf7 16 fxg6 ctJeS 17 'i'hS h6 18 :If7 ctJg4! 19 i.xh6 i.xh6 20 g7 'i'gS 21 i.h7+ ~xh7 22 'i'xgS i.xgS 23 :If8 i.e6 0-1 Burkart­Schlosser, Tecklenburg 1988) IS .. .fxg6 16 i.e4! (16 i.c2 h6 17 ctJe6 i.xe6 18 dxe6 d5+ Bereolos-Oestrei, USA Team tournament 1993) 16 ... :If8 17 :Ixf8+ 'i'xf8 18 ctJe6 'i'e7 19 i.gS j"f6 20 i.h6 i.d7 21 'i'b3 b6 22 :In ctJg4 and White's initiative does not compensate for the sacrificed pawn, Buck­Aldag, Niedersachsen 1995.

10 dxe5 The alternatives are all

weaker: a) 1Q ... ctJhS? 11 0-0 i.g4 12

e6!?±. b) 1Q ... ctJg4? 11 h3 (Also pos­

sible is 11 O-O!? dxeS 12 ctJgS with an initiative) 11...ctJh6 12 0-0 (12 g4!?) 12 ... ctJfS 13 ctJe4! dxeS 14 fxeS ctJd4 IS ctJxd4 :IxeS (1S ... cxd4 16 ctJd6±) 16

Page 11: Beating King's Indian and Benoni

ttJgS .:!::!:xdS 17 lDxfl ~h4 18 'i'b3! with a big advantage for White, Capit-Llanes, Ibercaja 1995.

c) 1O ... ttJfd7?! 11 exd6! (Of course, White can transpose to the classical line after 11 0-0 dxeS 12 fxeS ttJxeS 13 .lif4), when Black has tried a number of moves, but none have achieved equality:

c1) 11...a6 12 a4 tZJf6 (Or 12 ... 'i'b6 13 0-0 'i'xd6 14 ttJgS ttJf6 IS .liB ttJbd7 16 ':!::!:el .:!::!:xel+ 17 'i'xel;J; b6? 18 ttJxf7!± Rapoport-Widera, Ka­towice open 1995) 13 0-0 (White is also better after 13 ttJeS 'i'xd6?! 14 0-0 ttJbd7 IS ttJc4, Cockcroft-Planas, Spain 1983) 13 ... i.g4 14 ttJeS .lixe2 IS 'i'xe2 'i'xd6 16 'i'f3 ttJbd7 17 tZJc4 'i'b8 with an edge for White, Lehmann-Toran, Mu­nich 19S4.

c2) 11...ttJf6 12 ttJeS (12 0-0) 12 ... ttJbd7 13 a-a!? ttJb6 14 a4 'i'xd6 IS ttJbS 'i'd8 16 d6 ttJbdS 17 .lic4 i.e6 18 fS! gxfS 19 ~gS a6 20 d7! i.xd7 21 ttJxf7! with a dangerous attack, Feld­man-Shianovsky, Kiev 1986.

c3) 11...bS 12 ttJxbS! (This seems even stronger than 12 a-a!? a6 13 fS! ttJf6 14 fxg6 hxg6 IS i.f4 as in the game Z. Szabo-Kecskes, Budapest 1995) 12 ... 'i'aS+ 13 'it>f2±.

c4) 11...ttJb6 12 ttJbS i.g4 13 0-0 tZJa6 14 h3 i.fS IS ttJeS! i.xeS 16 fxeS .:!::!:xeS, Bellinger­Shaw. COIT. 1994. and now the

Main Line with 9 .. .lle8 9

simple 17 ttJc3 would have con­firmed White's advantage.

11 fxeS tZJg4 After 11... ttJfd7, besides 12

i.gS and 12 i.f4 White has the interesting additional alternative 12 e6!? fxe6 13 dxe6 ttJf6 14 'i'xd8 .:!::!:xd8 IS ~c4 with an initiative.

12 ~gS For the lines 12 a-a!? ttJxeS

13 .lif4 (or 12 i.f4 ttJxeS 13 0-0) and 12 e6 see Game 4, Lagontrie-Boulard.

12 'i'b6 The best move. The main al­

ternative 12 .. .f6 is examined in Game 3, Cranboume-Crespo. Instead, after 12 .. .'ilN as the black queen is less active than it is on b6, although having said that an eventual ttJdS from White will not win a tempo, which is important in some variations.

12 ... 'i'aS 13 0-0 ttJxeS 13 ... ttJd7? 14 e6! fxe6 IS

dxe6 ttJdeS 16 ttJxeS ttJxeS 17 e7 c4 18 'i'dS+ 'i'xdS 19 ttJxdS gives White the advantage, Kluss-Zuelke, Nuremberg 1990.

H.Ji6 White can play also 14 ttJxeS

i.xeS (14 ... .:!::!:xeS IS 'i'd2 'i'M transposes into the game Krupkova-Repkova, note c2 to Black's 14th move in this game), when IS i.c4 practically forces Black to choose between IS ... 'i'M (see line c, note to Black's ISth move in this Ilame) and IS ... ttJd7 16 d6

Page 12: Beating King's Indian and Benoni

10 Main Line with 9 .. J~e8

i.d4+ 17 ~hl lLleS 18 iLdS iLe6, where apart from 19 ctJe4 (Nei-Westerinen, Helsinki 1966), White has the simple 19 i.e7!? with an initiative. It is necessary to note that although IS i.bS?! i.d7 16 'ilNf3 is suc­cessful after 16 ... iLxbS? 17 'ilNxf7+ ~h8 18 iLf6+ iLxf6 19 'ilNxf6+ ~g8 20 'ilNf7+ ~h8 21 a4 iLd7 22 l:1ael± (Bronznik­Dzulganov, USSR 1990), it can be met by 16 ... fS 17 iLc4 'ilNb4! 18 'ilNd3 'ilNxb2 with good pros­pects for Black.

14 ... iLe6 Less precise is 14 ... ctJbc6 IS

ctJdS iLe6 16 ctJc7 ctJd7 17 ~hl h6 18 iLh4 l:1ac8 19 ctJxe6 l:1xe6 20 iLc4 l:1ce8 21 iLxe6 l:1xe6± Petursson-Thorsteins, St John open 1988.

IS ctJdS ctJbd7 IS ... c4 allows an unpleasant

exchange of the important dark­squared bishop after 16 ctJf6+, while Is ... iLxdS?! is not at all in the spirit of the position. Af­ter 16 ~xdS ctJbd7 17 ctJxeS ctJxeS 18 l:1adl White is better, Bennett -De la Rosa, Geneva open 1994.

16 iLd2!? The immediate win of the ex­

change gives Black good com­pensation, as in the game Tim­mermanns-Aepfler, Germany 1994: 16 ctJc7 c4 17 ctJxa8 l:1xa8 18 ~d2 ctJxf3+ 19 iLxf3 'ilNxd2 20 iLxd2 iLxb2 21 l:1ael iLd4+ 22 iLe3 iLxe3+ 23 l:1xe3 l:1b8 24l:1dl ctJf6.

After 16 iLd2!? Black has tried two possibilities and expe­rienced problems in both:

a) 16 ... ctJxf3+ 17 iLxf3 ~a6 18 iLc3 iLxc3 19 bxc3 iLxdS, Neb. Ristic-Los, Belgrade GMA 1988, and now 20 iLxdS! would have applied unpleasant pres­sure.

b) 16 ... 'ilNd8 17 iLc3 ctJxf3+ 18 iLxf3 l:1b8 19 ctJc7 iLxc3 20 bxc3 iLc4 21 ctJxe8 iLxfl 22 'ilNxfl 'ilNxe8 23 l:1el ctJeS 24 iLdS ~g7 2S iLxf7 'ilNxf7 26 l:ixeS with a slightly better ending, Cranbourne-Burijovich, Buenos Aires 1992.

13 0-0 After 13 ctJa4 'ilNaS+ (The

queen excursion 13 ... 'i'b4+?! 14 iLd2 'i'e4 IS ctJc3 'ilNfS favours White. The game Kabiev­Podolny, COIT. 1975, continued 16 0-0 iLxeS 17 h3 ctJf6 [17 ... iLxc3! 18 hxg4 ~e4] 18 ctJgS iLd4+ 19 ~hl 'ilNeS 20 iLf4 'i'e7 21 d6 'ilNf8 22 ctJb5 with a big advantage) 14 iLd2 'iYd8 IS iLgS (What else?) Black has a pleasant choice between repeating the position with Is .. .'~IVa5+ and playing 15 .. .f6, which is stronger here than after 12 iLgS.

However, White does have a very interesting alternative to 13 0-0:

13 'ilNd2!? This move was invented by

M. Blokh, a long-standing spe­cialist in the Four Pawns At­tack.

Page 13: Beating King's Indian and Benoni

13 ... lLlxe5 After 13 ... lLld7?! 14 d6?! is

suspicious, e.g. 14 ... ttJdxeS IS lLldS?! "iWxb2! 16 "iWxb2 lLld3+! (Not 16 ... lLlxf3+? 17 gxf3 ~xb2 18l'Ibl+-) 17 ~f1? (17 ~d2!?) 17 ... lLlxb2 18 ~b5 ~e6 19lLlc7 l'Ied8 20 lLlxa8 l'Ixa8 21 l'Ie 1 a6 22 l'Ixe6! (22 ~e2 bS+=) 22 .. .fxe6 23 .ltd7 ~f7 24 ~e2 ~f6 with a, clear edge for Black, Komarov -Lamoureux, Cannes 1993. It is better to play 14 e6! fxe6 (14 ... lLldeS IS ~bS!? needs further tests) 15 dxe6 "iWxe6 (1S ... ~xc3?! 16 bxc3 "iW xe6 17 O-O! ctJdf6 18 l'Iae 1 "iWc6 19 h3! ctJe4 20 "iWf4 h6 21 hxg4 hxg5 22 ctJxg5 ctJxg5 23 "iWxg5 ~g7 24 ~c4! l'Ixe1 2S l'Ixel "iWf6 26 l'Ie7+ 1-0 Blokh­Feldman, USSR 1982) 16 ctJdS !? ~eS 17 ctJxe5 "iW xe5 18 ~f4 'iVe4 19 ctJc7, when White is clearly better (Blokh).

140-0-0

This is the point of Blokh's idea. The king moves out of immediate danger and the b2-pawn is protected at the same time,

Main Line with 9 .. . ':e8 11

14 . .,ctJa6 Black has also tried a number

of other moves: a) 14 ... c4?! IS ctJxeS ~xeS

16 ~xc4 "iWcs 17 ..tb3 ctJa6 18 l'Ihf1! ~g4 19 d6! l'If8 20 d7! ~xdl 21 l'Ixdl± Blokh-Lelu­aschvili, corr. 1989/90.

b) 14 ... .ltd7?! 15 ctJxe5 .ixeS 16 l'Ihel ctJa6 17 d6! "iWa5?! 18 ~c4 ctJb4, Bellinger-Ardin, corr. 1994, and now 19 ~f2! gives White a clear advantage.

c) 14 ... 'i'aS, when instead of lSl'Ihel?! ctJbd7 16 d6 ctJb6! 17 .ltbS .ltd7 18 ctJxeS 'uxeS 19 l'IxeS ~xeS 20 .ltxd7 ctJxd7+ Blokh-Mitenkov, USSR 1980, White should play first IS lLlxeS!? (Blokh).

d) 14 ... ~fS IS ctJxe5 .ltxe5 16 l'Ihe I! 'i'b4 ! (otherwise 17 g4!) 17 a3! 'i'b3 18 g4 ..td7 19 ..tf3 and White's initiative is worth more than a pawn, Blokh­Schneider, corr. 1992.

15 ctJxe5 ':'xeS It is too dangerous to play

15 ... .ltxe5?!, e.g. 16 ~b5! l:lf8 (16 ... .ltxc3? fails to 17 'i'xc3 'i'xbS 18 .lth6 f6 19 ~xf6 "iWc4+ 20 ~b1 "iWe4+ 21 ~a1 'i'e7 22 :hel!) 17 :hel (17 ~e7!?) with strong pressure for White, for example 17 .. .f6 18 d6! ~g4 19l'Ixe5!±.

16 ':'hel 16 ..tf4?! is premature. After

16 ... l'Ie817l'Ihe1?! ..td7 18 d6 .lte6 19 ..txa6 'i'xa6 20 ~bl ..td4 Black is clearly better, Blokh-Lukin, Russia 1992.

Page 14: Beating King's Indian and Benoni

12 Main Line with 9 ... 1:1e8

16. ctJb4?! In the event of 16 ... .1i.d7 17

d6 l:lae8, the move 18 .1i.f4 is much stronger than it was two moves ago. After 18 .. J:iSe6 19 .1i.c4 ~xel 20 ~xel llxel+ 21 'iVxel White keeps the initiative despite the exchange of rooks (21. .. Vib4? 22 iLxf7 + !).

17 .1i.c4! ~xel 18 ~xel! J.fS 19 d6!?

All White's pieces are well placed and his chances are pref­erable, Blokh-Sergeev, USSR 1983.

13 ctJxe5 Apart from 13 ... c4+, Black's

alternatives are unattractive: a) 13 ... 'iVxb2?! 14 ctJbS!

ctJxeS IS ~bl with advantage to White.

b) 13 ... iLfS?! 14 ctJa4! 'iVaS IS ctJh4! ctJxeS (1S ... .1i.xeS? 16 iLxg4 iLxg4 17 'iVxg4 iLd4+ 18 ~hl 'iVxa4 19 ctJfS! gxfS 20 ~xfS hS 21 'iVg3+-, Semkov) 16 ctJxfS gxfS 17 llxfS ctJbd7 18 ctJc3± Semkov-Apicella, Bul­garia-France 1990.

c) 13 ... h6?! 14 ctJa4 (14 iLf4!? gS IS iLel) 14 .. ,VWc7 IS d6 'iVc6 16 .1i.e7 .1i.e6, Kak­ageldyev-Lerner, Riga 1972, and now 17 ~c I! would have been in White's favour.

d) 13 ... ctJd7 14 ctJa4!? ViaS IS iLd2 'iVd8 16 e6! fxe6 17 ctJgS iLd4+ 18 ~hl ctJxh2 19 ~xh2 exdS with complicated play in which White's chances must be better, Dittmar-Krug, Wiesbaden open 1990.

e) 13 ... c4+'? 14 Whl Now neither 14 ... ctJf2+? IS

~xf2 'iVxf2 16 ctJe4 'iVb6 17 ctJd6 ~f8 18 iLe7 ctJd7 19 ctJxc4+- nor 14 ... h6?! IS iLf4!? (IS e6 fxe6 16 'iVc2 eS 17 d6 .1i.fS 18 'iVa4 ctJc6 19 ctJdS 'iVxb2 20 d7 is unclear, Rojo­Sanz Alonso, Linares 1990) IS ... gS 16 iLel± can be recom­mended, but instead

14 ... ctJd7!? is an interesting idea of

Murey's which deserves close attention. Black opts for quick development instead of taking the pawn. White now has a choice of three moves:

el) IS d6?! is not very promising: IS ... ctJf2+! 16 ~xf2 'iVxf2 17 ctJdS?! (17 iLxc4?! ctJxeS 18 ctJe4 Vib6 19 ctJxeS ~xeS 20 ctJf6+ .1i.xf6 21 .1i.xf6 ~fS 22 .1i.e7 .1i.d7 and White has insufficient compensation for the exchange, Kouatly-Povah, Ramsgate 1979; 17 iLh4!?) 17 ... ctJxeS 18 ctJe7+ (It is too late to play 18 iLh4? due to 18 ... ctJg4! 19 ctJe7+ ~f8 20 iLxc4 'iVxh4!+ Peev-Trapl, De­cin 1978) 18 ... ~h8 19 ctJxeS iLxeS 20 ctJxc8 ~exc8 21 .1i.f3 ~d8!+ Pastor-Lybin, Frydek Mistek open 1996.

e2) IS e6 Many of the following com­

ments are based on Murey's previously unpublished analy­SIS.

IS ... fxe6 IS ... ctJdeS? is iust bad: 16

Page 15: Beating King's Indian and Benoni

tbxe5 tbxe5 17 'iWa41±. 16 dxe6 tbdf6 16 ... tbde5? 17 tbxe5 tbxe5

18 tbd5 'iWxb2 19 tbf6+± and 16 ... 'iWxe6? 17 tbd4! 'iYd6 (17 ... ~xd4? 18 .txg4 'iWd6 19 tbb5 'iWc5 20 tbxd4 ~xg5 21 tbe6 'iYe5 22 'iWf3 ~h8 23 'ilNf7 I:lg8 24 I:lael 'iWxb2 25 tbg5+-) 18 .txc4+ 'it>h8 19 'iWxg4 tbb6 (19 ... 'iYxd4 20 'iWxd4 ~xd4 21 I:lael±) 20 tbdb5! 'iWc6 21 'iYg3 tbxc4 22 tbc7± .te5? 23 tbxe8.

And now: e21) 17 tbd4? ~xe6 18 h3

I:lad8=F. e22) 17 e7? I:lxe7 18 tbd5

tbxd5 19 'iYxd5+ i.e6 20 'iYe4 I:lc7! 21 I:ladl I:lf8 22 h3 tbf6 23 'iYM i.d5=F Jones-Povah, Chester 1979.

e23) 17 .txc4?! .txe6 18 ~xe6+ I:lxe6 19 'iY d4 tbe4 20 'iYxb6 axb6 21 tbxe4 :xe4 22 h3 was approximately equal in Vaisser-Murey, Paris 1990.

e24) 17 h3! 'iYxb2! Both 17 .. tbf2+? 18 I:lxf2

~ xf2 19 .txf6 .txf6 20 tbe4+­and 17 ... tbe3? 18 .txe3 ~xe3 19 ~xc4 ~xe6 20 I:le 1 'ilNxe 1 +

Main Line with 9 ... ::'e8 13

21 'ii'xel .txc4 22 'ii'M± are bad for Black.

After 17 ... 'ii'xb2 White has a choice:

e241) 18 I:lc1? tbh5! 19 hxg4 (19 tbd4? tbg3+ 20 'it>gl ~xc3!-+; 19 tbe4 tbgf6 20 tbxf6+ .txf6=F) 19 ... tbg3+ 20 'it>h2 tbxfl + 21 .txfl .txc3=F.

e242) 18 tba4? meets with a fantastic refutation: 18 ... tbf2+!! 19 J::rxf2 tbe4! 20 J::rfl tbg3+ 21 ~gl ~xal 22 'iWxal tbxe2+ 23 ~f2 .txal 24 I:lxal .txe6! 25 I:lel I:lac8 26 I:lxe2 c3-+ Kak­ageldiev-Murey, COIT. 1972.

e243) Taimanov's proposi­tion 18 'iWel!? has never been tested.

e244) 18 'iWc1! 'iWxc1 19 I:laxc1 h6

After Nunn's suggestion 19 ... tbh5 20 tbe4 (Murey gives 20 hxg4 tbg3+ 21 'it>gl! .txc3 22 .txc4! .txe6 [Not 22 ... tbe4? 23 I:lxc3!; 22 ... ~b2?! 23 I:lfel!] 23 J::rxc3 tbxfl 24 'it>xfl .txc4 25 I:lxc4 I:lac8, when the re­sulting ending is about equal) 20 ... .txe6!? 21 hxg4 i.xg4, White can continue with 22 ~xc4+ 'it>h8 23 tbd6 tbg3+ 24 'it>g 1 with an advantage.

20hxg4! 20 i.d2?! favours Black after

20 ... tbh5 21 hxg4 tbg3+ 22 ~h2 tbxfl + 23 .txfl .txe6 24 tbb5 I:lec8! 25 tbd6 c3!

20 ... hxg5 21 tbxg5 21 tbb5? is poor: 21...~xe6

22 tbc7 tbe4! 23 tbxe8 I:lxe8 24 .txc4 tbg3+ 25 'it>g 1 ~xc4 26

Page 16: Beating King's Indian and Benoni

14 Main Line with 9 ... l:i.e8

~xc4 ttJxfl 27 ~xfl iH6=t. 21.. . ttJh7! 22 ttJce4 After 22 ttJxh7? ~xh7 23

ttJb5! ~xe6 (23 ... .txe6?! 24 ttJc7 c3 25 iLb5±) 24 ..txc4 .l:i.e4 White is not better.

22 ... ttJxg5 23 ttJxg5 ..th6 24 ,Uf6! ~g7 25 'uf7+ ~g8 26 ,Uf6

with a draw by repetition of moves (Murey).

e3) 15 Ji.xc4! The only move that promises anything for White, e.g. 15 ... ttJdxe5 16 ttJxe5 ttJxe5 (l6 .. .txe5 17 d6!) 17 ..tb3;i;:; Vaisser-Murey, Evry (rapidplay) 1993.

14 ttJxe5 14 ~d2 is a little investigated

alternative, while 14 d6 is con­sidered in Game 2, Vaisser­Degraeve.

14 ..txe5 14 .. Jhe5 is rather unpopular,

but not so stupid after 15 ~ d2 .i.f5 and now:

a) 16 g4?! is too aggressive: 16 ... c4+ 17 ~f2 ..td3 18 ..txd3 cxd3 19 ~f4 f5 20 ~dl ttJd7 21 .l:i.xd3 ~f8 with an initiative for

Black, Krupkova-Grabics, Zan­ka 1995.

b) Interesting, but not suffi­cient for equality, is the imme­diate exchange sacrifice after 16 iLf4 ttJd7?! 17 .i.xe5 ttJxe5 as in the game C. Hansen-Yurtaev, Gausdal1990.

c) 16 iLc4, after which Black has tried:

cl) 16 ... ttJd7 17 d6 ~b4? (l7 ... .te6! 18 ttJd5? ~xd5! 19 ..txd5 .td4+-+) 18 ~f4! h6 19 iLe7 ~a5, Vaisser-Gufeld, No­vosibirsk 1971, and now the simple development 20 ~afl would have assured White of an advantage.

c2) 16 ... ~b4 17 ~f4 h6! 18 iLd8 b6 was unclear in Krup­kova-Repkova, Prague 1996.

15 ..tc4 15 ~d2!? is not played often,

but deserves further analysis. One recent example: 15 ... ttJd7 16 ~adl a6 17 ~hl .i.d4 18 ~f4 f5, Thoma-Kahn, Budapest 1995, and now 19 d6! would have given White a strong at­tack.

15 ..tb5?! is an old and un­successful attempt to meet the Black set-up. After 15 ... .i.d7 (This position also can arise after 15 .tc4 .tf5 16 .tb5 ..td7) 16 ~f3 both 16 .. .f5 17 .i.c4 ~xb2 18 d6+ ~h8 19 ~ac1 ~c6 20 ~h3 ttJd7 21 ttJe2 .i.xd6 22 ~cdl .te5, Nei­Ciocaltea, Zinnowitz 1966, and 16 .. .f6 17 a4 ~xb5 18 ttJxb5 ttJd7 19 d6 'ilVc6, Kakageldiev-

Page 17: Beating King's Indian and Benoni

Zaid, USSR 1973, give Black an obvious advantage.

15 'i'xb2 This is one of the critical po­

sitions in the 9 .. J:te8 system. Apart from the text move, Black also has a choice between:

a) 15 ... 'i'd6? 16 lbb5 i.xh2+ 17 Whl 'i'g3 18 i.f4 'i'h4 19 ..txh2 'i'xc4 20 lbd6 1-0 Serdt­Podgornik, Bled 1992.

b) 15 ... lbd7? 16 d6 i.d4+ 17 Whl lbe5 18 lbd5 'i'xd6 19 lbf6+ Wf8 20 i.h6+ We7 21 lbd5+ Wd7 22 'i'a4+ b5 23 ..txb5+ 1-0 Dobos-Petersen, Aarhus 1991.

c) 15 ... 'i'b4 The theory of this line has an

interesting history. Initially it was considered good for Black on the basis of the game Janosevic-Forintos, Vrnjacka Banja 1973: J6 'i'b3?! i.f5 17 d6 'i'xb3 18 axb3 i.xd6 19 lbd5 lbd7 20 l:ixf5! gxf5 21 i.b5 i.e5 22 i.xd7 l:ied8 23 i.xd8 lixd8 24 lbe7+ Wf8 25 i.xf5 rJi;xe7 26 l:ixa7, although

Main Line with 9 .. J~e8 15

White succeeded in achieving a draw in the end. Then the eITor­filled game Szabo-Pietzsch, Salgotarjan 1967: 16 'i'f3 i.f5 17 g4?? 'i'xc4?? 18 gxf5 prom­ised a clear advantage to White. Sometime later an evident im­provement for Black was dis­covered: 17 ... 'i'xb2! 18 lbe2 i.d4+! 19 lbxd4 'iYxd4++. Fi­nally, 16 'i'f3 i.f5 17 i.b5 came into practice and again the evaluation of the line was changed, as we can see:

16 'iYf3 i.fS White's attack is too strong

after 16 .. .f5?! 17 'i'd3 ~g7 (l7 ... iYxb2 18 d6+ Wg7 19lbb5 i.d4+ 20 Whl iYxal 21 l:ixal i.xal 22 'iYbl!±) 18 l:iael i.d7 19 i.d2 'iWb6 20 l:ixe5!, Strat­ing-De Bruijne, Leeuwarden 1995.

17 i.b5 White has another try that is

worth further analysis: 17 i.b3 'iWg4 18 "i'xg4 .txg4 19 d6 i.e6 20 lbd5, Bellinger-Burd, COIT . 1994.

17 .. .1::£8 18l:iae1 Here White can also consider

18 i.e7!? i.g4! 19 'i'd3 'i'd4+ 20 'i'xd4 .txd4+ 21 rJi;hl ':c8 22 h3 i.d7 23 .tc4 with good compensation for the pawn as in the game Vaisser-Degraeve, Cannes 1990.

ll..J6 The knight move 18 ... 4.Jd7?

is bad: 19 i.xd7 .td4+ 20 rJi;hl ..txd7 21 l:ie7 .tf5 22 g4 "i'xb2 23 lbdl± Vegh-Pesztericz,

Page 18: Beating King's Indian and Benoni

16 Main Line with 9 ... .l:::.e8

Hungarian Team Ch. 1994, while White has strong pressure after 18 ... i.xc3?! 19 'i'xc3 'i'xc3 (19 ... 'i'xb5? 20 .th6) 20 bxc3 a6 21 i.e8! or 18 ... i.g4!? 19'i'd3 'i'd4+ 20 'it>hl. Perhaps Black should consider the very sharp 18 ... i.d4+ 19 'it>hl 'i'xb2.

19'iJ.xe5! This is stronger than 19 a3?!

i.d4+ 20 'it>hl 'i'xb2 21 i.h6 'i'xc3 22 i.xf8 'i'xf3 23 l::I.xf3 'it>xf8 24 l::I.e8+ 'it>f7 25 d6 a6 26 l::I.e7+ 'it>f8 27 i.c4 tZJc6 28 l:i'.xh7, Matsula-Kharitonov, Ki­rov 1993, as now 28 ... l:i'.d8! would have given Black an ad­vantage.

19 ... fxe5 20 i.h6l:i'.f7 21'i'g3 a6 22 i.e8! :e7 23 iLxg6! hxg6 24 :xf5

with a winning attack. This variation is not forced but it is rather convincing.

d) 15 ... i.f5!? 16 tZJb5! 16 i.b5?! i.d7 transposes

into the inferior line 15 iLb5?! given above.

Here Black has a wide choice:

dl) 16 ... l::I.f8?! 17 a4 'i'a5 18

i.e7 tZJd7, Sakovich-Didishko, USSR 1973, and now 19 d6!? would have promised an ad­vantage.

d2) 16 ... h6 17 l:i'.xf5!? (If this is found not to work, White can follow the game Hausner­Tobyas, Czech Ch. 1992, with 17 i.xh6 'i'f6 18 'i'd2 a6 19 tZJc3 b5 20 g4, where the posi­tion was not so clear) 17 ... gxf5 18 d6! hxg5 19'i'h5 l:i'.f8 20 l:i'.f1 with a very strong attack.

d3) 16 ... tZJd7 17 a4! (It is too early to play 17 d6? .te6! 18 'i'b3 .txd6 19 tZJxd6 'i'xd6+) and now:

d31) 17 ... 'i' a5? 18 d6 i.e6 19 .txe6l:i'.xe6 20 'i'f3 f6 21 'i'xb7 tZJb6 22 ~th6 f5 23 l:i'.ae 1 1-0 Vaisser-Jojic, Paris 1990.

d32) 17 ... a5?! 18 l::I.a3! .td4+ 19 'it>hl tZJe5 20 l::I.b3! i.g4 21 'i'c1 'i'a6! 22 h3 is better for White, Vaisser-Podvrsnik, Ptuj 1989.

d33) Black has at his disposal an interesting queen sacrifice: 17 ... a6!? 18 a5 axb5 19 axb6 bxc4 which awaits its first test.

d34) 17 ... f6 18 a5 'i'd8 19 iLf4!? (The white bishop seems more safely placed on f4 than on h6. The game Vaisser­Akopian, USSR 1988, saw in­stead 19 i.h6 'it>h8 20 d6 i.xb2?! [It would have been better to play 20 ... a6! 21 tZJc7 .td4+ 22 'it>hl tZJe5 with com­pensation] 21 l::I.a2 tZJe5 22 i.e2 i.d4+ 23 tZJxd4 cxd4 24 'i'xd4 l:i'.e6, and now 25 l:i'.d2 would

Page 19: Beating King's Indian and Benoni

have promised an advantage to White, e.g. 25 ... lZJc6 26 'iYb2! lZJxa5 27 g4 .te4 28 d7!) 19 ... ~h8 20 d6 a6 21 Cjjc7 with complex play in which White's chances must be better, Strat­ing-Nijboer, Amsterdam open 1995.

d4) 16 ... a6! This appears to give Black a

forced draw. 17 d6 axb5 In the case of 17 ... Cjjd7 18

lZJc7 .i.d4+ 19 ~h1 'iYxd6 20 lZJxa8 ~xa8 21 iYb3!? Black does not have enough compen­sation for the exchange.

18 .txf7 + ~xf7 18 ... ~g7? 19 .i.xe8 SLd4+ 20

~hl 'iYxd6 21 i..xb5± Vaisser­Kozlov, USSR 1971.

19 ~xf5+ gxf5 Again Black must accept the

sacrifice. 19 ... ~g7? fails due to 20 d7 lZJxd7 21 'iYxd7+ ~h8 (21...~g8 22 iYf7+ ~h8 23 ~xe5!±) 22 ~f7 1:.f8, Van der Doef-Buisman, corr. 1974, and now White could have won with 23 1:.afl! SLg7 24 ~xf8+ ~xf8 25 ~xf8+ SLxf8 26 iYf7.

20 'iYh5+ ~f8 Spasov, as Black against

Wessman, Tunja 1989, went wrong here with 20 ... ~e6? 21 'iHxe8+ ~xd6 22 ~e1! c4+ 23 SLe3 lZJc6 24 'iYxa8 SLd4 25 ~dl ~c7, and could have lost the game after 26 ~xd4! lZJxd4 27 'iHf8.

After 20 ... ~g7 White has a perpetual check with 21 SLh6+

Main Line with 9 ... .l::!.e8 17

~f6 22 .tg5+. Since 21 ..th6+ ~f6 22 ~fl? loses because of 22 ... ..td4+! (not 22 ... .txh2+? 23 ~h1 ~e5 24 i..g5+ ~e6 25 'iYe8+ ~d5 26 ~dl+ mating, Strating -Bloemhard, Holland 1992) 23 ~hl ~e5, the only try for a win is 21 'iYxe8; but this is less promising than in the game Wessman-Spasov.

I analysed the critical posi­tion after 20 ... ~f8 in the local press in the annotations to my game against Kozlov, and con­cluded that White has nothing more than a draw by perpetual check: 21 i..h6+ .tg7 22 ~f1 c4+ 23 ~h1 'iHf2! 24 ..txg7+ ~xg7 25 iYg5+ ~f7 26 'iYh5+. When Yudovich then published this analysis in In/ormator this sharp position became the sub­ject of much theoretical debate. Pukshansky claimed that White could win by 21 ..th6+ i..g7 22 .txg7+ ~xg7 23 'iYxe8 c4+ 24 ~hl iYxd6 (Or 24 ... iYf2? 25 d7 ~xa2 26 ~el lZJc6 27 d8'iH Cjjxd8 28 ~e7+ mating) 25 ~el ~a6(?) 26 ~e7+ ~h6 27 iYf8+ ~g5 28 'iHg8+! winning, and

Page 20: Beating King's Indian and Benoni

18 Main Line with 9 ... .l::i.e8

over a period of some fifteen years his analysis was passed from book to book. The power of White's attack was then con­firmed by the game Reinemer­Hoerstmann, Wittlich 1990: 25 ... ~f6? (Instead of 25 ... lIa6?) 26 h4! lilxa2 27 'iYhS+ ~g6 2S 'i'gS+ ~f6 29 g4! f4? 30 'i'hS+ ~g6 31 'i'eS+ ~g7 321Ie7+±. I also believed all this analysis without sufficient checking un­til I played a little-known game against Bauer (French Team Cup 1992). My young and rela­tively inexperienced opponent knew nothing about this varia­tion but found the only defence: 21 ~h6+ ~g7 22 ~xg7+ ~xg7 23 'i'xeS c4+ 24 ~hl 'i'xd6 25 lIe 1 WIIf8! Now White has nothing more than a perpetual check after 26 'iVe5+ (26 'i'xb5 tzJa6) 26 ... ~gS 27 'iYd5+ ~hS.

Other options do not give White more than a draw:

d41) 21 ii.h6+ ii.g7 22 'iVxf5+ ~gS 23 lilfl? (23 'iVd5+ ~hS 24 ~xg7+=) 23 ... ~d4+ (Thanks to this check Black can beat off the white attack and keep a great material advan­tage) 24 ~hl tzJa6 25 d7 lIedS.

d42) 21 lIfl? c4+ (Not 21...tzJd7? 22 'iVxh7 lile6? 23 lIxf5+ ~f6 24 ii.h6+ ~eS 25 'iVgS+ tzJfS 26 'i'xfS+ ~d7 27 'iYf7++-) 22 ~hl 'iYxd6 23 'i'xh7 lIe6 24 'i'xf5+ ii.f6 25 i.xf6 lIxf6 winning.

d43) 21 'iYxh7 'i'xd6 22 i.h6+ 'iVxh6 23 'iYxh6+ i.g7

(Possible is 23 ... ~e7!?) 24 'iYd6+ ~gS 25 'i'd5+ ~hS 26 'i'xf5 tzJc6 27 'i'h5+ with per­petual check, as in Vaisser­Apicella, French Ch. 1996.

So we can conclude that at the moment the line 15 ... ~f5 16 tzJb5 leads to a more or less forced draw.

16 d6

16 1If8! This natural move was first

played in the game Gorovaya­Kviatkowskaja, COIT. 1970. Dur­ing the course of the following 20 years, however, it was for­gotten and only in 1990 did Marin bring it back into prac­tice. Thanks to its re-emergence I was able to check my 1969(!) vintage analysis in the present game. Others are much weaker:

a) 16 ... i.f5? 17 lIxf5! (In B. Vladimirov-Doda, Leningrad 1967, White inverted his moves, 17 i.xf7+? ~xf7 ISlIxf5+, and could have been punished with IS ... ~g7!) 17 ... gxf5 IS i.xf7+! ~fS (After lS ... ~xf7? 19 'iYh5+

Page 21: Beating King's Indian and Benoni

<;t>f8 the quickest win was dem­onstrated in Vaisser-Jaikovsky, Moscow 1967, which up until now had followed Vladimirov­Doda: 20 ~h6+! !1Lg7 21 ~f1!) 19 ~xe8 'iVxc3 20 ~e7+ <;t>xe8 21 'iVh5+ <;t>d7 22 'iVxf5+ <;t>c6 23 'iVe4+ <;t>d7. For a long time this position was considered to be drawn. In reality, instead of taking the perpetual check White can play 24 ~d1! and Black is helpless. For example: 24 ... ttJc6 25 'iVf5+ <;t>e8 26 'iVe6 or 24 ... 'iVb2 25 'iVa4+! b5 26 "WI g4+ winning.

b) 16 ... !1Le6? 17 ~xe6 fxe6 18 'iVf3 ttJd7 19 'iVf7+ <;t>h8 20 "WIxd7 'iVxc3 21 ~adl and the passed d-pawn assures White of a clear advantage.

c) 16 ... !1Ld4+?, after which I caused unnecessary problems for myself by playing for beauty in the game Vaisser-Zagorov­sky, USSR 1976: 17 'iVxd4? cxd4 18 ~xf7+ <;t>g7 19 ~xe8 dxc3! with an unclear position. Instead, the simple 17 ~hl gives White a clear advantage:

c1) 17 ... iH5 18 ~xf5! gxf5 19 ~xf7+ ~xf7 (l9 ... ~f8 20 ~xe8 "WIxc3 21 ~c1 +-) 20 "WIh5+ <;t>f8 21 ~h6+ !1Lg7 22 "WIxf5+ <;t>g8 23 ~f1 1-0 Arenci­bia-Martin del Campo, Bayamo 1989.

c2) 17 ... ~f8 18 ttJb5! 'iVxal 19 'iVf3 'iVb2 20 ttJxd4 'iVxd4 21 !1Lxf7+ <;t>g7 22 ~f6++-.

17 ~xf7+ White's other choices are:

Main Line with 9 ... .:i.e8 19

a) 17 ~xf7? is just plain bad: 17 ... ~xf7 18 ~xf7+ <;t>g7-+.

b) The aforementioned game Gorovaya-K viatkowskaja con­tinued 17 ttJb5? "iYxal 18 "iYf3 ~b2 (We see a big difference here compared to note c2 on the previous move, where Black's bishop was placed on d4) 19 !1Lxf7+ (19 !1Le7!?) and now the quickest way for Black to win was 19 ... <;t>h8! 20 !1Le7 !1Lg4! 21 'iVxg4 'iVxb5 22 'iVe4 ~xf1 +.

c) 17 ~c1!? seems less clear. At least the variation proposed by Konikowski can be im­proved: 17 ... !1Lxc3 18 !1Le7 ~d4+ 19 <;t>hl ttJd7 20 !1Lxf8 ttJxf8 and now instead of 21 ~xf7 !1Le6 22 ~xf8+ ~xf8 23 !1Lxe6+ 'it'h8+ (Konikowski) White should play 21 'fJ.b1! ~c3 22 !1Lxf7+ 'it'g7 23 ~b3 ~a5 24 ~e2! with an advantage.

17 'it'g7 Black can also play 17 ... ~xf7

18 ~d5 !1Ld4+ 19 'it'hl !1Lf5 20 ~abl !1Lxc3! (Not 20 ... ~xc3? 21 .:i.xb7 ttJd7 22 ~xd7 ~af8 23 ~xf7 ~xf7 24 d7+-) 21 ~xb2 !1Lxb2 22 g4 !1Le4+ 23 ~xe4 ~xf1+ 24 ~g2 ~f7 25 ~e8+ ~f8 26 'iVb5 !1Ld4 27 ~xb7 ttJd7 28 ~xd7 ~f2+ 29 'it'g3 ~af8 30 'iVe6+ 'it'h8! 31 ~d5 with a probable draw.

18 ~dS! After 18 ~d5? ~xc3 19

~ad 1 !1Lf5 20 ~f3 !1Ld4+ 21 'it'hl 'iVxf3! 22 gxf3 ~xf7 Black is clearly better, Semkov-Marin, Berga 1990.

Page 22: Beating King's Indian and Benoni

20 Main Line with 9 ... J:.e8

18 'i'xc3 White has an advantage after

both: a) 18 .. ..l:hfl+?! 19 'i'xfl i.f5

20 .l:.e1! 'i'xc3 21 i.xb7 .txd6 22 .txa8; and

b) 18 ... .tf5?! 19 'i'c1! 'i'xc3 (l9 ... 'i'xc1 20 1:i.axc1 lLlc6 21 lLlb5;!;) 20 .th6+ ~h8 21 'i'xc3 i.xc3 22 .txf8 lLld7 23 i.e7 i.xal 24 .l:.xal; but

c) J8 ... lLld7! could have posed some problems. White can probably hold a draw, but not more after 19 .l:.xf8! lLlxf8 20 'i' c1! 'i'xc3 21 i.h6+ ~h8 22 'i'xc3 i.xc3 23 .l:.fl i.d4+ 24 ~hl i.f5 (24 ... lLle6 25 .l:.f7) 25 g4 .txg4 26 i.xf8.

19 I!xf8 ~xf8 Not 19 ... 'i'd4+? 20 .l:.f2±.

20 'i'f1+ i.f5 21 .l:.dl!

This is the point. Black has no comfortable defence to the twin threats of 22 i.xb7 and 22 g4. If now 21. .. 'i'b4 then 22 a3!

21 lLld7 22 g4 ~g7

23 gxf5 'i' c2? 23 ... h6! was necessary (If

23 ... .l:.f8?! then 24 i.e6!±). For example, 24 i.e7!? (Or 24 .txb7 .l:.b8 25 i.d2 'i'd4+ 26 ~hl 'i'h4 27 'i'g2 .l:.xb7 28 'i'xg6+ ~h8 with perpetual check, Kahn-Gladyszev, Buda­pest 1996) 24 ... 'iYc2 25 'i'f3 'iYxh2+ 26 ~fl gxf5 27 'iYxf5 'i'f4+ 28 'iYxf4 i.xf4 29 i.xb7 .l:.b8 30 i.c6 lLle5 31 i.e4 i.g5 32 .l:.d5 lLlf7 with equality, EI­bilia-Berthelot, French Team Ch.1993.

24 .l:.d2 'iYxf5 25 'iYxf5 gxf5 26 i.e7 i.d4+ 27 ~f1 lLle5

27 ... i.f6 doesn't work be­cause of 28 .l:.g2+ ~h8 29 i.e6 .txe7 30 dxe7 lLlf6 31 i.f7 and Black is helpless.

28 .l:.g2+ lLlg6 29 h4 ~h8

29 ... i.f6 also loses: 30 h5 i.xe7 31 dxe7 .l:.e8 32 hxg6 hxg6 (32 ... .l:.xe7 33 gxh7+ ~xh7 34 .l:.g5+-) 33 .l:.e2 ~f6 34 .l:.e6+ ~f7 35 .l:.b6+ ~xe7 36 .l:.xb7+ ~d6 37 .tf7. More re­silient, however, was 29 ... ~h6! 30 .tg5+ ~g7 31 h5 lLle5 32 .l:.e2, although White still has a big advantage.

30 .l:.xg6 hxg6 31 .txb7 .l:.b8 32 d7 33 d8'iY+ 34 .txd8 35 .tg5 36 ~e2

c4 .l:.xd8 ~g7 c3 1-0

Page 23: Beating King's Indian and Benoni

Game 2 Vaisser-Degraeve

Cappelle fa Grande 1987

1 d4 lbf6 2 c4 g6 3 lbc3 J..g7 4 e4 d6 S f4 0-0 6 lbf3 cS 7 dS e6 8 ~e2 exdS 9 cxdS ~e8 10 eS dxeS 11 fxeS lbg4 12 J..gS 'ii'b6 13 0-0 lbxeS

14 d6

Over the years the position after 14 d6 has been of great importance to the theory of the Four Pawns Attack. Currently, however, it is perceived as less attractive for White than 14 lbxeS (see the previous game), because Black has a choice between a practically forced draw in the line 14 ... 'ii'xb2 and complicated play with mutual chances after either 14 ... c4+ IS ~hl 'ii'xb2 or 14 ... J..fS.

14 ~fS!? Black has a wide choice here: a) 14 ... ~e6?! IS lbxeS!

~xeS (lS ... 'ii'xb2?! 16 'ii'a4! j;;ffR 17 j;;fJr 1+ Sipqmnnn-J1Jnl!e

Main Line with 9 ... J:.e8 21

corr. 1984; if now 17 ... J..xeS, 18 ~abl traps Black's queen) 16 lbdS ~d4+ (Black has nothing better: 16 ... J..xdS? 17 'ii'xdS .1i.d4+ 18 ~hl ~f8 19 ~xf7!+- or 16 ... 'ii'xb2?! 17 lbf6+ iLxf6 18 iLxf6 'ii'b4 19 iLf3 tLJd7 20 ~b 1 'ii' as 21 iLa 1 and White's attack is worth much more then two pawns, Kaplun-Vicin, Odessa 1980) 17 'ii'xd4 cxd4 18 tLJxb6 axb6 19 ~bS tLJc6, Peev-Vogt, Varna 1973, when White would have had a clearly better ending after 20 iLf6!?

b) 14 ... tLJbd7?! IS iLbS!? ~e6 (Not IS ... c4+? 16 ~hl tLJd3 17 iLxc4 tLJf2 + 18 ~xf2 'ii' xf2 19 ~h4! 'ii'xb2 20 ~xf7+!, but IS ... a6!? deserves attention) 16 iLe7 tLJxf3+ 17 'ii'xf3 ~d4+ 18 ~hl tLJeS 19 'ii'f4 ~xc3 20 'ii'h6 tLJd7? (20 ... ~d7 21 ~xd7 tLJxd7 22 bxc3 f6 was neces­sary, although White still has a strong initiative) 21 ~xf7! mating, Kouatly-Raupp, Berlin 1976.

c) 14 ... tLJxf3+ IS iLxf3 iLd4+?! (lS ... 'ii'xb2 16 tLJdS transposes to line d below and IS ... c4+ 16 ~hl 'ii'xb2 to line e below) 16 ~h 1 'ii'xd6? 17 .1i.dS! (Also possible is 17 tLJbS 'ii'b6 18 tLJxd4 cxd4 19 'ii'd2 J..fS 20 ~adl tLJc6 21 ~dS with strong pressure, Arencibia-Gomez, Cu­ban Ch. 1988) 17 ... ~e6 18 ~xb7 tLJd7 19 ~xa8 ~xa8 20 lbbS. White is clearly better ~ftpr h0th 'W, ,~rl" 21 01xrl4

Page 24: Beating King's Indian and Benoni

22 Main Line with 9 .. J:te8

cxd4 22 'ilNf3!, as in the game Kouatly-Teichmann, London 1978, and 20 ... 'ilNb6 21 ltJxd4 cxd4 22 b3 .1l.f5 23 'ilNf3!, Kouatly-Perdigo, Ales 1979.

d) 14 ... 'ilNxb2 15ltJd5ltJxf3+ Black has also tried

15 ... ltJbc6?! (15 ... ltJec6?! is rather artificial, since 16 ..tc4 gives White good chances, while 15 ... ..tf5 transposes to the main game) 16 .l:.bl (16 ltJf6+!?; 16 ..tf6!?) 16 ... 'ilNxa2 (l6 ... ltJxf3+? 17 ..txf3 is bad for Black. Now both 17 ... 'ilNd4+ 18 c;t>hl .l:.b8 19 ltJc7 ..td7 20 ltJxe8 .1l.xe8 21 ..td5, Semkov­Spassov, Bulgarian Ch. 1988, and 17 ... 'iIN e5 18 ltJf6+! ..txf6 19 ..txf6 'ilNxf6 20 ..txc6 ..tf5 21 i.xe8 .l:.xe8 22 .l:.xb7, Fang-Van Wely, New York 1993, gave White a clear advantage) 17 ltJc7 ..tf5 18 ltJxe8 .l:.xe8! (The only move. 18 ... 'ii'xbl? 19 'iVxbl ..txbl 20 ltJc7 is just win­ning for White, who also has a big advantage after 18 ... i.xbl? 19 ltJxg7 ..tc2 20 'iVel 'iVd5 21 ltJxe5 'iVxe5 22 i.f6 'iVxd6 23 i.c4, Vaisser-Belov, USSR 1983) 19 .l:.xb7 ltJd4! 20 ltJxd4 cxd4 21 i.f3!, when White's chances are slightly better, thanks to the d6-pawn.

16 i.xf3 ..td4+! Or 16 ... 'iVd4+ with the same

effect. This check is not only useful from a general point of view, driving the white king away, but, as we can see below, it if; ~rtl1~ 11'1 n~rP<;s~rv f0T (,0n-

crete tactical reasons. 17 c;t>hl 'iVxal 18 'iVxal ..txal

19.1:.xal

19 ... .l:.e5! Probably Black can also

achieve equality after 19 ... ltJd7 with precise play, due to the position of White's king on hI. The game P. Garcia-Nunez, Ciego de Avila 1989, continued 20 i.e7! (Not 20 ltJc7?! .l:.e5!) 20 ... .l:.b8 21 ltJc7 .l:.f8 22 .l:.el b5! (This is better than Black's play in the game Vaisser­Khodos, Krasnodar 1978 [with the white king on gl]: 22 ... c4?! 23 ..txf8 c;t>xf8 [22 ... ltJxf8 fails because of 24 .l:.e8 c3 25 ..te4 i.f5? 26 lIxb8 ..txe4 27 ltJe6! fxe6 28 d7 c2 29 .l:.xf8+ c;t>g7 30 .l:.f1 +-] 24 .l:.e8+ [Even stronger is 24 iLd5! ltJf6 25 .txc4±] 24 ... c;t>g7 25 c;t>f2 ltJf6 26 lId8 b5, and now 27 ltJe8+ would have been winning) 23 .txf8 (23 ..td5? was proposed by Nu­nez as being advantageous for White, but the position turns in Black's favour after 23 ... .tb7! 24 ..txb7 lIxb7 25 ..txf8 ltJxf8 26 .!I.f'R rS;;f!7 17 .'t1.e7 ~bR)

Page 25: Beating King's Indian and Benoni

23 ... 'it>xf8 24 ~e8+ 'it>g7 25 iLc6 b4! 26 h3 (If 26 ..txd7? b3!=F. Note that this tactic doesn't work with the king on gl instead of hI) 26 .. .'~Jf6 and now instead of the losing 27 d7? iLxd7! the game should have continued 27 ~d8 c4! 28 ~e8+! (Both 28 d7? ~xd7 28 it.xd7 b3 and 28 ~a6? c3! are bad) 28 ... ~xe8 29 ~xe8 b3 30 axb3 cxb3 31 'it>h2! b2! 32 d7 iLxd7 33 ~xb8 it.xc6 with a drawn endgame.

20 iLf4! Here 20 ~e7+? is bad:

20 ... 'it>g7 21 ~xc8 (21 it.f4 ~d7 22 ..txe5+ ~xe5 23 ~xc8 ~xc8 24 it.xb7 ~d8=F Schol­seth-Nordahl, Gausdal 1993) 21...~d7 22 iLxb7 ~b8 23 iLc6 ~xg5 24 it.xd7 c4! 25 ~xa7 (25 h4 c3! 26 iLa4 ':'d5 27 ~xa7 [or 27 d7 ~xd7!-+] 27 ... ~b4 28 ~b5 ~xa4 29 ~xc3 .l:!xh4+ winning, Dobos-Schlosser, Bal­atonbereny 1989) 25 ... c3! 26 iLa4, Tozer-Schlosser, Oakham 1988, and now the simplest way was 26 ... .l:!c5 27 iLc2 (27 ~b5 is met by 27 ... .l:!bxb5! 28 iLxb5 c2!-+) 27 ... .l:!d5 28 .l:!c1 ':'xd6 29 a4 .l:!b2 30 h3 .l:!d2 31 it.e4 c2 32 'it>h2 f5 and Black wins.

If Black doesn't check on d4 (so that White's king stays on gl) our evaluation of this varia­tion must be changed, as the game Pesztericz-Boros, Buda­pest 1995, demonstrated. To avoid the problems of move "1

'm hf"r;., <'" W'" "t'lrt frn", ::> tl('-

Main Line with 9 .. . l:.e8 23

viation from line d analysed above: 14 .. :iWxb2 15 ~d5 ~xf3+ 16 iLxf3 'iVxal? 17 'ii xa 1 it.xa 1 18 .l:!xa 1 ':'e5 19 ~e7+! 'it>g7 20 ~xc8 ~d7 21 ..txb7 lIb8 22 iLc6 l:.xg5 23 iLxd7 c4 24 tiJxa7 c3 25 ':'cl! (Here this is possible) 25 ... l:.b2 26 iLc6 and White stands better.

There is one other nuance in this move order that is worth pointing out. After 16 .. :~Wxal? White can try 17 ~e7+?! ~h8 18 'iVxal iLxal 19 ":'xal, but in fact after the precise 19 ... 'it>g7! (but not 19 ... tiJd7? 20 ':'el ":'f8 21 it.h6 lId8 22 iLd2! f6 23 iLc3 lIf8 24 tiJd5 J:tb8 25 iLg4 b5 26 iLxd7 iLxd7 27 lie7+­Vaisser-Grigoriadis, Odessa 1977) 20 .l:!el it.e6! 21 iLxb7 tiJd7 22 i.xa8 ~xa8 White does not have enough compensation for the pawn.

20 .. Jhd5! The best plan is to return the

exchange. White is much better after both 20 ... tiJd7? 21 ~xe5 ~xe5 22 !leI and 20 .. .lH5? 21 ~e7+ 'it>g7 22 ~xf5+ gxf5 23 :e1 tiJd7 24 lIe8! ~b6 25 a4! with the idea of 25 ... iLd7 26 ~xa8 tiJxa8 27 a5+-.

21 iLxd5 ~c6 22 iLxc6 bxc6 23 lIc 1 iLd7 24 !lxc5 !lb8 25 h3

with equality e) 14 ... c4+ 15 'it>hl By including the check and

then playing 15 ... ~xf3 16 ..txf3 'iVxb2 Black tries to obtain the ~'lrn(' Txwitinn ~" thl" l::>st rliq

Page 26: Beating King's Indian and Benoni

24 Main Line with 9 ... l:le8

gram, but with an extra tempo ( ... c5-c4). Two sidelines cannot be recommended:

el) 15 ... ttJd3?! 16 ~xd3 cxd3 17 'iVxd3 .1Lf5 (Or 17 ... .1Le6 18 1::I.ac1 'iVa6 19 'iVxa6 bxa6 20 1::I.fdl with a stable advantage in the ending, Peev-Vukic, Kap­fenberg 1970) 18 'iV d2 ttJd7 19 1::I.adl1::1.b8, Gorelov-A. Shashin, Krasnodar 1980, and now 20 ttJd5! 'iVxb2 21 ttJe7+ would have given White the better po­sition.

e2) 15 ... ~e6?! 16 ttJxe5 ~xe5 17 ttJd5 'iVxd6 18 ttJf6+ 'it'f8 19 'iVxd6+ iLxd6 20 1::I.adl ~c7 21 ttJxe8 'it'xe8 22 1::I.c1 is clearly better for White, Konik­owski-Molderer, Balatonbereny 1989.

e3) 15 ... ttJxf3 16 ~xf3 'iVxb2 When I analysed this position

way back in 1979, I discovered what I believed to be a pleasing refutation of this move order. This line has not been in vogue since then, however, and to my regret no one has ever offered themselves as a victim.

In 1987 Konikowski pub­lished some analysis claiming that White wins easily after 17 ~d5 ~e6 18 ~xe6 fxe6 19 'iVf3 ttJc6 20 'iVf7+ ~h8 21 .1Lf61::1.g8 22 1::I.b 1 etc., and ever since then I have been hoping that some­body would find 19 ... 1::I.f8! (in­stead of the suicidal 19 ... ttJc6?) 20 'iVe4 ttJc6 21 'iVxe6+ 'it'h8 22 ttJd5 1::I.f5! with an advantage for Ph~k ~nr1 "'("11'1<''1 th"'rp +0rf' trv

to trap me. But sadly this has not happened and I have there­fore decided to publish my home preparation in this book.

17 .1Ld5! .1Le6!

The only move. All the alter­natives are losing: 17 ... .1Lf5? 18 .1Lxf7+ 'it'xf7 19 'iVd5+ 'it'f8 20 .1Le7+ 1::I.xe7 21 dxe7+ ~xe7 22 1::I.ael+; 17 ... 1::I.f8? 18 iLe7! (18 1::I.xf7?! 1::I.xf7 19 .1Lxf7+ 'it'h8! is less clear) 18 ... 'iVxc3 19 1::I.c1 'iVe5 20 .1Lxf8; and 17 ... 'iVxc3? 18 iLxf7+ 'it'h8 19 i.xe8 'iVxal 201::l.f8+!

18 ~xf7! ! 'it'xf7 If 18.Jxc3?! 19 1::I.e7! Black

has two choices: 19 ... 1::I.xe7?! 20 dxe7 ttJc6 21 iLxe6+ 'it'h8 22 1::I.cl± and 19 ... ttJc6! 20 1::I.xe8+ 1::I.xe8 21 d7 1::I.f8 22 iLxe6+ ~h8 23 1::I.c1 'iVe5 24 'iVd5t, while if 18 ... ttJc6 White can play 19 1::I.xg7 + ~xg7 20 i.xc6 bxc6 21 1::I.c1! with a strong initiative. Black could try one more alter­native: 18 ... .1Lxf7 19 ~xf7+ ~h8 20 .1Lxe8 'iVxc3 with com­plicated play.

19 d7!! Th" ("Inl" m0vpf

Page 27: Beating King's Indian and Benoni

19 .. .'~Jc6! 19 ... ttJxd7? IS weaker: 20

Jl.xe6+ ~xe6 21 ~xd7+ Wg8 22 ~xe6+ Wh8 23 ~cl h6 24 ~h3±.

20 dxe8~+ ~xe8 21 Jtxe6+ 21 ~f3+? doesn't work after

21...Wg8 22 Jtxe6+ ~xe6 23 IHlttJe5 24 ~d5 ~b6.

21...~xe6 22 ~d7+ ttJe7 23 Mfl + .1f6 24 .1xf6

Black is the only one with chances to win the endgame after 24 ttJd5? (Or 24 ~d4? ttJf5) 24 ... ~f2! 25 ~xe6+ Wxe6 26 ~xf2 Jtxg5.

24 ... ~xf6 When I originally analysed

this position, I believed that White was winning here, but I had missed Black's simple de­fence after 25 ~e I? ~b4 26 a3 ~d6+. What a pity! White has nothing better than a perpetual check after:

25 ~xf6+ Wxf6 26 ttJe4+ WO 27 ttJg5+ Wf6 28 ttJxh7+ WO 29ttJg5+ Wf6.

e4) Black can avoid all these complications by a more precise move order:

15 ... ~xb2! 16ttJd5 16 ttJxe5!? works well after

16 ... ~xc3? 17 Jtxc4 ~xe5 18 .1xf7+ Wh8 19 ~xe8 ~xg5! 20 ~c1 ~d7 21 ~c7 with the ad­vantage, but 16 ... ~xe5! 17 Jl.d2 poses some problems for White and needs to be tested. Another interesting possibility is 16 ~cl !?, when one nice variation runs 16 ... ttJxf3 17 ~xf3 .1xc3

Main Line with 9 ... ~e8 25

18 ~bl! ~xa2! 19 ~dS .1fS 20 ~xf5! gxf5 (20 ... l:.el+? 21 ~xel ~xel 22 l:.xf7±) 21 ~xf7+ WxO 22 'iYhS+ Wf8 23 .1h6+ ~g7 24 ~xfS+ Wg8 25 'iYd5+ (25 ~fl doesn't win be­cause of 25 ... c3!) 2S ... Wh8 26 ~xg7 + with a perpetual check.

16 ... ttJxf3 17 ..txf3 Not 17 ~b I? 'iYxe2! 18 ttJe7+

~xe7!+. 17 ... ~xal 18 ~xal i.xal 19

~xalttJd7! The alternative 19 ... ~e5 is

similar to line d with one ex­ception. After 20 ..tf4 ~xd5! 21 ..txd5 ttJc6 White can play 22 .1xc4 with compensation for the pawn. 19 ... ttJd7! is more ambitious.

20 i.e7 ~b8 21 ttJc7 ~f8 22 ~el b5

We have arrived at the posi­tion studied in line d above with an important extra tempo for Black ( ... c5-c4). The question is now whether White can even hold a draw. The game Mayer­Anageldiev, USSR 1977, (reached by another move or­der) continued 23 ~c6 c3 (23 ... b4!?) 24 i.xb5? (24ttJd5!) 24 ... c2! 25 ~c1 .1a6 26 a4 ~xb5 27 axb5 ~fc8 with a clear advantage for Black.

15 ttJd5 Black can quickly gain an

advantage after 15 ~b5?! ttJbc6 16 ttJxe5 ~xe5 17 ..tc4 (17 a4) 17 ... ~xb2 18 ~cl ~e6 19 i.xe6 ~xe6 20 ~f3 ~f8, Kou­atly-Jadoul,Ostend 1984.

Page 28: Beating King's Indian and Benoni

26 Main Line with 9 ... 1:1e8

15 ~xb2 16 ltJe7+

Nowadays Black has found good antidotes to this move, so it is White's tum to propose something new. Perhaps the clever 16 ~c1, waiting for the queen's knight to move? After 16 .. .'~Jbd7 the check 17 ltJe7+! is increased in strength, while if 16 ... ltJc6 White simply plays 17 ltJc7, keeping his pawn on d6. Another try was not successful in the game Fang-Minasian, Philadelphia open 1991: 16 ltJh4? ~c2 17 '+lUd2 ~d4+ 18 '+lUxd4 cxd4 19 ltJc7 ~bc6 20 ~b5 ~f8 and Black is better.

16 ~xe7 This is much stronger than

16 ... ~h8?! 17 ~xf5 gxf5. 17 dxe7

Only once has White tried the unhappy 17 ~xe7?! After the reply 17 ... ~bc6 18 ~c1 ~xe7 19 dxe7 ~xf3+ 20 ~xf3 ~e8 21 ~h1 ~d4 Black's strong bishops assure him of an ad­vantage, Safin-Akopian, Yur­mala 1985.

17 ~bc6 The knight is better placed

here than on d7. 18 ~c1?!

White needs an improvement here:

a) 18 'iVe1? was refuted in the game Rechlis-Badea, Haifa 1989: 19 ... h6 19 ~h4 g5! 20 ~g3 '+lUxal! 21 ~xa1 ~xf3+ 22 ~xf3 ~xa1 23 ':'xa1 ~xe7 24 ~xb7 ~d8 with a clear advan-

tage for Black. b) 18 ~hl!? seems the best

here, posing a nice trap: b1) 18 ... ~xf3?! 19 ~xf3

~xa1? 20 ~xc6! (Nunn) 20 ... '+lUxd1 21 ':'xd1+- !lic-San­dic, Yugoslavia 1989.

b2) The ambitious 18 ... h6 19 ~h4 ~g4?! met with a precise answer in the game Barsov­Arsovic, Budapest 1991: 20 ~d6! ~e3 (Or 20 ... .:.e8!? 21 ':'ae1! ~c3 22 '+lUxc5 ~xe1 23 ':'xe1 and the knight on g4 is in danger) 21 ':'fe1 c4 22 .tf2 ~c2 23 ~xc4! ~xe1 24 ':'xe1 ~xf2 25 e8'+lU+ ':'xe8 26 ':'xe8+ ~h7 27 .txf7 ~f1+ 28 ~g1 .td4 29 .ii.g8+ 1-0.

b3) After 18 ... .:.e8 White can play 19 ~xe5 (19 ':'c1?! trans­poses to 18 ':'c1?! ':'e8 19 ~h1 considered below) and now Black has tried:

b31) 19 ... '+lUxe5 20 ~c4 ~e6?! (20 ... ~xe7!?) 21 ~xe6 '+lUxe6 22 ':'b 1 f6 23 ':'e1 ~e5 24 ':'xb7! fxg5 25 '+lUd8 cJi;f7 26 ':'f1 + .tf6 27 ':'b8± Guidi­Mauro, corr. 1984.

b32) 19 ... ~xe5 20 ':'c1 ~d4 21 ':'xf5 gxf5 22 ~c4 ~e5?! (Possible is 22 ... ~xe7!?) 23 ':'bl! '+lUf2? (23 ... '+lUa3 24 ~b5! ~c6 [24 ... ~g4? 25 '+lUxg4!+-] 25 ~a4!;!;) 24 ~a4! ~c6 25 ~d5+- Moutousis-Cela, Zoub­eri Zonal 1993.

All these variations need further tests.

18 ':'e8 19 ~xe5

.

Page 29: Beating King's Indian and Benoni

After 19 'it>h1, as well as 19 .. .'~Jxe7 20 t2Jxe5 'iVxe5 21 iLc4liJc6 22 'iVb3, Peev-Makro­poulos, Bulgaria-Greece 1973, 22 ... 'i'c7!? 23 .l::i.cel! with com­pensation, Black has the un­pleasant resource 19 ... h6! 20 ..th4 liJg4! (Or 20 ... t2Jxe7 21 liJxe5! 'i'xe5 22 iLb5 t2Jc6 23 .l::i.e1 iLe4 24 'i'd7, Peev-Sykora, Moscow 1977, and now the move 24 ... .l::i.e6 would have been approximately equal) 21 !!e1?! (21 'i'd2 'iYxd2 22 t2Jxd2 liJxe7 23 .l::i.xc5+) 21.. . .l::i.xe7! 22 .l::i.xc5 ..tf6 23 iLxf6 'i'xf6 24 'it'g 1 ..te4!+ 25 h3? 'iYf4 26 'i'c1 'i'g3 0-1 Schoen-Safin, Ger­many 1995.

19 'i'xeS 20 i.bS ~e6?!

My idea of 20 ... ~e4! was successfully tested by Marin against Barsov (Budapest open 1990). That game continued 21 ..th4 (21 'i'd7 'iYd4+; 21 'iYb3!? ..td5+, Marin) 21...'iYb2 22 'iYg4 j"d4+ 23 'it>hl f5 24 j"c4+ 'it'g7 25 'i' g3 liJxe7 with a clear ad­vantage for Black.

21 j"xc6 'i'd4+! A necessary intermediate

check. 21... bxc6? loses after 22 'i'd8 'iYd4+ 23 'it>hl 'iYd7 24 .l::i.cdl ~d4 25 ~f6! h6 26 .l::i.fe1! 'it>h7! 27 .l::i.xe6 'i'xe6 28 j"xd4 'i'xe7 29 'i'xe7 .l::i.xe7 30 ..txc5.

22 'it>h1 bxc6 23 iLe3 'iYxd1?!

23 ... 'iVxe3? 24 'iWd8 was bad, so the only move was 23 ... 'i'h4! with an unclear position.

Main Line with 9 ... ~e8 27

24 .l::i.fxd1 ~d4?

After 24 ... ~d5 25 j"xc5 ~f6 26 .l::i.e 1 ~e6 White has a slight plus but one which will be very difficult to convert into a win

25 .l::i.xcS! ~xe3 Black loses quickly after

25 ... ~xc5? 26 .l::i.d8 ~d7 27 j"xc5, and more slowly after 25 .... iJ6 26 .l::i.d8! .l::i.xd8 27 exd8'iV+ j"xd8 28 .l::i.xc6.

26 .l::i.d8 iLxcs 27 ~xe8+ 'it>g7 28 .l::i.g8+ 'it>xg8 29 e8'iV+ j"fS

29 ... 'it>g7 would not have helped to save the a7-pawn after 30 'iVxc6 iLb6 31 'i'c3+! 'it>g8 32 a4.

30 a4 ..tb3? Only active counterplay

would have given Black any chances to save the game: 30 ... c5! 31 'iVa8 c4 32 'iVxa7 c3 33 'iVc7 iLg7 34 a5 iLf5 35 a6 c2 36 a7 .i.e4 37 'iVc8+ iLf8.

31 as! .i.c4 32 'iVxc6 iLe2 33 a6 1-0

Page 30: Beating King's Indian and Benoni

28 Main Line with 9 ... l:Ie8

At last the a7-pawn falls (33 ... hS 34 'iYb7 SLcS 3S 'iYb8+ ~h7 36 'iWeS wins a bishop).

Game 3 Cranbourne-Crespo

corr. 1988-89

1 d4 4Jf6 2 c4 g6 3 4Jc3 SLg7 4 e4 d6 S f4 0-0 6 4Jf3 cS 7 dS e6 8 SLe2 exdS 9 cxdS l:i.e8 10 eS dxeS 11 fxeS 4Jg4 12 SLgS

12 f6?!

This move is certainly worse than 12 ... 'iYb6. Instead of trying to ride the storm of White's ini­tiative with an extra pawn, Black will suffer for nothing.

13 exf6 SLxf6 14 'iYd2!

Also possible are 14 ~xf6 'iYxf6 IS 0-0 4Je3 16 'iYd2 4Jxfl 17 l:i.xfl SLfS and 14 ~f4 c4 IS 'iY d2 ~fS 16 4Jd4 ~xd4 17 'iYxd4 4Jc6 18 'iYcs 4JceS 19 0-0, Grivas-Fernandez, Sharjah 1985, with complicated play in both cases.

14 SLfS Black's problems are not re­

solved by other moves: a) 14 ... bS transposes to Game

11, Blokh-Kitchev. b) 14 ... 4Jd7?! IS 0-0 ct:JdeS

16 SLxf6 4Jxf6 (Or 16 ... 'iYxf6 17 4JgS! 'iYb6 18 4Jge4 c4+ 19 ~hl SLfS 20 d6! 4Jd3, Peev­Donner, Cienfuegos 1973, and now according to Peev White could have won after 21 d7! l:i.eS 22 SLxg4 SLxe4 23 l:i.ael! 'iY d4 24 l:i.xe4 l:i.xe4 2S 4Jxe4 'iYxe4 26 'iYgS! 'iYd4 27 'iYe7!) 17 SLbS! 4Jxf3+ 18 l:i.xf3 l:i.f8 19 d6 SLfS 20 l:i.el! with a large advantage, Arencibia-Salcedo, Bogota 1990.

c) 14 ... 4JeS?! IS a-a-a! 4Jxf3 16 ~xf6 4Jxd2 17 ~xd8 l:i.xd8 18l:i.xd2 with a clear advantage.

d) 14 ... ~xgS IS 'iYxgS and now:

dl) IS ... 'iYb6?! 16 a-a-a! 4Jf2 17 d6 'iYd8 18 SLc4+ ~g7 19 'iYxcs 4Jxdl 20 l:i.xdl SLfS 21 4JdS 4Jc6 22 4Jc7 with a clear edge for White, Kuempers­Eitel, Bayern 1991.

d2) IS ... 4JeS 16 'iYxd8 (Even stronger is the move 16 a-a-a!?) 16 ... 4Jxf3+ 17 ~f2 l:i.xd8 18 ~xf3 4Jd7 19 l:i.he 1 with a small but stable advantage for White, Vaisser-Levic, Vrnjacka Banja 1986.

d3) IS ... 4Je3 16 'iYxd8 l:i.xd8 17 ~f2 ct:Jg4+!? 18 ~g3 lbe3 19 l:i.ael ct:JfS+ 20 ~f2 ct:Jd7 21 4Je4 l:i.f8 22 d6 with an un­pleasant position for Black,

Page 31: Beating King's Indian and Benoni

Forintos-Enklaar, Wijk aan Zee 1974.

d4) 15 ... ltJa6 16 "ii'xd8 .l:i.xd8 1 7 h3 ltJe3 18 ~f2 t'Llf5 19 iLxa6!? bxa6 20 .l:i.ad 1 ..l:i.b8 21 b3 ltJd6 22 ltJe5 .l:i.f8+ 23 ~g3 .~f5 24 .l:i.hel± Forintos-Ghit­cscu, Wijk aan Zee 1974.

d5) 15 ... il.f5 16 "ii'xd8 (Or 16 h3 ltJe5 17 "ii'xd8 t'Llxf3+ 18 ~f2 .l:i.xd8 19 ~xf3 t'Lla6, and now instead of 20 g4?! jLd3! 21 .l:i.he1 .l:i.d7 22 .l:i.e3 c4 23 .l:i.ael ltJc5 with equality, Gnichtel­Dobos, Budapest 1996, White should try 20 .l:i.adl! with the idea of 20 ... c4 21 .l:i.d4! which promises an advantage for White) 16 ... .l:i.xd8 17 h3 ltJf6 18 g4! ~e4 19 g5 il.xf3 20 ~xf3 ltJe8 21 0-0 with a clear advan­tage, Krupkova-Peng, Kishinev Women's Interzonal 1995.

d6) 15 ... "ii'xg5 16 ltJxg5 and: d61) 16 ... ltJe3 17 ~f2 (17

~d2!?) and now: d611) 17 ... ltJf5?! 18 ~b5 (18

ltJe6!?) 18 ... ~d7 19 ltJe6 .l:i.c8 20 .l:i.adl a6 21 ~e2± Kniest­Heck, Giessen open 1993.

d612) 17 ... il.f5?! 18 ltJe6!? (18 h3 h6?! 19 ltJf3 a6?! 20 g4 ~d7 21 ~d3 ~g7 22 .l:i.ae1 1-0 was Vaisser-Khalafian, Yerevan 1996) 18 ... ~xe6 19 dxe6 .l:i.xe6? (19 ... ltJc2±) 20 .l:i.ae 1 ttJc6 21 ~f3 .l:i.ae8 22 ltJe4+- Reinemer­Maniocha, Porz open 1993.

d613) 17 ... ltJc2! 18.l:i.acl (18 ~adl!? ~f8+ 19 ltJf3) 18 ... ~f8+ 19 ~f3 ltJd4!? (19 ... ttJb4?! 20 ~hd 1 ~f5 21 ltJe6 ttJd3+ 22

Main Line with 9 ... kte8 29

.l:i.xd3 ~xd3 23 ltJxf8 ~xf8 24 d6! ltJd7 25 .l:i.el ~e8 26 ~xe8+ ~xe8 27 ~xb7± Kahn-Balogh, Budapest 1993) 20 .l:i.he1 ltJa6 21 a3 iLg4 22 ~g3 ~xf3 23 ltJxf3 ~ad8 24 ~e5;l; Dobos­Balogh, Budapest 1995.

d62) 16 ... ~f5 17 h3 (It would be interesting to tryout 17 ltJe6!?) and now:

d621) 17 ... ltJe3? 18 g4! h6! (Not 18 ... ~d7? 19 ~d2! ltJg2 [19 ... ~e5 20 ltJge4+- Vaisser­Arizanov, Pula open 1988] 20 ~hf1 h6 21 ltJge4 ~g7 22 ~f2 ltJh4 23 ~af1 ~e7 24 d6 1-0 Petursson-Blumberg, San Ber­nardino 1989) 19 ltJf3 ~e4 20 ltJxe4 ~xe4 21 ~f2 ltJxd5 22 .l:i.hdl ltJb6 23 .l:i.d6 ~f7 24 il.d3 ~e6 25 ~xg6+± Vegh-Meta­xas, Iraklion 1992.

d622) 17 ... ltJe5!? was played in Kakageldyev-Kapengut, Len­ingrad 1969. Here 18 O-O-O!? c4 19 ltJge4 or 18 ltJge4 would have kept a small plus for White.

15 O-O! ~xg5 White was preparing 16 ~f4,

followed by h2-h3 and g2-g4, so this exchange seems logical. What happens if Black waits?

a) The old game Vaisser­Kaminnik, Rostov 1970, saw 15 ... ltJa6?! 16 ~f4! ltJb4 and now instead of 17 ~b5?! c4! with counterplay, White should have continued 17 h3 ltJe5 18 ltJxe5 ~xe5 19 ~xe5 ~xe5 20 g4 il.d7 21 a3 ltJa6 22 "ii'f4 with a clear edge.

Page 32: Beating King's Indian and Benoni

30 Main Line with 9 ... .l:Ie8

b) 15 ... ltJd7 16 ..tf4!, a move I proposed as an improvement to my game against Thipsay, Delhi 1987, where after 16 h3?! ..txg5 17 ltJxg5 ltJe3 18 IIxf5 ltJxf5 19 ..tbS! Black could have continued 19 ... ltJd6!, when White can equalise only with precise play: 20 ltJe6 ~e7 21 ltJc7 IIac8 22 ltJxe8 IIxe8 23 a4! a6 24 ..txd7 ~xd7 25 ~f2! My suggestion was checked in the game Arencibia-Paneque, Holguin (match) 1988, where after (16 ..tf4) 16 ... ltJge5 Aren­cibia has suggested 17 d6! ltJxf3+ 18 ..txf3 ..td4+ 19 Whl ltJeS 20 ..tdS+ Wh8 21 ltJb5 with a clear edge for White. Another good option for White is 16 ..tbS!? instead of 16 ..tf4.

16 ~xg5?! Here I want to bring to your

attention some old analysis of mine, which I have kept under wraps for more than a quarter of a century:

J6ltJxgS! ltJe3 17 ..tb5!

Also possible is 17 IIxf5!? gxf5 18 ltJce4! f4, as in Kret­Findlay, Toronto 1995, and now

19 ~c3! would have assured a strong initiative.

17 ... ltJxfl White is better in the case of

17 ... ltJd7 18 IIxf5! ltJxf5 19 ltJe6 or 17 ... IIe5 !? 18 ltJf3 ltJxfl 19 IIxfl IIe7 20 d6.

18 IIxfl In return for the sacrificed

exchange White has a strong attack.

18 ... IIf8 Other moves give White an

advantage which is somewhere between clear and decisive:

a) 18 ... ltJd7 19 d6! b) 18 ... ..td7 19 d6! c) 18 ... ltJc6 19 g4! h6

(19 ... ..txg4 20 ~f4) 20 gxf5 hxg5 21 f6 Wf7 22 ~xgS a6 23 ~h6.

d) 18 ... IIe5 19 ~f4! ~e7 20 ltJf3 IIe4 21 ltJxe4 ~xe4 22 ~d6!

19..tc4! White must be precise. 19

d6? spoils all the previous ef­forts: 19 ... ltJc6! 20 ~d5+ (Or 20 ..tc4+ Wh8 21 ltJce4 ~a5! with enough play) 20 ... Wg7! 21 ~xf5 ~xf5 22 ltJe6+ Wh8! 23 ltJxd8 IIxdS 24 ltJxb7 ltJa5! 25 ltJxa5 IIxd6 with an unclear position.

19 ... h6! 19 ... ltJd7? 20 d6+ Wh8 21

IIel! ltJf6 22 IIe7 is desperate for Black, while White is also much better after 19 ... ~f6 20' d6+ Wh8 21ltJd5.

20 d6+ Wg7 21 :xf5 hxg5 22 :xc5 ltJc6 23 ltJe4

Page 33: Beating King's Indian and Benoni

With one (and soon a second) pawn for the exchange and a powerful attack White is clearly better.

16 'i'xg5?! Black has two alternatives: a) 16 ... t2Je3? is not good: 17

'i'h6! 'i'e7 (If 17 ... t2Jxfl 18 t2Jg5 'Wie7 19 ~xf1 'i'g7 20 'i'xg7+ rJ;;;xg7 21 g4 and, even without queens, the attack con­tinues) 18 .tb5! .td7? (better is 18 ... t2Jd7, but still not good enough) 19 ~ael .txb5 20 t2Jxb5 t2Ja6 21 t2Jg5 and White won, Peev-lanosevic, Nis 1972.

b) The best option is 16 ... t2Jd7 17 h3 (17 .tb5!?) 17 ... t2Je3 18 'i'h6 t2Jxfl 19 t2Jg5 'i'e7 20 d6 and Black now has a choice between perpetual check with 20 ... 'i'g7 21 .tc4+ ~h8 22 t2Jf7+ and a complicated and risky continuation 20 ... 'i'e3+ 21 ~xfl t2Jf8 22 ~d 1 ~ad8!, Schrammel-Zsu. Polgar, Fon­yod 1983.

17 t2Jxg5 t2Je3 18 .tb5! ~d8?

Main Line with 9 ... .l::i.e8 31

18 ... ~f8!? 19 ~f2 a6 was necessary in order to minimise White's advantage. Now White sacrifices a whole rook and starts a terrible attack. The end of the game is a splendid and thematic illustration of the enormous strength of White's pieces in harmony with the passed d-pawn.

19 l:.fel!! t2Jc2 20 :e7! a6

Nothing can save Black: a) 20 ... t2Jxal 21 g4! .tc2 22

.tc4 :d7 23 d6+ 'iith8 24 ~e8+ rJ;;;g7 25 t2Je6+.

b) 20 ... h6 21 :fl hxg5 22 g4 a6 23 .lta4!

c) 20 ... t2Jd7 21 ~fl t2Jf6 22 .tc4.

White has a decisive advan-tage in all three cases.

21 .tc4 b5 22 ~f1 bxc4 23 g4! t2Je3

Otherwise: a) 23 ... .ltxg4 24 ~ff7 .tf5 25

~g7+ ~h8 26 :xh7+ ~g8 27 t2Jce4 t2Jd7 28 t2Je6 winning.

b) 23 ... h6 24 t2Jh7! .txg4 25 ~ff7 t2Jd7 26 t2Je4 mating.

c) 23 ... t2Jd4 24 gxf5 gxf5 25 t2Jxh7 winning.

24 gxf5! t2Jxf5 Or 24 ... t2Jxfl 25 f6 :f8 26

~g7+ 'iith8 27 ~xh7+ ~g8 28 f7+ :xf7 29 :x.f7 t2Je3 30 h3! and wins.

25 ~c7! t2Jd7 26 t2Jce4! t2Jf8 27 t2Jf6+ ~h8 28 ~xf5! gxf5

Page 34: Beating King's Indian and Benoni

32 Main Line with 9 .. Jle8

29 CDe6 lid6? This accelerates the inevita­

ble end. 30 lig7 1-0

Game 4 Lagontrie-Boulard

French Carr. Ch. 1987

1 d4 CDf6 2 c4 g6 3 CDc3 ~g7 4 e4 d6 5 f4 0-0 6 CDf3 c5 7 d5 e6 8 ~e2 exd5 9 cxd5 lie8 10 e5 dxe5 11 fxe5 CDg4

12 0-0 12 ~f4 ttJxe5 13 0-0 reaches

the game continuation via a transposition of moves.

Another interesting and little explored idea for White is

12 e6 fxe6 13 d6!? 13 0-0 has proved to be less

promising: 13 ... exd5! 14 ttJxd5 (Or 14 ~g5 'iYd6! 15 h3 ttJf6 16 ~xf6 ~xf6 17 ttJxd5 'it>h8 18 ttJxf6 'iYxf6 19 ~b5 ttJc6 20 'iY c2 .1i.f5 with pleasant equality for Black, Klompus-Waagmees­ter, COIT. 1990) 14 ... ~e6 15 ttJf4 (White's show of aggression 15 ~c4 ttJc6 16 .1i.g5 ttJf6 17 ttJe5, as in Kotov-Burechell, Stock­holm 1959, fails to 17 ... ttJxe5! 18 ~xf6 ttJxc4 19 ~xd8 liaxd8 20 ttJe7+ 'it>h8 21 ttJxg6+ hxg6 and Black's three pieces are stronger than the queen) 15 ... 'iYxdl 16 lixdl .1i.f7 17 ttJg5 ttJe5, Khodos-Portisch, Lipetsk 1968, and now instead of 18 ~e3? ~c4! White could have achieved some compensa-

tion for the pawn with 18 CDxf7 ttJxf7 19 .1i.c4.

13 ... ~d7 Several other moves have

also been tried here: a) 13 ... ttJe5 (13 ... a6 14 a4!?;

13 ... IU8!?) 14 0-0 (14 ttJe4!?) 14 ... ttJxf3+ 15 ~xf3 .i.d4+ 16 'it>hl 'iYxd6 17 ttJb5 'tid7 18 ttJxd4 cxd4 19 ~h6 ttJc6 20 lic 1 'ti d6 21 'tib3 .i.d7 22 'tixb7 liac8 23 'tia6 e5 24 .i.e4 with rich play for only one pawn, Berkovich-A. Kuzmin, Moscow 1981.

b) Two examples after the move 13 ... ttJc6:

b 1) 14 ttJg5 ttJh6 15 0-0 ttJf5 16 ttJge4 h6 17 ttJb5 lif8 18 ttJc7 lib8 19 ~c4 ttJcd4 20 ttJxc5 'tixd6 21 ttJ5xe6 .1i.xe6 22 ttJxe6 'tic6! 23 ttJxg7+! 'it>xg7 24 ~d3;l; Vasilchenko­Kovalev, Katowice 1990.

b2) 14 0-0 ~d7 15 ttJe4 'tib6 16 ttJfg5 c4+ 17 'it>h1 ttJe3 18 .i.xe3 'tixe3 19 lif3 'ti d4 20 'tiel lif8 (If 20 ... h6?! 21 lidl 'iYe5 22 'iYh4! ttJd4 23 lixd4! 'iYxd4 24 ttJf7 with a strong at­tack) 21 lid1 'iYe5 22 ~xc4 h6

Page 35: Beating King's Indian and Benoni

23 Mxf8+ Mxf8 24 '2lf3;J;; Michaelsen-Uhlmann, German Bundesliga 1995.

c) 13 ... ~b6 14 '2lg5 '2le5 15 0-0 '2lbc6 16 '2lge4 (16 ~el!? l/ld4 17 ~h4 h6 18 CZ'lge4 ctJf5 19 ~h3 looks reasonable for White) 16 ... ctJd4 17 CZ'la4?! ~c6 18 '2laxc5 b6 19 ~xd4! (Probably not quite correct, but amusing) 19 ... '2lf3+ 20 .txf3 ~xd4+ 21 'it>hl ~b5 22 .te3 .f'Lg7? (22 ... ii.xe3!?) 23 a4! ~xb2 24 d7 .txd7 25 '2lxd7 1.Jed8 26 Mabl ~a3 27 CZ'lef6+ '.'Qh8 28 ii.xa8 Mxa8 29 .td4 'i'xa4 30 Mf4 Md8 31 ii.e5 ~a5 32 Mbfl ~d2 33 '2le8 'it>g8 34 n.f8+ 1-0 Bach-Watzke, Dres­den 1996.

14 '2lg5 ctJe5 15 0-0 Ac6 16 .lfH4 CLlbd7 17 ~e1 h6

Black tried another plan in the game Bach-Schoene, RLNN 1990: 17 ... c4!? 18 'it>hl b5 19 ~g3 b4 20 ~h3 (A typical ma­noeuvre of the white queen) 20 ... '2lf8 21 '2lce4 ii.xe4 22 '2lxe4 ~ a5 23 CLlg5! h6 24 '2le4 ~d5 25 ~e3 CLlfd7 26 Madl with complicated play not unfa­vourable to White.

18'2lge4a619a4 Now instead of 19 ... Mb8?! 20

~ g3 b5?! 21 axb5 axb5 22 ii.xe5! '2lxe5 23 Ma7 ii.d7 24 ii.xb5! c4 25 Mc7 with a big advantage for White, Michael­sen-Holzer, Oberwart open 1992, it was better to give back a pawn immediately: 19 ... b5!? 20 axb5 axb5 21 Mxa8 ~xa8 22

Main Line with 9 ... Me8 33

ii.xe5! ii.xe5 23 ii.xb5, when equality is not far off.

12 '2lxe5 12 ... ii.f5?! allows a transpo­

sition to the game Semkov­Apicella in the line 12 .tg5 ~b6 13 0-0 ii.f5 after 13 .tg5! (13 ii.b5!?) 13 ... ~b6 14 ctJh4, with an advantage for White.

13 ii.f4 '2lbd7 Instead of this solid move,

Black can play: a) 13 ... a6?! is original, but

looks suspicious, e.g. 14 ~d2 '2lbd7 15 d6 b5 16 '2ld5 Ma7, Ellenbroek -Legemaat, Enschede 1995.

b) 13 ... CLlxf3+ 14 .txf3 may transpose to 13 ... '2lbd7 after 14 ... CLld7. Less well founded is 14 ... ~f6?! 15 .tg3! '2ld7 16 d6 Mb8 17 'it>hl and Black's queen is obliged to go home: 17 ... ~ d8 with a good position for White.

c) 13 ... ~b6 may also trans­pose to 13 ... ctJbd7 after 14 d6 '2lbd7. A difference arises if Black takes the b2-pawn as in the main line 14 d6 (Game 2).

Page 36: Beating King's Indian and Benoni

34 Main Line with 9 .. J~e8

The position of White's bishop on f4 instead of g5 is in White's favour in the vanatton 14 ... 'i'xb2 15 ctJd5 ctJxf3+ 16 i.xf3 'ilid4+ 17 ~hl 'i'xal 18 'i'xal i.xal 19 ~xal, as Black has no good defence to the twin threats of 20 ctJc7 and 20 ctJf6+.

d) 13 ... i.f5!? This move seems to be the

best here. It is natural to de­velop the bishop first.

14 ctJxe5 Axe5 15 'iii d2 The old game Chacet-Murey,

Moscow 1960, saw instead 15 i.xe5!? ~xe5 16 'iii d2 ctJd7 17 ~adl (If 17 g4 i.xg4! 18 i.xg4 ~g5 19 h3 f5 winning the sec­ond pawn). Now instead of 17 ... 'i'b6? 18 g4 i.e4 19 i.b5! c4+ 20 ~f2 i.f3 21 i.xd7 i.xd 1 22 'i'xdl with a clear advantage for White, Black should have played 17 ... h5!, preventing 18 g4, with the better chances.

15 ... ctJd7 16 B:adl Perhaps it was time to ex­

change bishops with 16 i.xe5!? 16 ... a6 16 ... i.d4+!? 17 ~hl a6 18 g4

i.xc3 19 bxc3 i.e4+ looks sat­isfactory for Black

17 i.xe5 ctJxe5 18 'ilif4 c4 19 g4l

The game Vaisser-Renet, Brussels Zonal 1993, continued 19 ... 'ilib6+ 20 ~hl (20 ~f2!?) 20 ... i.d3? (20 ... 'i'xb2!?) 21 i.xd3 ctJxd3 22 'ilixf7+ ~h8 23 ~d2 with an advantage for White, thanks to the strong passed d-pawn.

14 d6 Taking into account the

problems that White met in the game S. Ivanov-Shulman (see commentary to the next move), Belov has proposed 14 i.b5!? 'ilib6 (stronger is 14 ... a6!) 15 ctJxe5 i.xe5 16 i.xe5 ~xe5 17 'ilif3 ~f5 18 'i'e2 with a draw by repetition of moves.

14 ctJxf3+ Let us see what happens if

Black tries to vary: a) 14 ... ~b8 15 ctJd5 b5 16

ctJc 7! ~f8 17 ctJxe5 ctJxe5 18 'ilid5 is slightly better for White.

b) 14 ... a6?! 15 ctJd5 ~f8 16 ctJg5 ctJc6?! 17 Ac4 b5?

Page 37: Beating King's Indian and Benoni

runs into 18 ttJxf7 ! :lxf7 19 ttJe7+ ttJxe7 20 J.xf7+ <Ji;xf7 21 dxe7 J.d4+ 22 ~xd4! ~xe7 23 ~g5+ 1-0 Alcock-Fayne, corr. 1990.

c) 14 ... ~b6!? This move, which was con­

sidered for a long time to be incorrect, was rehabilitated in the game S. Ivanov-Shulman, Minsk 1995, which continued: 15 ttJxe5? ttJxe5! 16 .i.xe5?! (Belov suggested as an im­provement now or one move earlier 16 ttJd5!? ~xd6 17 .Jtg5 ~c6 18 ttJe7+ :lxe7 19 J..xe7 .Jtf5 with an unclear position. In fact I prefer Black here) 16 ... :lxe5 17 d7 .Jtxd7 18 ~xd7 ~xb2. The point! White cannot keep his extra piece and finds himself in a difficult position. N ow instead of 19 :lael? :lf5! with a winning position for Black, it would have been better to play 19 ~xf7+ Wh8 20 J.f3 ~xc3 21 ~xb7 :lae8, although Black keeps an edge because the a7-pawn is untouchable (22 ~xa7?? 'iWd4+ 23 'ii'xal!-+).

An answer, from White's viewpoint, to the Shulman chal­lenge was given in the game Vaisser-Nataf, French Team

. Ch. 1997 (added as this book was going to press):

15.tb5!

(seejollowing diagram)

Here Black again has a choice:

Main Line with 9 ... :e8 35

el) Accepting the sacrifice 15 ... ttJxf3+? 16 ~xf3 .Jtd4+ (or immediately 16 ... .Jtxc3) 17 Wh1 ..txc3 18 'iVxc3 'iVxb5 gives White a decisive attack: 19 :lfel! ~f8 (No help is offered by 19 ... ~xe1+ 20 ~xe1 ttJf8 21 .Jth6 ttJe6 22 ~f6 ~d3 23 'iW d8+! mating) 20 .Jth6 f6 21 .Jtxf8 Wxf8 22 'iWh3! h5 23 'ii'e6 cJ;;g7 24 'ii'e8 'i'xb2 25 ~e7+ Wh6 26 h4! winning.

c2) White is also clearly bet­ter after 15 ... ~f8?! 16 a4! c4+ 17 Wh1 ttJd3 18 .Jtg3 ttJxb2 19 'ii'c2 ~a5 20 ttJd5 a6 21 .tel c3 22 .Jtxc3 J..xc3 23 ttJxc3 axb5 24 'iVxb2 b4 25 ttJe4.

c3) Black again runs into problems after 15 ... a6?! 16 .i.xe5 .Jtxe5 (16 ... axb5? 17 ttJd5) 17 .Jtxd7 .Jtxd7 18 ttJd5.

c4) 15 ... c4+! 16 cJ;;hl ttJd3 17 J.xc4! ttJxf4 18 ttJg5

18 J..xf7+ was a very inter­esting alternative: 18 ... cJ;;xf7 19 ttJg5+ Wg8 20 :lxf4 liJf6! (The only defence, but one which allows Black to hold a dra,w. Others are weaker: 20 ... h6? 21 ttJd5! ~c5 [21...~xd6 22 'i'b3!]

Page 38: Beating King's Indian and Benoni

36 Main Line with 9 .. J:te8

22 !leI !lel+ 23 'i'xel 'i'xd5 24 'i'e8+ ttJf8 25 !lxf8+ ~xf8 26 'i'xg6+ ~g7 27 'i'h7+ ~f8 28 !lf1+ winning; 20 ... ttJc5?! 21 ttJa4! ttJxa4 22 'i'd5+ ~e6 23 ttJxe6 'i'c6 [23 ... ~xb2? 24 ttJg5+ ~h8 25 'i'f7 ~g7 26 !lh4+-] 24 'i'xc6 bxc6 25 ttJxg7 with an advantage in the end­game; and 20 ... ttJf8?! 21 'i'd5+! ..te6 22 !lxf8+ !lxf8 23 'i'xe6+ ~h8 24 ttJf7+ !lxf7 25 'i'xf7 with a slight plus for White, for example 25 ... 'i'xb2 26 !ldl! :1f8 27 'i'e7 'i'b4 28 h3 and the pawn on d6 is very annoying) 21 ttJd5! (21 !lxf6? doesn't work due to 21...~xf6 22 'i'd5+ ~e6! 23 ttJxe6 'i'xb2! 24 !lfl iNxc3 and White's initiative doesn't compensate for the material losses) 21...'i'xb2 22 ttJe7+ (In the case of 22 !lb1 ttJxd5!? 23 !lxb2 ttJxf4 24 'i'b3+ ~e6 25 ttJxe6 ~h8! 26 !lf2 !lxe6 Black has more than enough compensation for a queen) 22 ... !lxe7 23 dxe7 ~d7 24 !lb 1 (Black is better after 24 !ld4? ~c6! 25 !ld2 'i'b5 26 !ld8+ ttJe8) 24 ... iNe5! 25 !ld4! 'i'xe7 (25 ... 'i'xg5? 26 !lxd7 !le8 27 'i'b3+ ~h8 28 'i'f7 'i'f5 29 !lbd1 'i'b5 30 !ld8 is win­ning for White) 26 'i'b3+ ~h8 (26 ... ~f8?! is dangerous for Black: 27 !lxd7! 'i'xd7 28 ttJe6+ ~e7 29 ttJxg7 !lc8 30 !ld1 ttJd5! 31 'i'f3! with an at­tack) 27 ttJf7+ ~g8 28 ttJg5+ with a draw by a perpetual check.

18 ... ttJe5 ! White has a strong attack af­

ter either 18 ... ttJe6? ! 19 !lxf7 ttJxg5 20 !le7+ ~f8 21 'iVd5 or 18 ... 'i'xb2 19 !lxf4! 'i'xc3 20 ~xf7+ ~h8 21 !lh4.

19 !lxf4 ttJxc4 A very complicated position

arises after 19 ... 'i'xb2 20 !lxf7! ~h8!

20 !lxc4 b6' Now the best was 21 ttJd5!

'i'xd6 22 ttJc7 'i'xd1 + 23 !lxd1 hxg5 24 ttJxe8 ~e6 25 ttJc7 !lc8 26 ttJxe6 !lxc4 27 !ld8+ ~h7 28 ttJxg5+ ~h6 29 ttJxf7+ with an ending which is slightly more pleasant for White.

15 ~xf3 ttJe5 15 ... ~d4+?! 16 ~h1 ttJe5 is

less precise because of 17 ttJb5! (17 ~d5!?) 17 ... ttJxf3 18 'i'xf3 ~f5 (18 ... ~d7 19 ttJxd4 cxd4 20 ~g5!±) 19 ttJc7 ~xb2 20 ttJxe8 ~xa1 21 ttJc7 with the advantage.

16 ttJb5!? This is stronger than 16

~d5?! ~e6 17 ttJe4 c4! with

Page 39: Beating King's Indian and Benoni

good counterplay. 16 l:if8

The alternatives are unpleas­ant for Black:

a) 16 ... ctJxf3+?! 17 'iVxf3l:if8 18 ctJc7 l:ib8 19 l:iael, Destre­becq-Garabedjan, St Etienne 1980, transposes to the position considered below with a rook on f1 instead of aI, a difference that favours White.

b) White is better after 16 ... c4?! 17 ctJc7 ctJd3 18 ctJxe8 'iVxe8 19 ~g3.

c) Black cannot be satisfied with 16 ... ~f5?! 17 ctJc7 ctJxf3+ (l7 ... ctJd3!?) 18 'iVxf3 il.xb2 19 ctJxe8 il.xal 20 ctJc7 .id4+ 21 ~e3 l:ic8 22 Jtxd4 cxd4 (22 ... 'iVxd6 23 ctJb5 'iVb6 24 ~al 'iVxb5 25 'iVc3+-) 23 'iVf4 and White has a large advan­tage, Szabo-Zuckennann, Las Vegas 1973.

17 18

ctJc7 l:iel

l:ib8 ctJc6!

The most accurate move. Black can also try:

a) 18 ... ctJxf3+?! 19 'iVxf3 'iVf6

Main Line with 9 .. J::te8 37

20 ctJe8! 'iVd4+ 21 ~e3 'iVxb2 22 l:iabl 'iVc3 23 ctJxg7 'iVxg7, Balogh-Ribli, Budapest 1972, when despite Black's extra pawn, White is slightly better.

b) A very unusual and un­clear position arises after

18."tLlc4!? 19 b3!? .id4+ 20 'iVxd4 cxd4 21 bxc4

Who is better? A very good question! Now let us return to 18 ... tLlc6!

19 l:ie7! 20 ~xc6 21 'iVel

iLxb2 bxc6 .ifS?!

This is the critical moment of the game:

a) After 21...i..xal? White

Page 40: Beating King's Indian and Benoni

38 Main Line with 9 .. Jie8

wins, e.g. 22 ~xal f6 (22 ... .tf5 endgame after 28 .:txg6+?! 23 ~f6!+-) 23 a3! .tf5 24 .:txf6 29 ~xf6+ ~xf6 30 .:txf6 ~a2+ 'iith8 25 .th6 .:tg8 26 ~f7 is certainly not better for White . .:tbl+ 27 'iitf2 .:tb2+ 28 'iitn .:tb3+ 29 .te3.

b) After 21.. . Ji.d4+ 22 'iitfl! (22 'iithl? .txal 23 ~xal .tf5-+) 22 ... .:tb2 23 .:tdl .:tf2+ 24 ~xf2 .txf2 25 'iitxf2 White keeps the initiative in a compli­cated position.

c) 21. .. .:tb4!

22 .:tdl .td4+ 23 'iithl ~b2 24 .te5 .tf2

24 ... ~b8? 25 ttJe8! winning. 25 ~a5

25 iNc3? .:tc2 26 iNn .th4 turns the game in Black's fa-

According to Boulard this vour. move allows Black to draw. 25

22.te5 26 Not 22 Ji.h6? ~xd6 23 .txf8 27

'iitxf8 24 .:te8+ 'iitg7 25 .:txc8 28 ~d4+ 26 'iithl .txal 27 ttJe8+ 29 'iith6 28 ~c1 + ~f4 29 ~xal .:tc4 and Black wins.

22 ... .txal 23 ~xal The same result follows 23

.txal ~xd6 24 ttJe8 ~xe7! 25 ~xe7 .:tbl+ 26 'iitf2 .:txal 27 ttJf6+ 'iitg7 28 ttJh5+.

23 ... .tf5 24 ~c3 This is more precise than 23

'iitf2. 24 ... l:te4 25 .th8 f6 26 .txf6

~xd6 27 l:tg7+ 'iith8 28 lH7+ with a perpetual check. The

.txb2

.te5 ~d2 iNf4

Ji.g4 .txdl .tg4 .td4 iNc8

Page 41: Beating King's Indian and Benoni

30 ttJeS? This allows Black to achieve

a theoretical draw by sacrificing his queen. 30 ii.xd4! cxd4 31 h3 would have retained some win­ning chances.

30 ... 'WixeS 31l:i.xeS :'xeS 32 . .txd4 exd4 33 h3 ~e6 34 'Wixd4 Ji,xa2 35 'Wixa7 ~e6 36 Wife7 g5 37 d7 Ji,xd7 38 'Wixd7 1:tel+ 39 ~h2 h6 40 'Wixe6 lie6 41 'Wid5 ~g7 42 ~g3l:i.g6 43 ~g4 l:i.e6 44 'Wi d4+ 1/2_1/2

GameS Vaisser-Ibragimov

Bern 1992

1 d4 ttJf6 2 e4 g6 3 ttJe3 JLg7 4 e4 d6 5 f4 0-0 6 ttJf3 e5 7 d5 e6 S Ji,e2 exd5 9 exd5 l:i.eS

10 ttJd2

If White is not ready for the enormous complications that arise after 10 eS then he can opt for the quieter but no less dan­gerous 10 ttJd2. White's overall plan is first to finish his deve1-

Main Line with 9 .. .lle8 39

opment and then to prepare e4-eS.

10 ttJa6 Black has no less than five

plausible alternatives: a) The rare 1O .. .ltJg4?! does

not seem good enough for equality after 11 Ji,xg4 and now:

al) 11...Ji,xc3?! 12 bxc3 'Wih4+ 13 g3 'Wixg4 14 'i'xg4 ii.xg4 IS ~f2

Not IS lib 1 ?! ctJd7! IS ... 5 16 ~el ctJd7 17 c4

'!i.e7

and now instead of 18 lib 1 ?! ~ae8 19 ~xb7 fxe4 20 l:i.e3 (20 ttJfl ttJeS!) 20 ... Ji,fS with equality, Christiansen -Ghitescu, Thessaloniki Olympiad 1984, it would have been better to play 18 iLb2! l:i.ae8 19 ~e3 fxe4 20 l:i.ae 1 Ji,fS (White is winning after 20 ... ctJb6? 21 ttJxe4 ctJxc4 22 ttJf6+ ~f7 23 ctJxe8 ~xe3 24 ttJxd6+ ctJxd6 2S ~xe3) 21 g4! iLxg4 22 ctJxe4 'it>f8 23 ctJxd6 I:f.xe3 24 l:i.xe3 ~xe3 2S 'it>xe3 and White has an advantage in the ending.

a2) 11...'Wih4+ 12 g3 'Wixg4

Page 42: Beating King's Indian and Benoni

40 Main Line with 9 .. Jle8

13 "iYxg4 il.xg4 14 CDbS (14 ~f2!?) 14 ... 4Ja6! IS CDxd6 (After IS h3?! llxe4+! 16 CDxe4 il.f3 17 0-0 il.xe4 18 CDc3 iLd3 19 llf3 4Jb4 20 il.d2 ~d4+ 21 ~h2 iLc4 Black has more than enough compensation, Pan­chius-Liberzon, Israel 1983) IS ... 4Jb4 and in this position White has tried:

a21) 16 4Jxe8?! CDc2+ 17 ~fl iLh3+ 18 ~e2 llxe8 19 llb 1 and Black has an initiative that guarantees at least a per­petual, e.g. 19 ... iLg4+ 20 ~fl iLh3+ 21 ~e2 iLg4+ 22 ~d3 4Jb4+ 23 ~c4 iLe2+ 24 ~b3 iLd3 2S llal iLc2+ (2S ... 4Jc2!? 26 llb 1 4Jd4+ 27 ~c3 ~xb 1 28 4Jxbl11xe4+) 26 ~c4 ~d3+.

a22) 16 O-O?! 4Jc2 17 4Jxe8 ~d4+ 18 ~hl llxe8 19 llbl ~e2 20 eS, Peicheva Juergens­Nicki, Dortmund 1993, and now Black could have obtained a good position with 20 ... iLxfl 21 4Jxfl lld8 22 d6 f6.

a23) I recommend the im­provement 16 h3!, which gives an advantage to White.

b) The inclusion of the moves 1O ... a6?! 11 a4 weakens the important b6-

square and cannot be recom­mended for Black. If Black forces the sequence ... a7-a6, a2-a4 early in the Classical lines of the Modem Benoni then the best solution for White is to transpose to this line of Four Pawns Attack. Take a look at the way I was trapped by Yr-

jola, Sochi 1994: 1 d4 CDf6 2 c4 e6 3 CDf3 cS 4 dS exdS S cxdS d6 6 CDc3 a6?! 7 a4 g6 8 4Jd2! CDbd7 9 e4 ~g7 10 ~e2 0-0 11 0-0 lle8 12 f4! and we arrived at the position of variation b2.

bl) 11...4Jg4 12 CDc4! (This is better than 12 iLxg4?! 'ii'h4+ 13 g3 'ii'xg4 14 'ii'xg4 iLxg4 IS ~f2 ~d4+ 16 ~g2 CDd7 17 h3 ~xc3 18 bxc3 iLe2 19 lle 1 ~d3 20 lle3 c4 21 iLa3. 4Jb6! 22 ~xd6 CDxdS 23 llxd3 cxd3 24 c4. Boleslavsky considered this position as unclear, but the game Otero-Miguel, La Coruna 1993, proved that only White has any problems: 24 ... CDc3!? 2S ~f3 bS! 26 axbS axbS 27 llxa8 llxa8+) 12 .. .fS! (It is too dangerous to play 12 ... iLxc3+? 13 bxc3 llxe4 14 0-0 fS IS as CDf6 16 CDb6 lla7 17 iLf3 lle8 18 c4 CDbd7 1911bl± Fridstein­Landgraf, corr. 1967) 13 iLxg4 fxg4 14 eS dxeS IS 0-0 exf4 (Or IS ... e4 16 iLe3 b6 17 as bS 18 d6 iLe6 19 CDb611a7 20 iLxcS;l;) 16 iLxf4 iLd4+ 17 ~hl and White is slightly better, Schmidt -Aepfler, Germany 1994.

b2) 11. .. 4Jbd7 120-0 and now: b21) After the slow

12 ... 11b8?! 13 ~hl 'ii'c7 (l3 ... c4 doesn't help: 14 eS! dxeS IS 4Jxc4 bS!? 16 axbS axbS, Lar­sen-Ljubojevic, Milan 1975, and now 17 CDd6! b4 18 4Jxe8 'ii'xe8 19 4JbS would have as­sured White of an advantage)

Page 43: Beating King's Indian and Benoni

14 ':'a2 b6 (Even worse is 14 ... c4?! IS eS! dxeS 16 CLJxc4 e4 17 fS! ':'a8 18 i.f4± Toth­Nunn, Reggio Emilia 1983) IS b3 CLJf8 (lS ... CLJg4? 16 CLJdbl), Izeta-Femandez, Spanish Ch. 1987, White could have ob­tained a clear advantage by playing 16 i.d3.

b22) 12 ... c4! 13 ~hl! CLJcS 14 eS dxeS IS fxeS

IS CLJxc4!? is not a bad move either.

IS ... .:.xeS Or IS ... CLJfd7 16 e6! fxe6 17

CLJxc4 with better chances for White.

16 CLJxc4

16 ... .:.e8 If 16 ... .:.fS 17 i.f4! is to

White's profit, e.g. 17 ... CLJfe4 18 CLJxe4 CLJxe4 19 "iVc2 CLJgS 20 as ':'xdS 21 CLJb6± Farago-Bistric, Sarajevo 1983; 17 ... i.d7 18 CLJe3 ':'xf4 19 ':'xf4 "iVe8 20 CLJc4 CLJxa4 21 CLJxa4 .i.xa4 22 ':'xa4 bS, Csonkics-Chelushkina, Sub­otica Interzonal 1991, and now the simplest was 23 CLJd6! "iVb8 24 CLJxbS winning; and 17 ... gS 18 i.e3 ':'xfl + 19 "iVxfl CLJfe4

Main Line with 9 .. J:te8 41

20 CLJxe4 CLJxe4 21 i.d3 CLJd6 22 .i.b6± Lukacs-Karlsson, Hel­sinki 1983.

In Glek-Tseshkovsky, Buda­pest 1989, Black tried to resolve the problems of this difficult position with an exchange sacri­fice: 16 ... .:.xe2!? 17 CLJxe2 i.g4 18 CLJeS CLJce4, and now instead of 19 i.f4 White could have played 19 CLJxg4 CLJxg4 20 'ilb3! with the better chances, e.g. 20 ... "iVh4 21 i.f4 gS 22 Si.g3 CLJxg3+ 23 "iVxg3 "iVxg3 24 CLJxg3±.

17 i.gS h6 18 i.h4 CLJce4 Now White has a pleasant

choice between 19 CLJxe4 ':'xe4 20 i.g3 CLJxdS? (20 ... hS!? Yr­jola) 21 ':'xf7! i.e6 22 lixb7 ~h8 23 'ii'c2, Yrjola-Vaisser, Sochi 1984, and 19 d6 gS (19 ... CLJxc3!? 20 bxc3 i.e6) 20 i.e 1 i.e6 21 CLJxe4 CLJxe4 22 i.aS as in the game Beliavsky­Velimirovic, Moscow Inter­zonal 1982, in both cases with a clear advantage.

c) 10 ... b6?! 11 0-0 i.a6 12 a4! i.xe2 13 "iVxe2

This rare line is similar to 11 CLJd2 in the system 9 ... it..g4. The difference is that here Black cannot meet the White move CLJc4 with ... CLJb6 and after, for example, 13 ... a6 14 CLJc4 ':'a7 IS 'iVf3 ':'d7! 16 i.d2 bS 17 axbS axbS 18 CLJxbS CLJxe4 19 i.aS "iVe7 20 fS White is better, A. Zaitsev-Zhuravlev, USSR 1965.

d) 1O .. .'~Jbd7 11 0-0 c4

Page 44: Beating King's Indian and Benoni

42 Main Line with 9 ... .l:i.e8

and now: dl) Black has easy play after

12 ~xc4?! ttJc5 13 e5 dxe5 14 fxe5 .l:.xe5 15 ttJf3 .l:.e8 16 ~hl a6 (The immediate 16 ... ttJfe4!? 17 ttJxe4 ttJxe4 deserves atten­tion) 17 a4 ttJce4 18 ttJxe4 ttJxe4 19 ~b3 ttJd6, Toth­Velimirovic, Budva 1981.

d2) 12 e5?! is premature due to 12 ... dxe5 13 ttJxc4 ttJb6!, when Black has more than com­fortable equality after 14 fxe5 ttJfxd5 15 ttJxd5 (15 ttJd6 ttJxc3 16 bxc3 is slightly better for Black because of 16 ... .l:.f8! 17 ttJxf7 ~xdl 18 ~xdl ..te6 19 ttJh6+ ~h8 20 .l:.el .l:.ae8, as in Kurtenkov-Ghinda, Primorsko 1985) 15 ... ~xd5 16 ~xd5 (16 ttJd6 ~c5+ 17 ~hl .l:.e7 is no better) 16 ... ttJxd5 17 ~f3, Si­nes-Ljuboevic, Yugoslav Ch. 1982, and now 17 ... ttJb6!?

d3) 12 ~h 1 ttJc5 13 e5 dxe5 14 fxe5

After 14 ttJxc4 Black's best is 14 ... exf4! 15 ..txf4 ttJce4 16 ~f3 ttJxc3 17 bxc3 ttJe4 18 ~b3 b6! 19 ~xe4 (Not 19 d6?! ~a6 20 ~xe4 .l:.xe4 21 ttJd2

.l:.e8 22 .l:.fel ~d7 23 ttJf3 ~b7+ Dlugy-Vaisser, Havana 1985) 19 ... .l:.xe4 20 ttJd2 .l:.e8 21 .l:.ael ~f5 22 c4 ~d7 23 ttJf3 .l:.xe 1 24 .l:.xe 1 .l:.e8 25 ~e5 ~g4 26 ..txg7 ~xf3 27 ~e5 ~f5 with equality, Meduna­Poloch, Ceske Budejovice open 1995.

14 .l:.xe5 15 ttJxc4 M.e.8 15 ... .l:.f5?! 16 ~f4! (16 ~f3

b6! 17 d6, Haba-Renet, Thes­saloniki Olympiad 1988, and now 17 ... .l:.b8 would have given Black enough play) 16 ... ttJh5 (16 ... ttJxd5?! 17 ttJe3!±; 16 ... g5 17 ~e5! ttJfe4 [Not 17 ... b5?! 18 ttJxb5 ttJce4 19 ~d4 ~xd5 20 ~f3 ~e6 21 b3± Pigusov­Chekhov, Irkutsk 1983] 18 ..txg7 ~xf1 + 19 ~xf1 ~xg7 20 ttJxe4 ttJxe4 21 .l:.dl;t) 17 ~e3 .l:.xf1+ (l7 ... b6!?) 18 ~xf1 b6 (Or 18 ... .txc3 19 bxc3 ttJe4 20 ~xh5 ~h4 21 ~f3 gxh5 22 ~d4± Kindermann-Danner, Zu­rich 1994) 19 .l:.dl, Pekarek­Suba, Warsaw 1987, and now after the best move 19 ... ~f6 White gets an advantage, for example, by playing 20 ~xc5 bxc5 21 ~xh5.

16 ..tg5 h6 17 ~h4 ttJce4 18 tt:Jxe4

18 d6 is also interesting. 18 .. .lixe4 19 ~g3 ~xd5! Not 19 ... ttJxd5? 20 .l:.xf7!

.te6 21 .l:.xb7± Tasic-Murey, Cannes 1992.

20 ~xd5 ttJxd5 21 ~f3 . .l:.d4 22 .l:.ad 1 ttJb6

It is worse for Black to play

Page 45: Beating King's Indian and Benoni

22. Ji:Je7? 23 ':'xd4 ~xd4 24 ':'d 1 ttJf5 25 ~e5 iLxe5 26 .l::id8+ cJ;;g7 27 ttJxe5 cJ;;f6 28 .l::ie8, when the pin of the c8-bishop assures an advantage for White, Vaisser-Andres, Bayamo 1985.

23 .l::ixd4 ~xd4 24 ':'dl ~f6 25 b3 ttJxc4 26 bxc4 a5

White has more than enough pressure for a pawn, but the players agreed a draw in Chan­dler-Sax, Sarajevo 1985.

If White is not satisfied with this variation he can try:

d4) 12 a4 ttJc5 This line is particularly im­

portant because Black can prac­tically force this position with another move order, namely 1O ... c4 11 a4 ttJa6 12 0-0 ttJc5. Now White has two possibili­ties:

d41) 13 ~f3 b6 13 ... ~h6?! is suspicious.

White can play 14 e5 dxe5 15 fxe5 ':'xe5 16 ttJxc4 ~xc1 17 ':'xc1 ~e8, Zakharevich-Petru­shin, Azov 1993, and now 18 b4 with an advantage. The quiet 14 'iYc2 may be even stronger. The game Padevsky-Ciocaltea, Ha­vana Olympiad 1966, continued 14 ... ttJd3 15 ttJxc4 ttJxc1 16 'iYxc1 ~d7 (If 16 ... ~g4 White has a strong reply: 17 'i'dl! ':'c8 18 b3 ~xf3 19 'i'xf3 with a great advantage to White, Uh­mann-Felix, Karvina 1989) 17 b3 and White is clearly better.

The game Gutman-Petkevich, USSR 1967, saw instead

Main Line with 9 ... .l:te8 43

13 ... ~d7 14 e5 dxe5 15 fxe5 ':'xe5 16 ttJxc4 ':'e8, and now 17 ~f4 ttJfe4 (17 ... ~f5?! 18 ttJd6) 18 ttJxe4 ttJxe4 19 .l::ic 1 would have been slightly better for White.

14 e5 dxe5 15 fxe5 ':'xe5 16 ttJxc4 ':'e8

This is safer than 16 ... ':'f5 17 d6 ':'b8 18 ttJb5! ~e6 19 ttJe3 (19 'i'c2!? is not bad) 19 ... ~b3, Zaharevich-Gleizerov, USSR 1987, and now following analy­sis by Gleizerov, 20 ttJxf5! ~xdl 21 ttJe7+ cJ;;h8 22 .l::ixdl (22 ttJc6?! .txf3 23 ttJxd8 ~d5!+) with good compensa­tion for the queen.

lLd6 17 ~g5 is not dangerous for

Black. After 17 ... .tf5! 18 b4?! ~d3 the game turns in his fa­vour.

17 ... .te6! Worse is 17 ... .ta6?! because

of 18 ttJb5!

In this critical position 18 .txa8? does not work due to 18 ... .txc4 19 .tc6 l'1e6! with good chances for Black.

Against Granda Zuniga, Bue-

Page 46: Beating King's Indian and Benoni

44 Main Line with 9 .. Jle8

nos Aires 1992, Chemin con­tinued 18 ~e2 and now, as he noted, Black could have ob­tained sufficient counterplay with 18 ... LiJh5!? 19 g4 iVh4, threatening 20 ... LiJg3.

Perhaps White should try 18 LiJd5!? with some hopes for an advantage.

d42) 13 e5!? dxe5 14 LiJxc4 e4 (14 ... exf4 15 ..txf4 LiJce4 is also playable) 15 .i.e3 ttJd3 16 ..txd3 exd3 17 iV xd3 ..tf5 18 'li'd2 ~c8 19 b3 ttJg4! (Only the nice combination that starts with this move gives Black hopes of equality. In the case of 19 ... ttJe4?! 20 LiJxe4 ..txal 21 ttJed6 ..tc3, Balashov-Dvor­etsky, Moscow 1967, 22 iVc1! would have given White the better chances) 20 ..td4 ..txd4+ 21 'li'xd4 ~xc4! 22 bxc4 ~e3 23 h3 ~d3 24 iVxa7 iVh4 2S ~a2 ttJe3 26 'li'b8+ cJ;g7 27 iVeS+, Pomar-Toran, Palma de Mal­lorca 1966, and now instead of 27 ... f6? 28 iVe7+ cJ;h6 29 ttJe4 with a large advantage for White, Black should have played 27 ... cJ;h6! 28 ttJe4 (Or 28 ~c1?! ttJxg2! with a strong attack) 28 ... ttJxfl 29 ttJf6 ttJe3 30 ttJg8+ with a perpetual check (Dvoretsky). I would not be surprised if White can improve his play in this line.

e) 1O ... c4!? After 11 a4 (11 0-0 bS!) 11...ttJa6 12 0-0 LiJc5 play transposes to variation d4.

11 0-0 ~b8!? A clever, flexible move. At

the moment the black knight on a6 keeps its options open of hopping to either b4 or c7. It is less precise to play 11...c4?!, because of 12 eS dxeS 13 LiJxc4 and White has saved an impor­tant tempo when compared to line d2 above, by missing out 'it>hl. The game Hvenekilde-T. Horvath, Copenhagen 1983, continued 13 ... e4 14 ..te3 LiJb4! IS d6 .i.e6 16 ttJbS! ttJd3 17 ..txd3 exd3, and now White should have simply played 18 iVxd3 and if 18 ... a6 then 19 ttJc7! LiJxc7 20 ..tb6 with an advantage.

Another reasonable line is 11...ttJc7 12 a4

After 12 .i.f3 ~b8 13 ttJc4 b5 14 ttJxd6 (14 ttJaS?! ..td7 IS eS dxe5 16 fxe5 ~xeS 17 ..tf4 I:i.fS 18 ..tg3 b4=+) 14 ... iVxd6 IS eS iV d8! 16 d6 ttJe6 17 exf6 ..txf6 18 .tc6 .i.d7 19 .i.xd7 .td4+! 20 'it>hl 'li'xd7 21 fS gxfS 22 ttJdS 'it>h8 23 ttJe7 iVxd6 24 ttJxfS, Zilberman-Rechlis, Is­raeli Ch. 1994, White has a certain initiative for the pawn, but not more.

Page 47: Beating King's Indian and Benoni

12 ... b6 12 ... a6 13 j,f3 ':'b8 (l3 ... b5?!

doesn't work without prepara­tion: 14 axb5 axb5 15 ':'xa8 liJxa8 16 liJxb5 j,a6, Bronznik­Gazik, Bratislava 1992, and now White should have played 17 'iVa4! 'iVb6 18 ttJxd6 'iVxd6 19 e5 'iVb6 20 exf6 j, b5 21 liJc4 with a clear advantage) 14 a5! (14 ttJc4 allows the possibility of an interesting sacrifice: 14 ... b5 15 axb5 axb5 16 liJa5 .i.d7 17 e5 dxe5 18 d6 [18 fxe5 ':'xe5 19 liJb7!? ':'xb7 20 d6, suggested by Giffard, needs further analysis] 18 ... e4! 19 dxc7 'iVxc7 20 j,e2 c4 with full compensation for the piece, Kolbaek-Andersen, Vejle 1967) 14 ... il.d7 15 ':'el il.b5, Para­mos-Marin, Ibercaja open 1994, and now 16 liJdb 1 would have been slightly better for White.

13 iLf3 J:.b8 14 ':'el j,a6 After 14 ... a6?! 15 liJc4 b5 16

liJxd6 'iVxd6 17 e5 'iV d8 18 d6 ttJe6 19 axb5 axb5 20 exf6 .i.xf6 we obtain the position from the game Zilberman­Rechlis above with some im­portant differences: the white rook is on el and the a-file is open. This changes the evalua­tion to White's profit. The game Kozul-Marovic, Toronto 1990, continued 21ttJd5 il.d4+?! (After the best move 21...j,b7!? 22 f5!? il.xd5 23 'iVxd5 liJd4! 24 i.f4 White also keeps an edge, according to Kozul) 22 il.e3! i.b7 23 .i.xd4 .i.xd5 24

Main Line with 9 ... l:!.e8 45

.i.e5 i.xf3 25 'iVxf3 with a clear advantage for White.

An interesting alternative is the active 14 ... h5!?, as played in the game Rogers-Kristiansen, Thessaloniki Olympiad 1984, which continued 15 liJc4 (15 h3 liJg4 is not as dangerous for White as some commentators have claimed. After 16 eS! liJh6 17 liJc4 liJf5 with complicated play, I prefer White). Now in­stead of 15 ... i.a6?! 16 liJa3 .i.b7 17 .i.e3 'iVd7 18 h3 a6 19 liJc4 b5 20 liJxd6! ±, Rogers proposed 15 ... liJg4!? 16 j,xg4 .i.xg4 17 'iVd3 .i.c8! with an unclear position.

lSlb.dbl This is better than 15 liJfl

.i.xfll6 ':'xfl a6 17 WhlliJd7, Begovac-Wojtkiewicz, Bern 1991, when Black was without problems. 15 ... liJd7 16 liJa3 c4 17 liJcbS! .i.xb5 18 axb5 c3 19 ':'bl cxb2 20 il.xb2 i.xb2 21 ':'xb2 'iVf6

The game Malich-Tringov, Sarajevo 1965, arrived at this position. Now White should have played 22 'iid2, followed

Page 48: Beating King's Indian and Benoni

46 Main Line with 9 ... J::.e8

by 'bc4 and eventually e4-e5, with the better chances. Note that the capture 22 ... 'bxd5? fails to 23 e5!

12 Whl? Suddenly this lost tempo

turns the game in Black's fa­vour. Instead, 12 !tel is a more useful waiting move. The game might continue in the spirit of the previous line 11.. .'bc7, when the line recommended in ECD needs more tests: 12 !tb1 ~d7 (If 12 ... 'bc7?!, then 12 b4 or 12 a3 b5 13 b4, blocking the queenside) 13 e5 dxe5 14 fxe5 !txe5 15 'bc4, followed by 16 ~f4 with good compensation for a pawn.

12 13 a4

'be7 a6

Mikhail Gurevich demon­strated an interesting idea after Smirin's suspicious 13 ... h5?! (USSR Ch. 1988): 14 f5! a6 15 a5 gxf5 15 ii.xh5 'bb5 17 exf5 !te5 18 ~f3 ~xf5 19 'bc4 VjJfc7 20 ~d2 and White stood better.

14 as .td7

15 e5 The quiet 15 ~f3 is also in­

sufficient for equality: 15 ... 'bb5 (l5 ... ~b5?! 16 !tel ~d3 17 VjJfb3 VjJf d7 18 'bc4 ~xc4 19 VjJfxc4 'bb5 20 ~e3 'bd4 21 ~xd4 cxd4 22 'ba4 VjJf d8 23 VjJfb4 with an equal position, Arakelian-Ibragimov, Podolsk 1993) 16 e5 (After 16 .l:le1 'bd4 17 'bc4 ii.b5 18 'bb6 'bd7! the position of Black's knight in the centre is very annoying, while 16 'bxb5 ~xb5 17 .l:le1 c4 18 !ta3 !tc8 19'bfl'bd7, Toth-De Firmian, Biel 1986, is no better. Things would tum out quite differently if White had played 12 !tel instead of 12 Whl. Then he could play 16 'bc4 here with an advantage) 16 ... dxe5 17 fxe5 !txe5 18 'bc4 !tf5! 19'be3 !tf4 20 'be2 .l:lh4 21 g3 !te4! 22 ~xe4 'bxe4 with a strong ini­tiative for Black, Ufimtsev-Tal, USSR 1967.

15 dxe5 16 'be4 ~b5!

16 ... e4?! 17 ~e3 is worse.

Page 49: Beating King's Indian and Benoni

17 d6 ttJe6 18 fxe5 ttJd7 19 ~f4 ttJxf4 20 lIxf4 ttJxe5

20 ... i.xc4 looks good, when 21 i.xc4 ttJxe5 gives Black a plus.

21 ttJxb5 axb5 22 ttJb6 e4?!

It would have been better to play 22 ... :e6 23 d7 'iWc7!, threatening 24 ... ttJxd7. 24 i.xb5 does not avoid the threat: 24 ... ttJxd7! 25 ttJd5 "iYe5! and Black exchanges the dangerous d6-pawn while keeping his ex­tra pawn.

23 'iWd5 .l::te6 24 .l::td1 ttJd3 25 i.xd3 exd3?

A decisive mistake. Why not 25 ... .l::txd6 instead? After the queen sacrifice 26 .1t.xc4!? .l::txd5 27 i.xd5 ~h8 28 .l::txf7 the position is unclear.

26 d7 Black has nothing with which

to oppose the terrible passed d­pawn.

Main Line with 9 ... :e8 47

26 27 28 29

.l::te4 'iWxd3 "iYxb5

i.e5 'iWf6 .l::td8 i.e7?

A blunder in a lost position. 30 ttJd5 'ii'f5 31 .l::txe6 1-0

This game illustrates a situa­tion where White made a mis­take in the opening and found himself in an inferior version of the typical Four Pawns Attack middlegame. As you can see, if you have a good feel for this type of position and your oppo­nent does not, you still have chances, even if the opening goes slightly awry.

At the present time the main line 9 .. J~e8 10 e5 dxe5 11 fxe5 ttJg4 12 i.g5 "iYb6 13 0-0 leads logically to a more or less forced draw. White players hop­ing for more should consider 10 ttJd2 (Game 5), 12 0-0 ttJxe5 13 i.f4 (Game 4), 12 e6 (Game 4, pages 32-33) and 13 'ii'd2 (Game 1, pages 10-12).

Page 50: Beating King's Indian and Benoni

2 Main Line with 9 ... ..tg4

1 d4 ttJf6 2 c4 g6 3 ttJc3 i.,g7 4 e4 d6 5 f4 0-0 6 ttJf3 c5 7 d5 e6 8 .te2 exd5 9 cxd5 ~g4

9 ... ~g4 is Black's most solid continuation in the Four Pawns Attack. Once again, White's main idea will be to prepare e4-e5; and Black, whilst trying to hold up this advance, must also try to develop counterplay on the queenside. After an eventual exchange of Black's light­squared bishop on f3, White will also often use his strong centre as the basis for a direct pawn storm on the kingside with, for example g2-g4, h2-h4, and g4-g5.

The first two games in this chapter present the main line of this system. In Game 6, Vais­ser-Yrjola, Black plays for ... b7-b5, while game Vaisser­Berelovich (Game 7) is con­cerned with various plans in­volving ... c5-c4. Sidelines for Black are examined in Game 8, Vaisser-Smirin, while White's sidelines are seen in Game 9, Monin-Shchekachev.

Game 6 Vaisser-Yrjola Helsinki 1991

1 d4 ttJf6 2 c4 g6 3 ttJc3 i.,g7 4 e4 d6 5 f4 0-0 6 ttJf3 c5 7 d5 e6 8 i.,e2 exd5 9 cxd5 ..tg4

10 0-0 11 ~e1 12 h3 13 ~xf3

4Jbd7 ~e8 ~xf3

(see following diagram)

This is the critical position of the 9 ... ~g4 variation.

13 ~a5!

Page 51: Beating King's Indian and Benoni

The point of this move is that White now has no time to play 14 a4?!, preventing 14 ... bS, be-cause of 14 ... c4! IS .ie3 (1S ctJbS a6) IS ... ctJcS 16 .txcS 'i'xcS+ 17 'it>hl ctJd7 with a comfortable position for Black.

13 ... c4 and 13 .. Jk8, prepar-ing ... cS-c4, are considered in the next game. Here we discuss Black's other plans without ... cS-c4:

a) 13 .. J~b8!? 14 a4 hS (14 ... h6!?) IS ~e2 (1S g4!? hxg4 16 hxg4 ctJh7 17 gS is well justified here. White gains a tempo compared to variation b2) IS ... a6 16 as bS 17 axb6 ~xb6 18 'i'd3 'i'c8 19 ~a4 'ilHb7 20 'ilHc4 'i'c8?! 21 'ilHa2 ctJh7 22 .te3 .ih6 23 ctJbl 'ilHb7 24 ctJa3 gS 2S fxgS .ixgS 26 ctJc4± Vaisser-Brito, Las Palmas 1995.

b) 13 ... a6 14 a4 (Note that this move is not automatic: 14 g4!? h6 IS h4 bS 16 gS hxgS 17 hxgS 4:Jh7 also deserves atten­tion) and now:

bI) Black's attempt to attack

Main Line with 9 ... ~g4 49

on the queenside using piece play, without moving pawns, is suspicious: 14 ... 'ilHaS IS .te3 'i'b4 16 'ilHc2 ctJb6?! 17 .tf2 ctJfd7? 18 as ttJc4 19 ttJa2! 'i'bS 20.te2+-.

b2) A typiCal pawn sacrifice occurred in the game Yrjola­Pedzich, Cappelle la Grande 1992: 14 ... ~b8 IS g4 (1S as!?) IS ... h6 16 h4 ctJh7 17 gS c4 18 .ie3 ~c8?! 19 .tg4 ~c7 20 eS! dxeS 21 fS with a strong initia­tive.

b3) Passive play does not promise much: 14 ... ~c8 IS 'it>hl 'i'c7 16 ~e2 'i'b8 17 as ~e7 18 .te3 ~ce8 19 .tf2 .th6 20 'i'c1! bS (20 ... ttJeS 21 fxeS!) 21 axb6 'ilHxb6 22 .ih4 ~b8 23 g4 .ig7 24 'i'e1 ~be8 2S .ig3 h6 26 eS! dxeS 27 fxeS ctJh7 28 e6 fxe6 29 d6 ~t7 30 ctJdS 'i'b3 31 ~a3 'ilHc4 32 ttJe7+ ~fxe7 33 dxe7± Banikas-Arakhamia, Ika­ros open 1995.

b4) 14 ... c4 is discussed in the next game.

14 .te3

Page 52: Beating King's Indian and Benoni

50 Main Line with 9 ... .i.g4

If White wishes to avoid a theoretical battle, he may choose to pay more attention to 14 g4 or to the game Bulthaupt­Lindemann, German Bundes­liga 1996: 14 I:!.e2 b5 15 'iVe1 b4 16 CLJdl c4 17 ~d2 'iVb6+ 18 ~hl CLJc5 19 e5 CLJd3 20 'iVg3 CLJd7? (20 ... dxe5) 21 e6 CLJf6 22 f5 with an initiative.

14 bS The game Vaisser-

Kindermann, Biel 1991, saw instead 14 ... I:!.ac8?! (Black mixes two different plans) 15 g4! h6 16 h4 b5? (16 ... g5? is also bad: 17 hxg5 hxg5 18 e5! dxe5 19 fxg5 CLJh7 20 CLJe4 with a clear edge, but 16 ... CLJb6!?, liberating the d7-square for the king's knight, deserved atten­tion) 17 g5 hxg5 18 hxg5 CLJh7 19 .tg4 I:!.cd8 20 e5! dxe5 21 f5 e4! 22 fxg6 (22 f6? CLJhxf6 23 gxf6 CLJxf6 is in Black's favour) 22 .. .fxg6 23 .te6+ I:!.xe6? (23 ... ~h8 would have been more resilient, but White still has good attacking chances after 24 'iVg4) 24 dxe6 CLJe5 25 e7! I:!.e8 26 'iVd5+ ~h8 27 ~g2 I:!.xe7 28 I:!.hl I:!.f7 29 I:!.afl! I:!.xfl (29 ... 'iVc7 30 'iVa8+ .tf8 31 CLJd5+-) 30 ~xfl 1-0.

IS a3 CLJb6 Black players have also ex­

perimented with other moves in this position, which is important for the evaluation of the whole system:

a) 15 ... I:!.ac8!? 16 g4?! (This move is effective only if Black

cannot comfortably free the d7-square for the knight on f6; and here that is not the case) 16 ... CLJb6 17 e5 CLJc4 18 exf6 CLJxe3 19 I:!.xe3 I:!.xe3 20 fxg7 I:!.ce8 (One might think that, in comparison with the main line, White has an extra tempo, g2-g4, but in reality he has one tempo less, because the f3-bishop is not protected) 21 ~f2 b4! 22 axb4 'iVxb4 23 CLJe2 'iVxb2 24 I:!.xa7 c4! 25 I:!.a6 I:!.d3 26 'iVel I:!.ee3 0-1 Elbilia­Loeffler, Cannes open 1993. Three years later Elbilia im­proved upon his previous play with 16 ~f2! a6?! 17 ~g3 CLJb6 18 e5 dxe5 19 fxe5 CLJfd7, and now the simple 20 e6 would have fixed White's advantage, Elbilia-Grivas, Yerevan Olym­piad 1996.

b) 15 ... I:!.ab8!? 16.tf2 h6 and now, instead of the logical plan 17 I:!.e2 ! ? preparing 'iV eland e4-e5 with good chances, in the game Vaisser-Hemandez, Las Palmas 1995, White missed a tactical nuance: 17 e5? dxe5 18 fxe5 I:!.xe5! 19 I:!.xe5 CLJxe5 20 .te2 (The intended 20 .txc5? runs into 20 ... CLJxf3+ 21 gxf3 [or 21 'iVxf3?? b4 winning a piece] 21...'iVc7 with a strong initiative) 20 ... c4 and White did not have enough compensation for the pawn.

c) 15 ... b4 16 axb4 'iVxb4 17 'iVc2 CLJb6 18 .tf2 CLJfd7 19 I:!.e2 ~d4 20 ~hl .txf2 21 I:!.xf2 c4 22 I:!.e2 a5 23 .tg4 CLJf6 24 'iVd2

Page 53: Beating King's Indian and Benoni

4lxg4 25 hxg4 tiJd7, Lautier­Smirin, Cap d' Agde 1996, and now after 26 l::ra4! (Forcing Black's queen to the c5-square) 26 .. :iVc5 27 g5 White is better.

16 e5

The result of the spectacular game Kozul-Nunn, Wijk an Zee 1991, (negative for White) practically closed discussion about the move 16 .tf2; but in reality matters are not so clear. That game continued 16 .... ctJc4 17 'ii'c2 ctJd7 18 .te2 (18 a4? b4 19 ctJb5 a6!) 18 ... l::rab8 19 a4 b4, and now instead of 20 .txc4? bxc3 21 b3 a6! 22 l::rec 1 ctJb6 23 Jii.fl c4 24 .txc4 ctJxc4 25 bxc4 l::rb2 26 'ii'd3 l::rd2 27 'ii'f3 f5! with a decisive advan­tage for Black, White should have played 20 ctJb5! At the very least, the line proposed by Nunn: 20 ... ctJxb2 21 ctJxd6 b3 22 'ii'b 1 ctJxa4 23 ctJxe8? .txa 1 24 'ii'xal b2 25 'ifa2 'ii'xel +! 26 .txel bl'ii' 'and Black wins' can be improved by 23 l::ra3! 'ifb4 24 l::rxb3 'ii'xb3 25 'ii'xb3

Main Line with 9 ... .i..g4 51

l::rxb3 26 tiJxe8 with an unclear position.

16 ctJc4 16 ... dxe5?! 17 fxe5 l::rxe5 18

.txc5l::rxel+ 19 'ii'xel tiJbd7 20

.td4 is promising for White. 17 exf6 ctJxe3

17 ... $Lxf6? does not work: 18 .td2 tiJxb2 19 tiJe4! .td4+ 20 ~h2 'ii'd8 (Or 20 ... tiJxdl 21 .txa5 .txal 22 l::rxdl .tg7 23 ctJxd6+-) 21 'if c2 tiJc4 22 .tc3 f5 23 ctJg5 l::rxe 1 24 l::rxe 1 ctJxa3 25 'ii'e2! and White is winning.

18 l::rxe3 l::rxe3 19 fxg7 l::rae8 20 f5!

The alternative 20 'ifd2!? (Hoping to provoke 20 .. .f5?! 21 ~f2! l::r3e7 22 g4) worked out well in the game Elbilia-San Marco, Cannes open 1995: 20 ... 'ii'b6?! (Better is 20 ... c4!?) 21 f5 c4 22 ~hl gxf5 23 l::rfl l::r8e5 24 'iff2 b4 25 axb4 'ii'xb4 26 .te4! l::r3xe4 27 ctJxe4 l::rxe4 28 'ii'xf5 l::re7 29 'ii'f6 with a large advantage for White.

20 gxf5 In the game Dearing-Moss,

Hastings 1996, Black tried 20 ... b4 21 axb4 'ii'xb4 22 'ii'd2 'iff4?! (22 ... 'ifh4!?) 23 l::rfl 'ifxf5? and now 24 .te4! 'ii'g5 25 l::rf3 wins for White.

In my game with Kruger (San Bernardino 1989) Black played the unfortunate 20 ... 'ii'd8 21 'ii'd2 'ifh4? (21...a6!) which led to a clear advantage for White after 22 ctJxb5 'if g3 23 l::rfl.

21 'ifd2

Page 54: Beating King's Indian and Benoni

52 Main Line with 9 ... i.g4

21 b4?! Black's overly-aggressive

pawn move gives White a clear advantage. The critical position for this line arises after 21...a6!? 22 1:.f1 ~d8 (22 ... c4 23 ~d4) 23 tDdl!? (23 .i.dl ~g5 24 ~f2 1:.8e5 25 i..c2? b4) 23 ... 1:.3e5 24 ~h6. I believe that White's chances are better here.

22 tDe2 c4 23 tDg3 c3 24 bxc3 bxc3 25 ~c2 ~b6 26 'i1t>h2 ~b2 27 1:.a2! ~b6?

The endgame after 27 ... ~xc2 is not exactly a dream for Black, but it was still better than the text move.

28 tDxf5 29 tDh6+ 30 tDg4 31 tDxe5 32 g3 33 ~d3 34 1:.e2 35 1:.12 36 ~dl

1-0

1:.3e5 'i1t>xg7 ~d4 ~xe5+ 1:.c8 f5 ~f6 1:.b8 ~e5

We now look at Black's plans involving ... c5-c4 in the main line.

Game 7 Vaisser-Berelovich

Groningen 1993

1 d4 tDf6 2 c4 g6 3 tDc3 ~g7 4 e4 d6 5 f4 0-0 6 tDf3 c5 7 d5 e6 8 ~e2 exd5 9 cxd5 .i.g4 100-0 tDbd7 11 1:.e1 1:.e8 12 h3 ~xf3 13 i.xf3

13 1:.c8 Instead: a) First we consider Black's

play without 13 ... a6 14 a4. This gives White some additional possibilities and so cannot be recommended.

13 ... c4?! 14 i.e3 ~a5 14 ... a6? is too slow because

of 15 .i.d4! and the immediate 16 e5!

15 ~d4! After 15 'i1t>hl tDc5 16 i..xc5

Page 55: Beating King's Indian and Benoni

'iVxcs 17 eS the resulting posi­tions are similar to those in which the moves 13 ... a6 14 a4 are inserted. Nevertheless, we should note one particularity to this move order: instead of IS .. .'~JCS Black can play IS ... .l::te7!?, and after 16 .td4 CLJcS (Or 16 ... .l::tae8 17 .l::te2! bS 18 a3 a6 19 'iVgl 'iVc7 20 .l::tael with a pleasant position for White in Komarov-Niermann, Cattolica open 1993) 17 b4! cxb3 18 axb3 'iVb4 19 CLJa2 'iVbS 20 eS (20 .te2!?) and now not 20 ... CLJe8? 21 .l::te3± Glek-Koro­lev, corr. 1988, but the more resilient 20 ... dxeS! 21 fxeS CLJfd7. There is no direct refuta­tion: 22 b4 CLJd3 23 .te2 .ixeS (23 ... .l::txeS) is not good for White.

IS ... CLJcS If lS ... .l::te7 16 b4! is still pos­

sible. llih4l The most energetic way to re­

fute Black's set-up.

Neither of the responses con­sidered below promise Black an easy life:

Main Line with 9 ... .iLg4 53

al) 16 ... 'iVxb4?! 17 .l::tbl 'iVaS 18 .l::tbS 'iVa6 19 .txcS CLJxe4! (Much better than 19 ... dxcS? 20 eS CLJd7 21 d6 .l::tab8 22 a4 b6 23 CLJdS and Black is suffocating, Komarov-Strovsky, Belfort 1992) 20 CLJxe4 ~xbS 21 ..tf2! (21 ..txd6? .l::txe4 22 .l::txe4 'iVb6+=) 21... ~ a6 22 .l::te2 and White still has the better chances.

a2) 16 ... cxb3 17 axb3 ~b4 18 CLJa2 ~bS

Or 18 ... 'iVxb3 19 .ixcs ~xd1 20 .l::texdl dxcS 21 d6 .l::tab8 22 CLJc3 a6 23 .l::tdbl:t.

19..te2!

Instead, after the repetition of moves 19 CLJc3 ~b4 20 CLJa2 'iVbS, Kouatly and A. Kuzmin agreed a draw (Doha 1993). Let us try to continue the analysis.

19 ... ~xb3 After 19 ... ~d7!? 20 i.xcs

dxcS 21 ..tf3 CLJhS 22 i.xhS ..txal 23 ~xal gxhS 24 CLJc3 White's attack is going to be very dangerous.

20 .ixcs 'iV g3 Or 20 ... ~xdl 21 .l::taxdl dxcS

22 eS CLJd7 23 ..tbS .l::ted8 24 e6

Page 56: Beating King's Indian and Benoni

54 Main Line with 9 ... ~g4

fxe6 2S dxe6 4Jf6 26 !:!.xd8+ !:!.xd8 27 e7 with an advantage for White.

21 i.f2 'ilVxf4 22 'iVc1 'iVeS 23 i.bS!

The point of this move will become clear at the next note.

23 ... !:!.ec8 If 23 ... !:!.e7 24 i.d3± and

24 ... 4Jxe4? is now impossible due to 2S !:!.xe4.

24 'iVa3 4Jxe4 2S i.d3 fS 26 !:!'adl l:te8 27 i.xe4 fxe4 28 i.d4

White is clearly better. b) 13 ... a6 14 a4 c4 IS i.e3

'ii'a5 After the inclusion of ... a7-a6

and a2-a4 the resource b2-b4! is not available to White any more and 16 i.d4?! 4JcS 17 eS? 4Jd3! is favourable for Black. White therefore has to choose between two other possibilities:

b 1) The first is 16 'iIV e2 ! ? (Hitting the c4-pawn) 16 ... !:!.ac8 17 'iVf2 ctJcS (Black waited pas­sively, but without success, in Norwood-Westerinen, London 1988: 17 ... !:!.e7 18 !:!'adl 'iYb4?! 19 i.d4 !:!.ce8 20 !:!'e2 b6?! [20 ... h6!? 21 'iYg3 4Jh7] 21 'iVg3 [21 ~hl!?] 21...4JcS 22 eS 4Jfd7 23 e6 i.xd4+ 24 !:!.xd4 4Jf6? [24 .. .fxe6 2S dxe6 4Jxe6;l;] 2S fS±) 18 i.xcS 'iVxcs 19 'ilVxcs !:!.xcS 20 eS dxeS 21 fxeS 4Jd7 22 ctJe4! !:!.xdS (The only move. Otherwise 22 ... !:!.aS? 23 e6! fxe6 24 4Jd6± or 22 ... !:!.c7? 23 d6 !:!.c8 24 i.g4 !:!.cd8 2S i.xd7+-) 23 4Jf6+

i.xf6 24 i.xdS with a clear ad­vantage for White in the end­game, Lautier-Sutovsky, Til­burg 1996.

The second way is the pro­phylactic move:

b2) 16 ~hl This avoids a check after

16 ... 4JcS 17 i.xcs 'iVxcS. We consider three options for Black:

b21) 16 ... 4JcS 17 i.xcs 'ilVxcs 18 eS dxeS 19 fxeS 4Jd7

This is the critical position. 204Je4 Often White plays 20 e6 first.

The text move gives the addi­tional possibility of playing 21 4Jd6 after 20 ... 'iVe7.

20 .. .'ii'b4!? 21 e6 fxe6 If 21...4JeS 22 4JgS!? gives

White an initiative. On the other hand 22 exf7+ 4Jxf7! 23 d6? is similar to the game Glek­Kaminsky, Odessa 1989, (where White was better) but in that example Black's queen was on b6 and the a-pawns on their starting positions, so this does not work here. Black can simply plav 23 ... !:!.ad8 and if 24 'iVdS?

Page 57: Beating King's Indian and Benoni

then 24 .. J!e5. Instead of 23 d6 White should play 23 .i.g4 4Je5 24 ~e6+ transposing to the text.

22.i.g4! 22 dxe6?! 4Je5 23 ~d6 was

successful in the game Bode­Apicella, Paris 1990, after 23 ... ~a5? 24 e7! 4Jxf3 25 'iWe6+ 'It>h8 26 4Jd6! 4Jg5 27 'iWd7+-, but Black's play was improved in the subsequent game Hensberger-Reinderman, Bussum 1993: 23 ... a5! 24.l::i.adl 4Jd3+.

22 ... 4Je5 23 ~xe6+ 'It>h8 This type of position is gen­

erally slightly better for White (in practice the queen is some­times on e7 or c7 and the a­pawns on their starting squares). Here are two examples:

b21l) Kakage1dyev-Yurtaev, USSR 1983 (black queen on e7). 24 ~d2 .l::i.f8 25 .l::i.f1 4Jd3? (25 ... h6!) 26 4Jg5! .i.f6 27 d6! ~xd6 28 4Jf7+ .l::i.xf7 29 .i.xf7 ~xb2 30 .l::i.adl ~d4 (30 ... b5 31 ~xb2!) 31 ~xc4 ~xc4 32 ~xd3 ~xa4 33 .l::i.bl±.

b212) Summerscale-Buckley, Hastings Masters 1995 (black queen on e7, pawns on a2 and a7 and one move less). 23 .l::i.f1 .l::i.f8 24 ~c2 b5 25 b3 4Jd3 26 .l::i.adl ~h4 27 ~e2 .l::i.ae8 28 g3 ~h5? 29 ~xh5 gxh5 30 4Jd6±.

b22) 16 ... h6?! 17 ~d4 (17 .l::i.e2!?) 17 ... 4Jc5 18 ~xc5 ~xc5 19 e5 dxe5 20 fxe5 4Jd7 (One might suppose that Black has won a tempo compared with

Main Line with 9 ... i.g4 55

variation b 1, but in reality the move ... h7-h6 has just weak­ened the g6-pawn) 21 e6 4Je5 22 exf7+ 'It>xf7? (22 ... 4Jxf7!? 23 .l::i.e6!?) 23 d6!± 4Jd3? 24 ~d5+ 'It>f6 (24 .... ~xd5 25 .l::i.f1+!) 25 ~f3+ 1-0 Vaisser-Le Quang, Ostend 1992.

b23) 16 ... .l::i.e7 17 .l::i.e2! An unfortunate manoeuvre by

the white queen was met with a refutation in the game Peicheva-J. Polgar, Novi Sad Olympiad 1990: 17 ~d2?! ~ae8 18 ~f2?! 4Jxe4! 19 4Jxe4 .l::i.xe4 20 ~xe4 .l::i.xe4 and Black is slightly better.

Also less precise than the text move is 17 ~d4?! 4Jc5! 18 .i.xc5 (18 e5? 4Jd3!) 18 ... ~xc5 19 e5 .l::i.ae8.

17 ... J:.ae8 17 ... 4Jc5!? 18 .i.xc5 ~xc5 19

e5 is similar to variation b21. 18 ~d4 4Jc5 19 e5! 4Jfd7 In the case of 19 ... dxe5 20

fxe5 4Jfd7 21 d6 White simply wins an exchange.

2U..e6l

In this sharp position White has the better chances.

Page 58: Beating King's Indian and Benoni

56 Main Line with 9 ... j.,g4

14 cJth1! Kasparov succeeded in equal­

ising against Nogueiras (Bar­celona 1989) after 14 ii.e3?! bS! IS lbxbS lbxe4 16 ii.xe4 1:lxe4 17 lbxd6 ':xe3 18 ':xe3 ii.d4 19 ~f3 ':b8 20 cJth2 lbf6 21 lbc4 (21 lbe4 .txe3 22 lbxf6+ ~xf6 23 ~xe3 ~xb2=) 21.. . .txe3 22 ~xe3 ~xdS=.

14 15 a4 16 .te3

a6 c4 lbc5?!

It is better to take back with the queen on cS rather than with the rook, as we have seen in other games. Therefore Black should try 16 ... ~aS!? 17 ':e2 ':cd8, preparing 18 ... lbcs, as in the game Vegh-Reinderman (Haarlem 1994).

17 .txc5 18 eS 19 fxe5 20 e6

':xc5 dxe5 lbd7

This position is similar to that which arose in variation b21 above, but with the rook on cS.

20 lbe5

If 20 .. .fxe6 21 dxe6 lbeS 22 ~xd8 1:lxd8 23 ii.xb7 with an extra pawn for White (23 ... lbd3? loses to 24 e7 ':e8 2S ii.dS+).

21 exf7+ rJ;;xf7 Christine Foisor had serious

problems against Zsuzsa Polgar (Tilburg Candidates 1994) after 21...lbxf7 22 lbe4 (22 ':xe8+!? also seems strong: 22 ... ~xe8 23 lbe4 ':as 24 d6! lbeS 2S b4! cxb3 26 ~xb3+±) 22 ... .:aS 23 ii.g4 ~h8 24 ii.e6lbes 2S ~d2 h6 26 ':adl lbd3 27 ':f1 ':xdS? 28 ii.xdS ~xdS 29 lbf6! iLxf6 30 ~xh6+ 1-0.

22 lbe4! ':c8 A very nice mate awaited

Black after 22 ... lbd3

(see following diagram)

23 ':f1 ':xdS? (If 23 ... .:aS White can choose between 24 .tg4+ ~g8 2S .te6+ ~h8, transposing to the game Zsu. Polgar-Co Foisor in the previous

Page 59: Beating King's Indian and Benoni

note, or the sharp 24 b3!?) 24 tLld6+!! ~xd6 2S ~dS+ ~e7 26 hif7 mate.

23 24 25 26 27 28 29

jLg4! ~xg4 d6 ~adl ~d5! ~e2! ~d2

tiJxg4 ~g8 ~c6 ~e5 ~e8 ~e6

The desired construction has been built. White is winning thanks to the passed d-pawn.

29 h6 30 d7 ~e7 31 tLlc5 c3 32 bxc3 jLxc3 33 ~xe7 1-0

Main Line with 9 ... !tg4 57

Game 8 Vaisser-Smirin

peA (rapid play) Moscow 1996

1 d4 tLlf6 2 c4 g6 3 tiJc3 jLg7 4 e4 d6 5 f4 0-0 6 tiJf3 c5 7 d5 e6 8 ~e2 exd5 9 cxd5 ~g410 0-0

10 tiJbd7 The most precise way of pre­

venting e4-eS. Instead: a) The consequences of the

indifferent move 1O ... a6?! were grave in the game Vegh­Dambacher, Haarlem 1994: 11 eS! dxeS 12 fxeS tiJfd7 13 e6 fxe6 14 tLlgS! jLxe2 IS ~xe2 eS 16 jLe3 ~xf1 + 17 ~xf1 tiJf6 18 tLle6±. We see a similar out­come after 1O ... ~aS?! 11 eS! dxeS 12 fxeS tLlfd7 13 e6 fxe6 14 tLlgS! with a strong initiative.

b) 10 ... jLxf3?! As we shall see in Game 9

(Monin-Shchekachev) the line 1O ... tLlbd7 11 tiJd2 is harmless, so Black should not be in a rush to exchange this bishop. This move allows White to save the tempo h2-h3 and so cannot be recommended. If Black wants to play the system with ... tiJbd7, ... .1Lxf3 and ... tiJe8 discussed below, he should adopt the move order 1O ... tiJbd7 11 ~el jLxf3 12 jLxf3 tiJe8, as this cuts out one or two dangerous op­tions for White.

11 .1Lxf3 tiJbd7 12 ~el White has another aggressive

plan here: g2-g4-gS, .1Lg4 or ~c2, followed by the manoeu-

Page 60: Beating King's Indian and Benoni

58 Main Line with 9 ... i.g4

vre e4-e5! d6xe5, f4-f5. As an illustration the game Arencibia­Baron, Mamesa 1996, contin­ued: 12 g4!? h6 13 h4 lbh7 14 g5 a6 15 ~e3 b5 16 ~d2 lbb6 17 b3 b4 18 lbe2 ii.xal 19.1:Ixal a5 20 lbg3 a4 21 ~bl lbd7 22 e5! dxe5 23 f5 axb3 24 axb3 ~a6 25 i.e4 ~b6 26 ~f1 with a strong initiative and full com­pensation for the exchange.

12 ... lbe8 12 ... ~e8 allows White the

main line position without hav­ing wasted a tempo on h2-h3.

13.l:te2 This is an important multi­

functional move in the ... i.g4 system, but here 13 i.e3!? pre­paring i.e3-f2-g3 and eventu­ally i.g4 seems to me more precise.

13 ... a6?

It was necessary to play 13 ... lbc7! 14 a4 a6 15 ~el ~e8 16 il.e3 .l:Ib8 17 a5 with a com­plicated game, although White still has slightly better chances.

After the text move, White emerges on top with 14 ~el! 'Wie7 15 a4 lbc7 16 .i.e3 ~fe8

(l6 ... b5 17 e5!) 17 ~dl (again threatening 18 e5) 17 ... ~d8 18 .i.f2± Vaisser-Maki, Helsinki 1991. The manoeuvre i.c1-e3-f2-g3 (or h4) is a typical one in this whole system.

c) 1O ... .l:te8

Now the solid 11 h3 .i.xf3 12 ii.xf3, trying to transpose to the main line after 12 ... lbbd7, can be met by 12 ... lbfd7!? or 12 ... lba6!? 13 ~ellbd7. This is not bad for White, but is not to everyone's taste. It is more tempting to play:

11 e5! dxe5 In the game Jarovik-

Mirovschikov, Russian Ch. Semi-Final 1996, White ob­tained a strong attack after 11...lbfd7?! 12 e6 fxe6 13 lbg5 ii.xe2 14 lbxe2 lbf8 15 dxe6 lbc6 (l5 ... lbxe6? 16 ~d5 ~d7 17 f5 gxf5 18lbf4+-) 16 f5 d5?! (l6 ... gxf5!?) 17 lbf7 ~e7 18 i.g5 .i.f6 19 lbf4 il.xg5 20 lbxd5.

12 fxe5 il.xf3 If 12 ... lbfd7 13 .i.g5 f6

(l3 ... 'Wib6!? 14 lba4!) 14 exf6 .i.xf6 15 ~d2 with the better

Page 61: Beating King's Indian and Benoni

chances for White, Tozer-Lane, London 1988.

13 .txf3 lhe5 14.tf4 and now: c1) 14 ... lIe7 15 'i'b3 tbbd7?!

(i5 ... lId7 seems better, but still after 16 .te2! White has more than enough compensation for the pawn) 16 d6 !le6 17 tbd5! tbxd5 18 .txd5 ~d4+ 19 'It>hl ~f6 20 ~xb7 !lb8, Schon­Cvitan, New York open 1987, and now White should have played 21 ~xa7! !lxb2 22 'fic7 with a clear advantage.

c2) 14 ... lIe8 15 'i'b3! 'fib6 16 ~xb6 axb6

17 iLc7! The most precise. 17 tbb5?!,

Vaisser-Guyard, Aubervilliers (rapidplay) 1993, gave Black the possibility of 17 ... lIa4! with counterplay, while if 17 a4 then 17 ... tbfd7!

17",lUfd7 Or 17 ... tba6 18 d6! 18 lifel ! White is clearly better. For

example: 18 ... tba6 19 d6lIxel+ 20 lIxel ~xc3 21 bxc3 tbxc7 22 dxc7 !le8 23 'JJ.e7 lIxc7 24

Main Line with 9 ... iLg4 59

i.d5±. d) 1O ... tbfd7!? This clever

move poses some problems for White:

dl) 11 a4 tba6 12 tbb5 ~e7 13 h3 i.xf3 14 i.xf3 tbb4 15 e5 dxe5 16 d6 ~h4 17 i.xb7 lIab8 18 .i.d5, Piskov-Itkis, Kastel Stari 1988, and now 18 ... tbb6!? 19 i.b3 c4 20 ~d2 a5 would have given Black the better chances in a complex position.

d2) 11 tbg5?! (If White wants to exchange bishops here, in contrast to variation b, it is bet­ter to achieve this with 11 tbd2) 11.. .i.xe2 12 'fixe2 h6 13 tbf3 lIe8 14 ~c2 tba6 15 .i.e3 f5 16 tbd2 tbf6 17 h3 tbb4 with an initiative, Piskov-Neverov, Pod­olsk GMA qualifier 1989.

d3) The interesting move 11 tbb5!? has never been tried. Black has three reasonable pos­sibilities:

d31) 11...tbb6 12 a4 a6 13 tbc3t (if 13 tba3?! lIe8 14 h3 i.d7! 15 a5 tbc8 16 e5 ~xa5 and White does not have enough compensation for the pawn).

d32) 11...~b6 12 a4 c4+ 13 ~hl .i.xf3 (Otherwise 14 tbd2! follows) 14 .i.xf3 tbc5 15 ~e2t.

d33) 11...tbf6 12 h3 (12 e5!?) 12 ... i.xf3 13 i.xf3 lIe8 14 lIel a6 15 tbc3 tbbd7 16 a4 and we have arrived at a well-known position which favours White (see Game 7, Vaisser-Berelo­vich).

Page 62: Beating King's Indian and Benoni

60 Main Line with 9 .. . .i.g4

11 ~e1 tDe8!?

A rather unusual idea in this position.

12 a4 Instead: a) 12 h3?! (It is not necessary

to spend a tempo provoking this exchange) 12 ... ~xf3 13 ~xf3 ~bS leads to Game 9, Monin­Shchekachev.

b) 12 tDd2 is playable here. After 12 ... ~xe2 13 'iixe2 (13 lIxe2!?) 13 ... a6 14 a4 ~d4+ 15 ~hl tDef6 16 'iif3 lIeS 17 tDe2 'iie7 IS tDxd4 cxd4 19 b3 tDxd5 20 ~a3! tDe3 21 lIec1 'iie6 (21...tDc5!?) 22 ~b2 lIacS 23 lIxcS lIxcS 24 ~xd4 tDc2 25 lIc1 tDe5 26 ~xe5 dxe5 27 fxe5 tDd4 28 lIxcS+ 'iixcs 29 'iidl 'i'c3 30 tDf3 tDe6 the position is equal, Banikas-Beliavsky, Euro­pean Club Cup 1995.

c) 12 tDg5!? ~xe2 13 lIxe2! seems to me the most unpleas­ant for Black. Compared to the game Piskov-Neverov from the previous note, the position of Black's knight on eS instead of

d7 is in White's favour. 12 a6 13 ~e3 tDc7 14 ~f2 lIb8 15 ~h4 ~f6 16 ~g3 ~xf3 17 ~xf3 ~h4?

Black should not permit e4-e5. It was necessary to play 17 ... ~d4+! IS ~hl lIeS with counterplay.' "I, ", I

18 19 20 21 22 23

, I

eS! hxg3 fxeS tDe4 tDxcS tDe4

~xg3 dxeS 'iigS 'i'xeS 'i'd6

White's pawn sacrifice to ac­tivate the d-pawn is the key to this position.

23 24 ~h2 25 lIc1 26 d6 27 ~c7 28 tDgS!

'iib6+ 'iixb2 tDe8 tDef6 l:1fd8

Preparing a decisive piece sacrifice.

28 h6

Page 63: Beating King's Indian and Benoni

29 ~e2! First driving away Black's

queen ... 29 "iWb4

and now giving up the knight!

30 etJxf7! cj;;xf7 31 ~e7+ cj;;g8

31... cj;;f8 changes nothing. White plays 32 "iWd3 and after both 32 ... etJg4+ 33 ~xg4 "iWxg4 34 "iWc3 "iWh5+ 35 'itgl "iWdl+ 36 'itf2 etJe5 37 ~xe5 and 32 ... etJg8 33 "iWxg6! etJxe7 34 "iWxh6+ 'itf7 35 j,h5+ 'itg8 36 "iWe6+ 'ith8 37 dxe7 White wins.

32 "iWd3 etJg4+ 32 ... etJf8 doesn't help: 33

~g7+ 'ith8 34 "iWe3 g5 35 "iWe7 "iWd4 36 ~g6 etJg8 37 "iWf7 etJd7 38 j,e4 with mate in two.

33 ~xg4 "iWxg4 34 "iWb3+ 1-0

Game 9 ~onin-Shchekachev

St Petersburg open 1994

1 d4 etJf6 2 c4 g6 3 etJc3 iLg7 4

Main Line with 9 ... J..g4 61

e4 d6 5 f4 0-0 6 etJf3 c5 7 d5 e6 8 ~e2 exd5 9 cxd5 j,g4

10 0-0 Instead White can play: 10 e5?! This line is not very danger­

ous for Black. Nevertheless, he must play precisely. For exam­ple 1O ... etJfd7?! is suspicious: 11 e6! fxe6 12 etJg5 j,xe2 13 "iWxe2 exd5 14 ~e6+ 'ith8 15 etJxd5 etJf6, Sutter-Wojtkiewicz, Bern open 1991, and now the simple 16 etJf7+ ~xf7 17 'jVxf7 etJxd5 18 ~xd5 ~e8+ 19 'itf2 j,d4+ 20 'itf3 would have maintained White's advantage.

1O ... dxe5 11 fxe5 iLxf3 12 ~xf3

Not 12 gxf3? etJh5. 12 ... etJfd7 For 12 ... ~e8?! see the com­

ments in variation c, Black's 10th move in Game 8.

13 e6 etJe5

Hll=Q Weaker is 14 exf7 +?! ~xf7

15 0-0 etJbd7 and now both 16 j,e2 ~xfl + 17 ~ xfl "iWb6!, Ludden-De Vries, COIT. 1994,

Page 64: Beating King's Indian and Benoni

62 Main Line with 9 ... ~g4

and 16 ttJe4 ttJxf3+ 17 l:txf3 l:txf3 18 'iixf3 ttJe5, Knezevic­Gligoric, Yugoslavia 1970, give Black the better position.

After 140-0: a) Warning - don't believe

the published recommendation 14 ... ttJxf3+ 15 l:txf3 fxe6 16 l:txf8+ 'iixf8 17 ~e3! ttJa"6! 18 dxe6 'Wie7 19 iYb3 l:te8 20 l:tel ttJb4 (20 ... 'iixe6?? 21 ~g5+-) 21 iLf2 ~h8 22 a3 ttJc6 23 ttJd5 'iid6 24 e7, Schoen-Sherzer, New York open 1987, 24 .. . iLe5! 'and Black is slightly better'. After 25 ~h4! White's threats are very strong. For example, 25 ... iLg7 26 ttJf6! iLxf6 27 'ii'f7 i.d4+ 28 ~hl 'ii'b8 29 l:tdl! and Black is obliged to give back a rook.

b) 14 .. .fxe6 and now: bl) 15 i.e4l:txfl+ 16 'ii'xfl and: bll) 16 ... ~h8 The old Hort proposition 17

i.e3? was tried in the game Hajenius-Le Quang, Brussels Zonal 1993: 17 ... exd5? 18 i.xd5 ttJbd7 with a complicated game, but the simple 17 . ..cbg4! 18 iLxc5 (18 l:te 1 ttJxe3 19 l:txe3 i.d4) 18 ... 'ii'c7 19 'iib5 ttJa6 would have won for Black.

Instead of 17 iLe3? White should play 17 dxe6 and now Black has a choice between a draw after 17 ... 'iid4+ 18 ~hl ttJg4 19 e7 ttJa6 20 .txb7 ttJf2+ and a complicated game after 17 ... ttJbc6 18 iLf4 Vi'd4+ 19

~hl ttJg4 20 ii.g3 ~f8. b12) 16 ... ttJg4 seems attrac­

tive, but after 17 g3! ~xc3 18 bxc3 exd5 19 .tg2! ttJc6 20 Vi'dl White has good compen­sation for the two pawns.

b2) 15 ii.e3 ttJxf3+ 16 l:txf3 l:txf3 17 iYxf3 exd5 18 ttJxd5 ?bcti

and now: b21) 19 ~fl?! ttJe5 20 'ii'e4

Vi'd6 21 i.g5 ttJc6 22 i.f6 :gf8 23 'ii'c4 ~h8, when White's attack proved unsuccessful in Bischoff-Hellers, San Ber­nardino 1990.

b22) 19 i.xc5 ~h8 20 ~dl Vi' a5 and Black has an edge.

Page 65: Beating King's Indian and Benoni

10 ttJbd7 11 h3

For the main line 11 .sel see Game 6, Vaisser-Yrjola, and Game 7, Vaisser-Berelovich.

Instead: a) lLa4 This is not just a waste of

time. It can be considered as a useful waiting move at the mo­ment, as Black still has not cho­sen which piece is going to eS. White demonstrated some inter­esting ideas in Kahn-Mah, Bu­dapest 1995: 11...ttJeS 12 ttJgS!? .txe2 13 ~xe2 ttJc7 (l3 ... ~d4+!? 14 <Jihl ttJef6) 14 lla3 (An unusual way of pre­paring an attack on the king­side) 14 ... iVe7? (The combina­tion of a queen on e7 and a knight on c7 is unfortunate, and White makes perfect use of this. Better was 14 ... h6 IS ttJf3 .seS) IS eS! dxeS 16 ttJge4! ttJeS 17 fS gxfS?! IS .sxfS ttJd6 (Black has helped a lot and now White's attack is very strong) 19 ttJxd6 iVxd6 20 ttJe4! ~xdS 21 .sg3 <JihS 22 ~ g4 l:igS 23 ~hS iVe6 24 ttJgS! ~xfS 2S iVxh7+! 1-0.

Black should try to obtain a position resembling that ana-1 ysed in V aisser -Berelovich where the move a2-a4 is often I ess useful than l:ie 1 :

11...lle8!? 11...a6 12 .sel .seS 13 h3

transposes to the main line of the 9 ... .tg4 variation, but with fewer options for Black.

Main Line with 9 ... ~g4 63

12 h3 .txf3 13 Jtxf3 c4 14 .te3 iVaS

and now: al) IS iVe2?! was successful

in Priehoda-Hass, Katowice open 1990, after Is ... iVb4? (1S ... .sacS? 16 ttJbS!) 16 iVf2 .sacS (l6 ... a6 17 as) 17 .txa7 ttJcS (17 ... b6 IS as bxaS 19 .sa4±) IS .txcS .sxcS 19 .sael ttJd7 20 .tg4 ttJf6 21 eS!±. Black should play actively: IS ... ttJcS! 16 iVxc4 ttJfxe4 17 ttJxe4 ttJxe4 IS ~xe4 .sacS! 19 ~d3 iVb4 with full compensa­tion, e.g. 20 .tf3? .sxe3!

a2) The usual continuation is IS .td4 .se7' It is premature to play

IS ... ttJcS?! 16 eS! ttJfd7 17 e6 fxe6 IS dxe6 ttJxe6 19 .tdS (19 <Jih1!?) 19 ... .txd4+ 20 iVxd4 ~cS 21 .sad 1 and White is slightly better, Fang-Zapata, Philadelphia 1994.

16 <Jih2! This prevents the thematic

exchange sacrifice that is possi­ble in the case of 16 <Jihl a6 17 g4 .saeS! IS gS ttJxe4 19 ttJxe4 (19 .txg7 ttJg3+) 19 ... .sxe4 20

Page 66: Beating King's Indian and Benoni

64 Main Line with 9 ... iLg4

ii.xe4 .l:txe4 21 'iWf3 f5! and .l:teS 13 'iWf3 White is not better, Peev­Velimirovic, Sofia 1972.

16 .. .'::tJc5! 17 e5lfleS'

The game Hausner-Donchev, Pardubice 1994, continued IS lflb5 lflb3 19 exd6 .l:td7 20 ii.xg7 rt;xg7 21 'iWel "iVxel 22 .l:taxel lflxd6 23 lflxd6 .l:txd6 24 .l:te7 lfld2 with equality.

Other White 11th moves are not dangerous:

b) 11 rt;hl?! .l:teS 12 h3 ii.xf3 13 ii.xf3 b5! 14 "iVc2 b4 15lfldl .l:tcS 16 lfle3 c4! with good counterplay, Krcmar-Felix, Kar­vina 19S9.

c) 11 "iVc2?! .l:teS 12 ii.d2 a6 13 a4 .l:tc8 14 a5 c4 15 .l:ta4 lflc5! 16 .l:txc4 lflcxe4 17 lflxe4 lflxe4 IS ii.d3 .l:txc4 19 ii.xc4 'Wic7 20 ii.b4 lflc3! 21 'Wib3 lfle2+ 22 rt;hl? ii.xf3 23 .l:txf3 lfld4 24 "iV a4 .l:te7 0-1 Holder­ried-Groszpeter, Andorra open 1995.

d) 11 ii.e3?! ~eS 12 lfld2 ii.xe2 13 "iVxe2 b5! 14 "iVf3 b4 15 lfldl "iVe7+ Kaidanov­Lerner, Norilsk 19S7.

e) 11 lfld2?! ii.xe2 12 'Wi xe2

13 ... ~c8!? This strong move, preparing

14 ... c4, was introduced by Nunn. Here White has tried a number of possibilities:

el) 14 lflc4?! lflb6 15 lflxb6 (Or 15 lflxd6?! "iVxd6 16 e5 'Wi d7 17 exf6 ii.xf6 18 f5 ii.d4+ 19 rt;hl ii.xc3! 20 bxc3 lflxd5 21 "iVg3 lflf6!-+ Wells-Nunn, Borehamwood 1980) 15 ... 'Wixb6 16 ~hl 'Wib4 17 e5 dxe5 18 fxe5 ~xe5 19 ~f4 ~f5! 20 g4 lflxg4 21 "iVxg4 "iVxb2 22 lfle2 .l:teS 23 lflg3 ~xd5+ Michael­sen-Maus, Hamburg 1987.

e2) 14 lflb5?! 'Wib6 15 a4 c4+ 16 "iVf2 ~c5! 17 b4 cxb3 18 ii.a3 lflg4! 19 ii.xc5 lflxc5 20 lflc4 lflxf2 21 lflxb6 lflfxe4 22 lflc4 a6 23 lflbxd6 lflxd6 24 lflxd6 ~d8 25 lflc4 ii.d4+ 26 rt;hl ii.xal 27 ~xal .l:txd5 0-1 Maier-Enders, Meisdorfer Schloturnier 1996.

e3) The most aggressive plan was demonstrated by Nogueiras against Cvitan (Novi Sad Olympiad 1990):

14 rt;hl c4 15 g4?!

Page 67: Beating King's Indian and Benoni

Cvitan immediately made a serious mistake with IS ... liJcS? and White obtained a clear ad­vantage after 16 eS! dxeS 17 fxeS 1;IxeS 18 liJxc4 liJcd7 19 liJxeS liJxeS 20 ~ g3 liJexg4 21 ,~f4. Black should have played lS ... h6! first. If White contin­ues his plan with 16 h4, then 16 ... liJcS! and now 17 eS? dxeS 18 fxeS liJfd7! 19 ~xf7+ ~h7 is just bad for White. Black is also better after 17 gS hxgS 18 hxgSliJhS.

11 ~xf3 12 ~xf3

12 1;Ib8!? Black has a wide choice here:

Main Line with 9 ... ~g4 65

a) The immediate 12 ... liJe8!? deserves attention. If White plays slowly then punishment is quick: 13 ~el?! liJc7 14 a41;Ie8 IS ~g3 c4! 16 ~e3 liJa6 17 1;Iadl liJacS 18 ~d4 ~xd4+ 19 1;Ixd4 ~b6 20 as ~xb2 211;Ixc4 bS! 22 1;Ibl ~c2 23 1;IxbS ~d3 24 1;Ibb4 liJa6 with a large ad­vantage for Black, Fors-Akes­son, Stockholm 1990. White played more accurately in the game Adianto-Hulak, Jakarta 1986 (by transposition): 13 g4 a6 14 gS liJc7 IS h4!? liJbS 16 ~d2 liJd4 17 hS f6 18 h6 ~h8 19 ~g4 fS 20 exfS liJxfS 21 ~e2 1;If7 22 liJe4 liJb6 23 ~c3;t.

b) The following variation is rather popular, so it is useful to take a look at it:

12 ... .l:.e8 13 g4!? 13 1;Iel leads to the main line

of the 9 ... ~g4 variation. 13 ... h6 14 h4

and now: b 1) The piece sacrifice is not

well founded: 14 ... bS?! IS gS b4 (lS ... liJh7 16 liJxbS±) 16 gxf6 liJxf6 17 liJe2!? liJxe4 18

Page 68: Beating King's Indian and Benoni

66 Main Line with 9 ... i.g4

h5! and both of the following variations give a clear advan­tage to White:

bll) 18 ... gxh5 19 ~d3 (19 ~xe4!? :'xe4 20 tDg3±) 19 ... f5 20 .txh5 ~h4? (20 ... :'e7!?±) 21 .txe8 :'xe8 22 ~f3 tDf6 23 ~g2 tDh5 24 .te3 :'e4 25 :'ael .i.xb2 26 lbg3+- Scholseth­Moen, Norwegian Ch. 1992.

b12) 18 ... ~h4 19 ~g2 g5, Osmanbegovic-Rotstein, Mari­bor Pirc 1994, and now 20 .txe4! would have been strong: 20 ... :'xe4 21 :'hl ~g4+ 22 tDg3 ~xdl 23 :'xdl :'c4 24 tDf5 :'d8 25 ~h3! (25 ~f3?! .txb2!) with an edge for White.

b2) More solid is 14 ... tDh7 15 g5 hxg5 16 hxg5:

b21) Now 16 .. .f6?! is not good, as has been shown twice: 17 gxf6 ~xf6 (17 ... tDhxf6!?) 18 ~e3 (Or 18 ~g2 a6 19 a4 :'e7 20 e5! dxe5 21 d6 :'ee8 22 .td5+ ~h8 23 tDe4 ~f5 24 tDg5+- Moutousis-Hon Kah Seng, Thessaloniki Olympiad 1988) 18 ... :'e7 (18 ... g5 19 e5 ~h6 20 .tg4) 19 ~g2 b5? 20 e5! tDxe5 21 tDe4 ~h4 22 :'h 1 tDc4 23 .i.c1 1-0 Arencibia­Gonzalez, Capablanca Memo­rial 1993.

b22) After 16 ... a6 17 a4 c4 18 .te3 :'c8 White has a choice between the quiet 19 .td4 ~xd4+ 20 ~xd4 tDc5 21 :'adl with a roughly equal position, Monin-Purtov, Budapest 1993, and complicated play with 19 .tg4!? .txc3 20 bxc3 :'xe4 21

~f3! (Not 21 .td4? tDxg5Ff) 21...~e7 22 .td4 :'e8 23 :'a2 tDdf8, Stankovic-Paunovic, Ce­tinje 1992. Now instead of the continuation 24 "iYh3? :'xd4! 25 cxd5 b5+ White should have played 24 a5! with good com­pensation for the pawn .

b3) 14 ... h5!?

This actually seems to be Black's best:

b31) Existing theory gives the variation 15 gxh5 tDxh5 16 .txh5 gxh5 17 ~xh5 b5! 'with compensation'. This final evaluation can be accepted de­spite a spectacular White vic­tory in the game Dekusar­Borulia, USSR 1990: 18 e5 dxe5 19 fxe5 tDxe5? (It was necessary to play 19 ... :'xe5! 20 ~xf7+ ~h8 and if 21 ~f4 then 21...~xh4!) 20 .th6 tDg6? (Also bad was 20 ... b4? 21 ~xg7 ~xg7 22 :'ae1! f6 23 ~hl! with a mating attack, but it was better to play 20 ... ~d6!? 21 ~xg7 ~xg7 22 tDxb5±) 21 ~xg7 ~xg7 22 :'xf7+! ~xf7 23 ~h7+ and mate was not far off. However, the problem is

Page 69: Beating King's Indian and Benoni

that the very first move of this variation is wrong! Let us re­write it like this:

15 gxh5? ttJxh5 16 .ixh5?! 16 iVel is met by 16 .. .fS+.

Now instead of 16 ... gxhS?? Black has a nice, practically winning improvement:

16 .•. W!ixh4!! 17 iVg4 (Not 17 .Qg4? .ixc3! 18 '.tg2 ttJf6 19 .~f3 .id4-+) 17 ... iVxhS 18 ~xd7? (18 iVxhS gxhS+) 18 ... Mad8 19 iVb5 iVg4+! 20 '<t>f2 (20 '.th2 .ixc3 21 iVd3 ~d4-+) 20 ... .id4+ 21 '.tel .~xc3+ 22 bxc3 I:i.xe4+-+.

b32) So White should try IS gS ttJg4 16 .ixg4 hxg4 17 I:i.e 1 (Or 17 iV xg4 .ixc3 18 bxc3 ~xe4 19 .id2 iVe7 20 Mae1 ttJb6 21 iVf3 I:i.ae8 22 f5! ttJxdS 23 fxg6 fxg6 24 c4! ttJe3 25 ,~xe3 with a drawn ending, Bach-Trisic, Hamburg 1996) 17 ... c4 18 .ie3 (It would be in­teresting to check out 18 iVxg4 ttJcS 19 iVf3) 18 ... .txc3 19 bxc3 Mxe4 20 'iYxg4 'iYe7 21 ~f2 ttJcS 22 Mxe4 iVxe4 (Kouatly-Kindermann, Tmava 19R7) and here. accordinQ to

Main Line with 9 ... ~g4 67

Kindermann, White has enough compensation after 23 Me 1 iVxdS 24 ~d4 with the idea of h4-hS.

13 Mel ttJe8 If 13 ... b5?! 14 e5±.

14 ~g4?! 14 a4 a6 (14 ... ttJc7!?) 15 a5

bS 16 axb6 Mxb6 17 Me2;!;; would have corresponded more with the spirit of the position.

14 f5 15 exf5

Or IS .tf3 bS! 15 gxf5 16 .te2 a6

Avoiding 16 ... 0.c7 17ttJbS!? 17 .td3!

Or 17 a4?! ttJc7 18 as ~d4+ 19 '.th2 'iYf6 with an initiative.

17 ttJc7 17 ... bS?! hands the initiative

to White: 18 Me6! Mb6 19 g4. 18 iVc2

Now it is too late for 18 Me6?: 18 ... ttJxe6 19 dxe6 ttJb6 20 'iYb3 'iYc7 21 ttJd5 ttJxd5 22 iVxdS iVc6 with a clear advan­taQe for Black.

Page 70: Beating King's Indian and Benoni

68 Main Line with 9 ... j.g4

18 iVf6 With the threat of 19 ... c4.

19 lbe2

19 kibe8?! It would have been better to

play 19 ... lbxd5!? 20 lbg3 lbe7!, when the exchange sacrifice doesn't work: 21 i.c4+? d5 22 kixe7 iVxe7 23 i.xd5+ ~h8 24 lbxf5 iVel+ 25 ~h2 kixf5! 26 iVxf5 i.d4 27 i.e3 iVxe3 28 iVxd7 iVxf4+ 29 ~h1 i.e5 30 ~g1 iVe3+ 31 ~h1 iVg3 32 ~gl kif8-+.

20 kid1 The inaccurate 20 i.d2? just

gives away a pawn after 20 ... lbxd5! 21 iVb3 lb7b6, and both attempts to justify it don't work: 22 a4? kixe2! 23 kixe2 (23 i.xe2 iVd4+) 23 ... c4-+; and 22 i.a5? iVxb2 23 iVxb2 i.xb2 24 kiab1 c4! 25 i.c2 i.a3-+.

20 iVh4! 21 ~h2

21 i.xf5? kixf5! wins.

21 22 i.xf5 23 lbg3 24 iVf2 25 iVf3

lbf6 lbcxd5 ~h8 lbe7

Black's dream of a knight fork becomes real in the case of the stupid 25 i.e6? lbc6! 26 kixd6?? l:I.xe6 27 l:I.xe6lbg4+.

25 d5 26 i.e3?

The decisive mistake. White would still have been okay after 26 i.d2 d4 27 i.e6!

26 iVxg3+! 27 iVxg3 lLlxf5 28 iVg5

28 iVf2? lLlxe3 29 kidc1 lLlfg4+-+.

28 29 30 31 32 33

kie1 iVh5 kiac1 iVf3 kixe3 0-1

lLlxe3 lLle4 d4 c4 kixf4! kixf3

Page 71: Beating King's Indian and Benoni

3 Main Line: 9 ... b5 and rare moves

In this chapter we shall consider Black's alternatives to 9 ... .l'iIe8 and 9 ... il.g4 in the main line.

Game 10 Komarov-Frolov

Kiev 1995

1 d4 lbf6 2 c4 g6 3 lbc3 ~g7 4 c4 d6 5 f4 0-0 6 lbo c5 7 d5 e6 8 ~e2 exd5 9 cxd5

9 b5 First, let us take a look at

Black's rare alternatives: a) Black sometimes tries to

prevent e4-e5 by 9 ... lbfd7?!, but this cannot be recommended, because White can play 10 lbd2, when 9 ... lbfd7 is sense­less. In Taimanov's 8 ~b5+ (Chapter 7) White is even will­ing to play this position a tempo down. Here is a practical exam­ple: 9 ... lbfd7?! 10 0-0 lba6 11 'Ot>hl (11 lbd2 l:Ie8 12 il.f3 lbc7 13 a4 a6 14 lbc4 i.f8 is also very attractive for White, Michaelsen-Ziger, Eger 1988) 11.. . .l'iIe8 12 lbd2 lbb6 13 a4 ~d7 14 a5 lbc8 15 ~f3 .l'iIb8 16

lbc4 lbb4 17 e5! il.f5 18 il.e4 i.xe4 19 lbxe4 dxe5 20 d6 and White is clearly better, Peev­Angelov, Bulgaria 1974.

b) 9 ... a6?! Now the simplest way for White to play is 10 a4 and after 10 ... .l'iIe8, 11 lbd2! transposing to an advantageous variation of the line 9 ... .l'iIe8 10 lbd2. It is worth bearing this in mind, because Black can practi­cally force White to play a2-a4 earlier on, for example with 7 ... a6.

Instead White can also choose between 10 e5!? and 10 O-O!? b5 11 e5 with good pros­pects in both cases.

c) 9 ... lba6?!, when White has a pleasant choice between:

c1) 10 O-O!? lbc7 11 a4 b6 12 e5! dxe5 13 d6 lbe6 14 fxe5 lbd7 15 lbd5 (A typical ma­noeuvre) 15 ... .tb7 16 lbe7+ 'it'h8 17 lbg5 lbxe5 18 lbxh7! .l'iIe8 (l8 ... ~xh7? 19 l1a3!) 19 lbg5 :xe7 20 dxe7 ~xe7 21 ~el! and White is clearly bet­ter, Kaplun-Aarland, Riga 1982; and

c2) 10 e5!? dxe5 11 fxe5

Page 72: Beating King's Indian and Benoni

70 Main Line: 9 ... b5 and Rare Moves

4:Jg4 12 ~f4 (12 ~gS f6!?) 12 ... l:i.e8 13 e6!? fxe6 14 d6 ~d7 IS ~d2 (1S h3!? also looks reasonable: IS ... 4:Jf6 16 4:JeS) IS ... 4:Jb4 16 0-0 l:i.f8 17 h3 4:Jf6 18 ~c4 bS!? (Not 18 ... a6? 19 l:i.ael± Konikowski­Braune, corr. 1994) 19 4:JxbS (19 ~xbS!?) 19 ... 4:JbdS?! (It was better to play 19 ... ~xbS 20 i.xbS 4:Je4 21 ~e3 4:JdS 22 ~xe4 4:Jxf4 23 g3!, but White still enjoys a certain advantage due to the passed d-pawn) 20 i.h2 4:Jb6 21 i.e2± Coo­Ciocaltea, Havana 1965.

d) 9 ... c4?! 10 i.xc4 (100-0 is also quite possible, e.g. 10 ... bS 11 eS!) 1O ... 4:Jxe4 11 4:Jxe4 l:i.e8 12 4:Jfg5! (Less convincing is 12 O-O?! l:i.xe4 13 i.d3 l:i.e8 14 fS transposing to the variation 9 ~d3 c4; while even worse are 12 4:JeS? i.xeS! and 12 4:Jd2? fS in both cases with an advantage for Black) 12 ... h6 (12 ... ~b6? 13 ~e2!±) 13 0-0 hxgS 14 4:JxgS 'iYb6+ IS ~hl and the big question is whether or not Black has enough compensation for the pawn.

e) 9 ... 4:Jbd7 and now: e 1) 10 eS was dealt a strong

blow in the game Lazic-Todo­rovic, Cetinje 1993: 1O ... dxeS 11 fxeS 4:Jg4 12 e6?! (12 ~gS!? 'iYb6 is much like the main line of 9 ... l:i.e 8 and needs further tests) 12 ... 4:JdeS 13 4:JgS fxe6! 14 ~xg4 4:Jxg4 IS 'iYxg4 exdS 16 'iYh4 h6 17 4:Jf3 gS with a very strong attack for Black in

the case of 18 'iYg3. Lazic pre­ferred to give a piece back: 18 i.xgS hxgS 19 'iYxgS ~xgS 20 4:JxgS ~g4!, but the resulting ending is much better for Black.

e2) White doesn't need all these complications, because he can simply play 10 0-0 l:i.e8 11 "VJic2!? (This seems even more interesting than 11 4:Jd2, which transposes to the variation 9 ... l:i.e8 10 4:Jd2 4:Jbd7) White's knight stays on f3, supporting an eventual e4-eS break.

We now consider the system with 9 ... bS.

This old variation has had its 'ups' and 'downs' over the years. In the early 1990s, when 9 ... bS was in crisis, the Ukrain­ian grandmaster A. Frolov gave it a new breath of life with 1O ... 4:Jfd7!? This line is studied in the current game, while older lines are dealt with in Game 11, Blokh-Kitchev. Today we can again state that Black has no easy means of obtaining equal­ity in this system.

Page 73: Beating King's Indian and Benoni

Main Line: 9 ... b5 and Rare Moves 71

10 e5! 10 iLxbS?! 4Jxe4 11 4Jxe4

'i'aS+ 12 <;;t>f2 ~xbS 13 4Jxd6 'i'a6 doesn't promise much for White.

10 4Jfd7!? After 1O ... 4Jg4?! 11 h3 4Jh6

12 iLxbS 4JfS 13 0-0 'i'b6, Skembris-Vuruna, Vrnjacka Banja 1989, 14 a4! would have maintained a clear advantage for White; while for lO ... dxeS 11 fxeS 4Jg4 see the next game, Blokh-Kitchev.

11 iLxb5! The only move to guarantee

White an advantage. Here are some illustrations of the prob­lems White has had with other continuations after 10 ... 4Jfd7!?:

a) 11 exd6?! a6! b) 11 4JxbS?! dxeS 120-0 e4

13 4JgS 4Jf6 14 4Jc3 ~e8 IS iLbS iLd7 16 ~xd7 4Jbxd7 17 ~e 1 4Jb6+ Feldmanrr-Frolov, Sibenik 1989.

c) 11 e6? fxe6 12 dxe6 and now:

c1) 12 ... 4Jf6 13 4JxbS (13

~xbS dS!) 13 ... dS 14 4JgS 'i'e7 IS fS a6 16 4Jc3 gxfS 17 4JxdS 4JxdS 18 'i'xdS ~b7 with a strong initiative, Zakharevich­Maximenko, Moscow 1991.

c2) 12 ... 4Jb6 13 0-0 4Jc6 14 4JgS b4 IS ~f3 bxc3 16 iLxc6 cxb2 17 ~xb2 ~xb2 18 e7 ~xe7 19 'i'b3+ c4 20 'i'xb2 ~b8+ Purgin-Frolov, Smolensk 1992.

11 12 0-0

dxe5 ..ta6

Another possibility is 12 ... 'i'b6 13 a4! and now Black has tried: a) 13 ... .ta6 14 <;;t>hl (14

~el !?) 14 ... ..txbS IS axbS exf4 16 iLxf4 4Jf6 17 4JeS 4Jbd7 18 ~a6 'i'b7 19 4Jc6 4Jb6 20 ..td6 with a big advantage, Kahn­Kaeser, Budapest 1996.

b) 13 .. J:td8 was tried in Ger­ard-Hagege, French Team Ch. 1996. White's 14 d6 was not bad, but 14 fxeS!? seems better, since the variation 14 ... 4JxeS IS .tgS f6 (1S .. JH8 16 ~e7) 16 4JxeS fxgS 17 4Jf7 ':'f8 18 4JxgS± is quite convincing.

c) 13 ... exf4

Page 74: Beating King's Indian and Benoni

72 Main Line: 9 ... b5 and Rare Moves

Now Lautier, White against Rogers at the Yerevan Olym­piad 1996, lost a substantial part of his advantage by playing the natural:

c1) 14 ~xf4?! a6 15 a5 llVb7 16 ~xd7 ttJxd7 17llVd2 ttJf6 18 i.e5. Now instead of 18 ... i.g4? 19 l:ta4! Black should have played 18 ... i.f5!?, with no great problems for him.

It was better to play c2) l41M1 For example, 14 ... a6 15 a5

~b7 16 i.c4 ttJc6 17 ttJd5 ttJde5 18 ttJxe5 i.xe5 19 i.xf4 i.xb2 20 l:tbl i.f5 21 ttJe7+! ttJxe7 22 dxe7 l:tfe8 23 i.d5±. This was my preparation for my game against Frolov in Gronin­gen 1993. Unfortunately, he declined to use his favourite King's Indian on that occasion.

13 a4! An improvement on 13 i.xa6

ttJxa6 14 f5! c4! 15 ttJg5 ttJac5 with a balanced game, Berk­ovich-Frolov, Alushta 1992.

13 i.xb5

14 axb5 l:teS Or 14 ... exf4 15 i.xf4 ttJb6 16

ttJe5! with an advantage. 15 l:tel?!

More to the point was 15 d6! exf4 (l5 ... e4 16 ttJxe4!) 16 ttJd5! with a clear advantage.

15 exf4 16 i.xf4 ttJf6?

The decisive mistake. Black simply forgot that the a7-pawn was pinned. 16 ... ttJb6! was nec­essary, but after 17 i.g5!? f6 (l7 ... l:txel+?! 18 'Wxel llVd7 19 d6! ~xd6 20 l:td1±) 18 i.f4 White is slightly better.

17 l:txeS+ llV xeS IS b6 a6

18 ... ~d7 doesn't help: 19 bxa7 l:txa7 20 i.xb8 and White wins.

19 llVb3 llVd7 Or 19 ... ttJbd7 20 b7 l:tb8 21

l:txa6!+-.

20 i.e7! ttJeS 21 d6!

It was not late to let things slip: 21 b7? l:ta7 22 i.xb8? (22 'Wb6±) 22 ... l:txb7 23 'Wa4 ~xa4

Page 75: Beating King's Indian and Benoni

Main Line: 9 ... b5 and Rare Moves 73

24 1:txa4 1:txb8 and Black has no prospects of losing.

21 tDc6 22 b7 1:tb8 23 iLxb8 tDxb8 24 1:tdl tDxd6 25 ~d5 iLfS 26 tDe4 ~xb7 27 tDxd6 1-0

Game 11 Blokh-Kitchev

USSR (carr.) 1991

1 d4 tDf6 2 c4 g6 3 tDc3 J..g7 4 e4 d6 5 f4 0-0 6 tDf3 c5 7 d5 e6 8 iLe2 exd5 9 cxd5 b5 10 e5

10 11 fxe5 12 iLg5

dxe5 tDg4

Although 12 iLf4 is not rec­ommended by theory, it is bet­ter than its reputation. Keres surprised Spassky with this move in their Candidates match in 1965 and could have ob­tained an advantage. Black must respond actively:

a) 12 ... tDd7?! 13 e6! fxe6 14 dxe6 1:txf4 (Or 14 ... tDb6 15 ~xd8 1:txd8 16 tDg5) 15 'iVd5 'It>h8 16 ~xa8 tDb6 17 ~xa7 iLxe6 (The attacking 17 ... tDe3? 18 e7 tDxg2+ is not justified and ended in a fiasco in the game Todorovic-Ilic, Yugoslav Ch. 1988: 19 'It>f2 J..d4+ 20 'It>xg2 .i.h3+ 21 ~xh3 [21 ~g3!?+-] 21...'i'c8+ 22 ~g2 ~g4+ 23 ~f1 llxf3+ 24 ~el i.f2+ 25 ~dl 1:td3+ 26 ~c1

1:txc3+ 27 bxc3 ~xe2 28 'iHb8+ 'It>g7 29 'iVf8 mate) 18 0-0 (18 1:tdl!? deserves attention, e.g. iLd4 19 tDxb5 'be3 20 'bbxd4, Pelikan-Quinteros, Buenos Ai­res 1966) 18 ... 'be3 19 1:tf2 h4 (19 ... 'bg4? 20 'bg5!) and now instead of 20 'bb5? Keres should have tried 20 'bdl! 1:tf7! (Bondarevsky; if 20 ... 'bg4? 21 'bg5! llxf2 [21...'i'xg5 22 'i'b8+] 22'bxf2 ~xg5 23 J..xg4 'bc8 24 'i'a4+- Bartis-Szmetan, Argentina 1970) 21 'i'a5 'bg4 22 'i'xc5;;\;.

b) In 1967 Petrayev invented an improvement 12 ... h4! 13 'be4 'bd7, when Keres's plan doesn't work anymore:

bl) 14 e6?! fxe6 15 dxe6 (After 15 J..d6!? the line pro­posed by Petrayev as good for Black needs practical tests, i.e. 15 ... 'be3 16 ~b3 'bxg2+ 17 ~f2 c4) 15 ... l':.xf4 16 'i'd5 'it'h8 17 'i'xa8 'bb6 18 'i'c6 'be3 19 g3 'bc2+ 20 ~f1 (20 'It>f2? ~d4+ 21 ~g2 'i'xe4 22 J..b5 'be3+ 23 'It>gl 'iHc2-+ Walter­Schmidt, COIT. 1968) and now Black can obtain an advantage with 20 .. J'Ixf3+! (20 ... llf5!?) 21 i.xf3 'bxa 1. Originally this was rejected by some commentators because of 22 'bg5, but in real­ity this is harmless in view of 22 .. :iie7.

b2) It is better to play 14 J..g5!? 'i'b6 (14 .. .f6!?) 15 0-0 c4+! 16 ~hl 'be3 17 .i.xe3 'i'xe3 18 'bd6 'bxe5 19 'bxe5 ~xe5 20 'bxc4 'i'g5 21 l':.c1

Page 76: Beating King's Indian and Benoni

74 Main Line: 9 ... bS and Rare Moves

!iLa6 with approximate equality, Rodriguez-Cuartas, Buenos Ai­res 1973.

b3) Another interesting try is 14 a-a!? ctJgxe5 15 ctJxe5 ctJxe5 16 ctJxc5 ~b6 17 iLe3 ctJd7 18 ~c1.

12 f6 Black can also move his

queen, which is much better placed on b6 than on a5:

a) 12 ... ~a5?! 13 0-0 ctJxe5 (Or 13 ... b4 14 ctJe4 h6 15 !iLe7! ctJe3 16 ~ d2 ctJxfl 1 7 ~xfl !iLf5 18 ctJf6+ ~h8 19 ctJh4 iLc8 20 ~d3! ctJd7 21 ctJxg6+! with a winning attack, Polikarpov­Zvorykina, USSR 1964) 14 d6 (14 ctJxe5!? !iLxe5 15 !iLe7 is also better for White) 14 ... ctJbd7 (White obtains a clear advan­tage after both 14 ... !iLb7 15 !iLe7 ~e8?! 16 ctJxe5 iLxe5 17 d7 ctJxd7 18 ~xd7 ~b6 19 ~xb5, Summerscale-Rudd, British Ch. 1995, and 14 ... ~e6 15 ctJxe5 !iLxe5 16 !iLf3 ctJd7 17 j"xa8 ~xa8 18 ~f3! ~b8 19 ~adl, Blokh-Vexler, USSR 1978) 15

fiLe7 b4 (Better is 15 ... fiLb7±) 16 j"xf8 ~xf8 17 ctJxe5! ctJxe5 18 ~d5! ~b8 19 d7 1-0 Bron­znik -Ma jzlan, Bratislava 1992.

b) 12".~b6!? 13 0-0 leads to the old and, I believe,

the future main line of the 9 ... b5 system.

Black has a number of possi­bilities:

a) 13 ... h6?! 14 iLe7 ctJe3 15 ~d2 c4 16 ~hl ~e8 17 d6± Sakharov-Tukmakov, Moscow 1963.

b) 13 ... ctJxe5?! 14 ctJxe5! (More precise than the immedi­ate 14 fiLe7?! c4+ 15 ~hl ctJbd7! 16 a4 b4 17 a5 ~b8 18 ctJe4 ctJxf3 19 iLxf3 iLa6 20 ~ a4 ~b5! with counterplay, U dovcic-Vasiukov, Yugoslavia­USSR 1963) 14 ... !iLxe5 15 !iLe7 ctJd7 (Or 15 ... ~e8 16 d6 iLe6 17 !iLf3 ctJc6 18 ctJd5 with a strong attack) 16 d6 iLb7 17 lbd5! ~c6 18 !iLf3 ~g7 19 lbc7 ~b6 20 ctJxa8 ~xa8 21 ~hl± Gore­lov-Vasiukov, Moscow 1981.

c) 13 ... ctJd7 (Without ... c5-c4+ either immediately or a lit­tle later, Black experiences

Page 77: Beating King's Indian and Benoni

Main Line: 9 ... b5 and Rare Moves 75

more problems. The important developments in the theory of this line are due to Sosonko) 14 e6 (Note also Arencibia's idea: 14 d6 c4+ 15 ~d4 ~gxe5 16 ~xb6 axb6 17 iLe7) 14 .. .fxe6 15 dxe6 ~xe6 (15 ... ~df6? without ... c5-c4+ doesn't work: 16 e7 1:!,e8 17 iLxb5 1:!,xe7 18 ~d5+-) 16 ~xb5 1:!,b8 (Note that 16 ... ~e3?? loses immedi­ately to 17 ~cl!; 16 ... iLa6? 17 ~c7! is also not good for Black due to 17 ... iLxe2?! 18 ~xe6 iLxd 1 19 1:!,fxd 1 1:!,f7 20 ~d8! +­Sosonko-Hug, Geneva 1977; and 17 ... ~xe2 fails to the con­tinuation 18 Vi'xd7 iLd4+ 19 'it>hl 1:!,f7 20 Vi'c6 1:!,b8, Hovde­Schoppmeyer, COIT. 1983, and now 21 ~xa6!±) 17 ~c7 Vi'f7, and now instead of 18 h3 ~gf6 19 Vi'c2;!;; Baumgartner-Lemaire, COIT. 1994, 18 ~e5!? is more promising.

d) 13 ... c4+ 14 'it>hl ~d7 14 ... h6?! and 14 ... ~xe5?! are

analogous to variations a and b above respectively.

14 ... ~f2+?! 15 1:!,xf2 Vi'xf2 is even worse. Now 16 ~e4 Vi'b6 17 iLe7, hoping for 17 ... 1:!,e8? 18 ~d6 1:!,xe7 19 ~xc8 Vi'd8 20 ~xe7+ Vi'xe7 21 d6!±, can be met by 17 ... ~d7!, when the po­sition is not at all clear. There­fore White must consider 16 ~xb5! and if 16 ... ~d7 17 e6! with a clear advantage.

After 14 ... a6 the game Rytov­Zhuravlev, Tallinn 1973, saw 15 d6! iLe6 16 ~d4 ~xe5 17

iLe7!? ~bd7 18 ~xe6 fxe6 19 iLxf8 1:!,xf8 20 1:!,xf8+ iLxf8 21 ~e4 Vi'e3 22 Vi'c2 and White is better. In the recent game Kahn­Habibi, Balatonbereny 1995, White tried for even more with the combination 17 ~xe6!? fxe6 18 ~d5. The game contin­ued 18 ... exd5 19 Vi'xd5+ ~f7 20 iLe7 Vi'c6 21 iLf3 llixd5 22 iLxd5 ~d7 23 ~xf7 1:!,xf7, and now 24 l1fl ! instead of 24 iLxa8? was promising. Black should have played 18 ... 1:!,xfl+! 19 iLxfl exd5 20 Vi'xd5+ ~f7 21 iLxc4! bxc4 22 l1fl 1:!,a7 23 iLe7 Vi'c6 24 Vi'xf7+ 'it>h8 with an unclear position.

l.i.di 15 d6?! iLb7! is less con­

vincing. 15 .. .fxe6 16 dxe6 ~df6! 16 ... Vi'xe6? is worse, as

shown by Zaltsman-Kalinsky, USSR 1964: 17 ~xb5 l1b8 18 ~fd4! Vi'd5 19 iLxg4 Vi'xg5 20 iLe6+ 'it>h8 21 ~d6±.

17 e7 l:te8

In this critical position White has two ways to a small plus, namely variations d2 and d3:

Page 78: Beating King's Indian and Benoni

76 Main Line: 9 ... b5 and Rare Moves

d 1) Black can equalise with very precise play after 18 a4?! (18 h3?! h6 or 18 ... ttJh6 prom­ises nothing) 18 ... .i.b7! 19 ttJd4 iVcS! 20 .i.xg4 iVxgS 21 .i.f3 il.xf3 22 iVxf3 iVg4! 23 iVxg4 ttJxg4 24 ttJc6, Gorelov-Gleize­rov, USSR 1986, and now, as Gleizerov has shown, 24 ... ttJeS! 2S axbS ttJxc6 26 bxc6 il.xc3 27 bxc3 l:!xe7 results in equality.

d2) 18 ttJd4!? iVcs (18 ... a6 19 a4!) 19 ..txf6 ttJxf6 (19 ... i.xf6? is bad: 20 ttJdxbS il.xc3 21 ttJxc3 ttJe3 22 ~f8+! l:!xf8 23 il.xc4+! ttJxc4 24 exf8iV+ iVxf8 2S iVdS+ il.e6 26 iVxe6+ iVf7±) 20 ttJdxbS! il.e6! 21 ~d4!? (Or 21 iVd6 iVxd6 22 ttJxd6 l:!xe7 23 ..txc4 .i.xc4 24 ttJxc4 ttJe4!= Piskov) 21...iVxe7 (21...iVxd4? 22 ttJxd4 iH7 23 ttJc6 ~ac8 24 il.f3±) 22 il.xc4 .i.xc4 23 iVxc4+ Wh8;!; was Kharkova-Chelushkina, USSR Women's Ch. 1989.

d3) 18 iVd4! .i.d7 An obligatory move. If

18 .. .'i'xd4?! 19 ttJxd4 b4 (19 ... a6 20 ..tf3±) 20 ttJcbS ttJe4 21 .i.xc4+! Wh8 22 il.dS ttJxgS 23 l:!ae 1! il.d7 24 il.xa8 l:!xa8 2S h4 White gets a clear ad­vantage, Blokh-Gunnas, corr. 1994.

19 iVxb6! 19 l:!ae1 .i:.ac8! 20 iVxb6

axb6 21 h3 was Kaidanov­Gleizerov, Smolensk 1986, and now 21...h6! is unclear.

19 ... axb6 20 ttJd4 h6 21 .txf6 ttJxf6 22 ttJdxbS llxe7 23

il.xc4+ Wh7

Here White is slightly better (Kaidanov).

13 exf6! This move is good enough for

White, so it is not necessary to investigate the huge complica­tions which arise after 13 d6!? very deeply. Nevertheless, here is one fantasy variation: 13 .. .fxgS 14 iVdS+ Wh8 IS 'iYxa8 iVb6 16 ttJdS! iVaS+ 17 b4 cxb4 18 0-0 ~fS 19 ttJxgS! ttJd7 20 ttJf7+ Wg8 21 ttJe7+ Wxf7 22 iVdS+ with a winning attack.

13 il.xf6 14 iVd2

Page 79: Beating King's Indian and Benoni

Main Line: 9 ... b5 and Rare Moves 77

14 lle8!? Before this game this was the

only move that had posed any problems for White. The alter­natives are:

a) 14 ... Si.xgS?! IS tLlxgS! (1S 'i'xgS is slightly better for White, but he wants more) IS .. .tt:Ja6 (Or IS ... b4 16 tLlce4 'i'e7 17 O-O-O± Guidi-Troia, COIT. 1991) 16 h3! tLleS 17 d6! with a clear advantage for White, Blokh-Krasnov, USSR 1986.

b) 14 ... iHS Here Black has succeeded in

neutralising a dangerous idea of Sosonko's:

b 1) IS tLlxbS tLld7 White maintains a clear ad­

vantage after IS ... ~b6 16 d6! tLlc6 17 Si.c4+ ~h8 18 0-0 tLlaS 19 b3! tLlxc4 20 bxc4 a6 21 tLlc7 llad8 22 lladl± Karlsson­Meyer, Uppsala-Bremen 1977, or after IS ... a6 16 tLlc3 lla7 17 h3 tLleS 18 Si.xf6 ~xf6 19 0-0 tLlbd7 20 g4!, Konikowski­Stanojevic, Chianciano Team Ch.1989.

16 O-O?! ~b6 17 Si.c4 tLldeS 18 tLlxeS tLlxe5 19 d6+ tLlxc4 20 ~dS+ ~g7 21 Si.xf6+ :'xf6 22 ~xc4

(see following diagram)

Now Reshevsky made a mistake against Sosonko (Amsterdam 1977): 22 ... llb8? 23 llael! llf7 24 lle7 llxe7 2S dxe7 with a clear advantage for

White. In the game Gomjak­Shemagonov, USSR COIT. Ch. 1992, Black improved on this and equalised after 22 ... llaf8! 23 llael Si.e6 24 ~d3llxfl+ 2S llxfl llxfl + 26 ~xfl Si.d7 27 a4 a6 28 tLlc3 ~xb2.

Still, White is better if he plays 16 d6!? before castling.

b2) Another good option for White is to avoid wasting time in grabbing the black pawn on bS, as in the game Semkov­Tasic, Cannes open 1989: IS O-O!? b4 16 tLld 1 tLld7 17 tLlf2! tLlxf2 18 lhf2 .te4 19 .txf6! tLlxf6 20 tLlgS 'i'xdS (20 ... i.xdS 21 lldl±) 21 tLlxe4 ~xe4 22 .tf3 1-0.

15 O-O! The active position of

Black's rook makes IS tLlxbS? inappropriate. After IS ... .txgS! 16 ~xgS (Even worse is 16 tLlxgS?! tLle3! 17 tLle6? Si.xe6! 18 dxe6 ~h4+ 19 g3 ~e4) 16 ... llxe2+! 17 ~xe2 iVe8+ 18 ~d2 iVxbS 19 iVd8+ rJitg7 20 'iYxc8 ~xb2+ 21 ~d3, when apart from the perpetual check (with 21...~bS+), Black can

Page 80: Beating King's Indian and Benoni

78 Main Line: 9 ... b5 and Rare Maves

play for the attack with 21.. .ctJa6!? Here and later we have used and adapted some of Blokh's comments on his game.

IS b4 The game Neuman-Haller­

ova, Czech Republic Team Ch. 1996, continued 15 ... iLxg5 16 ctJxg5 ctJe3 17 ctJf7 (17 d6!?) 17 ... ~h4 (l7 ... ctJxfl!?) 18 :f3 ~d4 19 ~xd4 cxd4 20 ctJxb5 ctJc2 21 :c1 :xe2 22 ctJc7 with a clear plus for White.

16 ctJdl iLb7 17 ctJf2! iLxgS?!

17 ... ~xd5? was bad: 18 ~xd5+ iLxd5 19 ctJxg4 iLxg5 20 ctJxg5 :xe2 21 ctJf6+ ~g7 22 ctJxd5 :e5 23 :f7+ ~h6 24 h4+-. It was better to play 17 ... ctJxf2 18 :xf2 ctJd7 (Again the d5-pawn is poisoned: 18 ... iLxg5? 19 ctJxg5 ~xd5 20 ~xd5+ ~xd5 21 :dl .Jtxa2 22 .Jtf3 iLb3 23 :d6! ~c4 24 h4 ctJa6 25 :xa6!+-) 19 d6 with an advantage for White.

18 ctJxgS! ctJf6

The following thematic variation is convincing enough: 18 ... ctJe3 19 ctJfe4! ctJxfl 20 :xfl :f8 21 ctJe6 :xfl + 22 iLxfl ~e7 23 ctJ4xc5+-.

19 ctJg4! ctJxg4 19 ... ctJbd7 loses immediately

to 20 iLc4!, or 19 ... ctJxd5 20 ctJh6+.

20 iLxg4 ~xdS 21 ~f2!

The threat of 22 .Jte6+ forces a transition into a winning end­ing.

21 22 23 24 2S 26 27 28 29 30 31 32

~xd4 iLe6+ ctJxe6 :f6 :afl :n :g7+ :xd7 :xd4 :xb4 :b7

~d4 cxd4 :xe6 ctJc6 ctJeS ctJd7 iLdS ~h8 iLxe6 .Jtxa2 as 1-0

Page 81: Beating King's Indian and Benoni

4 Black plays 6 ... CtJa6

Now we move on to examine early deviations along the way to the main line of the Four Pawns Attack after the first few moves of the King's Indian De­fence. The material in this sec­tion is divided into three chap­ters: this chapter deals with 6 ... lba6!?, Chapter 5 covers 6 ... c5 7 d5 b5, while Chapter 6 deals with other systems. The first of these is undoubtedly the best choice for Black.

I d4 lbf6 2 c4 g6 3 lbc3 ~g7 4 c4 d6 5 f4 0-0 6 lbf3 lba6

Now one of Black's main weapons aQ'ainst the Four

Pawns Attack, 6 ... lba6 was in­troduced into modem tourna­ment practice by Igor Belov in 1987. Black prepares ... e7-e5 (immediately or after ... .i.g4), while keeping open the option of playing ... c7-c5. This system has one advantage for died-in­the-wool King's Indian players: the resulting positions are often more in keeping with the spirit of that opening than are other variations of the Four Pawns Attack.

Initially I experienced prob­lems after 6 .. ,cZJa6 and my re­sults were disastrous. I lost the first two games I played against it, and my total against 6 ... lba6 (+2=5-2) is awful compared to my general results with the Four Pawns Attack in the period 1982-1996 (+31=10-2). Never­theless, nowadays White has several ways of obtaining a pleasant position. The line with 7 ..id3 is covered in the Game 12 (Zsu. Polgar-Chiburdanidze), while 7 e5!? and other lines are dealt with in Game 13, Vaisser­Golnhev

Page 82: Beating King's Indian and Benoni

80 Black Plays 6 . ./iJa6

Game 12 Zsu. Polgar-Chiburdanidze St Petersburg (match) 1995

1 d4lbf6 2 c4 g6 3 lbc3 iLg7 4 e4 d6 5 f4 0-0 6lbf3 lba6

7 iLd3 iLg4 Both this move and 7 ... e5

(considered in variation c be­low) are sound for Black. The alternatives are less precise:

a) After 7 ... lbd7?! 8 0-0 e5 9 fxe5 dxe5 10 d5 ~e7 we reach a pawn structure typical of the ... lba6 line, but in this case the d7 -square is not a good place for Black's knight. The game Tsedev-Sammalvuo, Bratislava 1993, continued 11 a3 c6 12 b4 lbc7 13 ~b3 cxd5 14 cxd5 ~h8 15 .tg5 f6 16 iLh4 with advan­tage to White.

b) 7 ... c5?! 8 d5 e6 9 0-0 (9 dxe6;J;;) 9 ... exd5 10 exd5 trans­poses into the line 9 exd5 (see Game 16), but with an extra tempo for White (.te2-d3). Furthermore, Black's move ... lba6 would not be his first choice. These small differences swing the evaluation of the po­sition in White's favour. One example: 1O ... lbc7 11 a4 lba6 12 f5! lbb4 13 iLbl a6 14 ~el lbg4 15 iLg5 f6 16 iLd2 g5 17 'ii g3 h5 18 h3 lbh6 19 lbxg5! h4 20 ~xh4 fxg5 21 .1Lxg5± Tarasov-Krishilovsky, St Pe­tersburg 1995.

c) 7 ... e5!? 8 fxe5 8 dxe5?! is weaker: 8 ... dxe5 9

lbxeS (9 fxe5?! lbcS! 10 iLc2 ~xdl+ 11 ~xdl lbg4 12 ~e2 lbxeS+, Sokolin) 9 ... lbc5 10 iLe3 lbxd3+ 11 ~xd3 ~xd3 12 lbxd3 ':e8 13 lbe5 (If 13 e5 lbg4 14 .tgl .tf5 15lbc5 b6 16 lba6 .txe5! with a fine initia­tive) 13 ... lbg4 liz-liz occurred in the game Ivanov-Sokolin, Len­ingrad 1991. The position is equal after 14 lbxg4 iLxc3+ 15 bxc3 iLxg4 16 e5 f6.

8 ... dxe5 9 d5 c6 Black immediately tries to

undermine White's strong cen­tre. Another plan is to prepare counterplay on the kingside while restraining White's pawns on the opposite wing with 9 ... ~e7 (The immediate block­ade with 9 ... lbc5 10 .tc2 a5 is not a good choice, as White can prepare b2-b4 with an advan­tage: 11 0-0 lbh5 12 .te3 ~e7 13 a3 f5 14 b4, Kalousek­Jirovsky, Mlada Boleslav open 1995) 10 0-0 lbh5. Black's plan worked relatively well in the game Parker-Volke, Copenha­gen 1996, thanks to the help of his opponent: 11 a3?! lbf4 12 b4 (12 lbxe5?! ~c5+ 13 ~hl lbxd3 14lbxd3 ~xc4+) 12 ... c5! 13 b5 lbc7 (13 ... lbxd3 14 ~xd3 lbc7) 14 ':el (14 .txf4! exf4 15 e5 .txe5 16 ':e1 f6 17 .txg6! hxg6 18 lbxe5 fxe5 19 d6;!;) 14 ... lbxd3 15 'iixd3 lbe8 16 .te3 b6 17 a4 lbd6 18 a5 ':b8 19 axb6 axb6 20 ':a4 f5 21 lbd2 ~h4 with an initiative. As an improvement I suggest 11

Page 83: Beating King's Indian and Benoni

.tIel!? tDf4 12 it.fl. After 9 ... c6 White has a

choice: c1) Black has comfortable

equality after 10 .tg5?! h6! (1O ... "iib6?! 11 tDa4! "iVa5+ 12 ~d2 "iVd8 13 o-o;t 13 ... cxd5?! 14 cxd5 tDe8 15 .txa6 bxa6 16 :tb4 tDd6 17 l:i.c1 l:i.b8 18 "iiel± Garcia Palermo-Danailov, Ali­cante open 1992) 11 .i..xf6 (Not 11 i.h4? "iib6! 12 'i¥e2 tDc5 with an initiative) 11.. . .txf6 12 bIb 1 tDc7 13 0-0 cxd5 14 cxd5 4Je8, as in the game Garcia Pal­ermo-Comas Fabrego, Ibercaja open 1992.

c2) Interesting complications occurred in the game Arencibia­Pecorelli, Santa Clara 1996, with White coming out on top: 10 "iie2 tDc5 11 it.c2 cxd5 12 cxd5 b6 13 b4 .ta6 14 'i¥dl ctJfxe4 15 .txe4 ctJxe4 16 tDxe4 f5 17 tDeg5 e4 18 ctJe6 "iVe7 19 tDfd4 f4 20 "iib3 .td3 21 h4!+-.

c3) 10 O-O! cxd5 1O ... 'i¥b6+ is also possible

(Arencibia's 10 "iVe2!? allows White to avoid this possibility). After (10 ... "iib6+) 11 ~hl tDg4 12 "iie2 tDc5 13 .i..c2 f5, how­ever, White should not be too afraid of the attack, which is a little too direct and overly ag­gressive. Now, besides the modest 14 l:i.bl fxe4 15 tDxe4 cxd5 16 tDxc5 "iixc5 17 cxd5 i.f5 18 i.xf5 l:i.xf5 (l8 ... gxf5 19 tDh4!) 19 h3 tDh6 20 .txh6 i.xh6 21 .l::.tbdl l:!.e8 22 d6! ~d8 23 "iid3, Giardelli-De Souza,

Black Plays 6 ... tDa6 81

Mar del Plata open 1995, when 23 ... ~h8! would given Black an approximately equal position, White has a more energetic op­tion at his disposal, namely 14 ex/51 gxf5 (l4 ... .i.xf5 15 .i.xf5 gxf5 [15 ... l:i.xf5 16 h3±] 16lbh4 tDd3! [16 .. . lbe4 ?! 17 lbxe4 fxe4 18 tDf5±] 17 "iixd3 tDf2+ 18 ~xf2 'ilf'xf2 19 g3 with a White advantage) 15 h3! tDf6 (l5 ... e4 16 hxg4 exf3 17 l:i.xf3±) 16 .i.e3 f4 (16 ... tDfd7 17 tDa4 'iia5 18 lbxc5 tDxc5 19 a3 lbe4 20 lbxe4 fxe4 21 tDg5±) 17 .tf2 (17 tDa4!?) 17 ... lbh5 18 tDe4 tDxe4 19 "iixe4lbg3+ 20 .i.xg3 i.f5 21 'iixf5! l:i.xf5 22 .i.f2 and the conclusion is that White has picked up too many pieces for the queen.

11 cxd5 Now Black has three plausi­

ble continuations: c31) After the exchange on

d5, Black's attack from the pre­vious note is senseless, as White's bishop has c4 available to it, e.g. 11..."iib6+ 12 ~hl lbg4 13 "iie2 lbc5?! 14 .tc4! i.d7 15 h3 tDf6 16 .i.g5 tDh5 17 "iif2 tDa4 18 lbxa4 'ilf'xf2 19 l:i.xf2 .i.xa4 20 ~el± Pribyl­Beckhuis, German Bundesliga 1993/94.

c32) 11...b5?! is original, but weakens the queenside and can­not be recommended. In the game Parker-Burgess, English Team Ch. 1996, White was clearly better after 12 i.g5 (12 .txb5? lbxe4!) 12 ... b4 13 lba4

Page 84: Beating King's Indian and Benoni

82 Black Plays 6 ... lba6

h6 14 i.xf6 i.xf6 15 l;Ic1 i.g7 16 'iVe2 i.d7 17 b3 i.xa4 18 bxa4 lbc7 19 i.b5! lbe8 20 i.xe8 'ii'xe8 21 l;Ic6.

c33) 11...lbe8

This is one of the critical po­sitions of the 6 ... lba6 system. White plans to break the block­ade of the d6-square, thereby strengthening his passed d­pawn. Simultaneously he tries to diminish Black's counterplay on the kingside. I believe that White's prospects are slightly better.

Practical examples from this position are:

c3 31) White discourages his opponent from playing ... f7-f5: 12 'iVe2 lbac7 (After 12 ... lbc5 White should not allow his bishop to be exchanged: 13 i.c2 b6 14 M i.a6 15 b5;!;) 13 i.g5 f6 14 i.h4 'ii'e7 15 ~hl?! (The immediate 15 lbd2 was more precise) 15 ... lbd6 16 lbd2 iLd7 17 a4 i.h6 18 lbc4 lbce8 19 i.f2 lbxc4 20 i.xc4 lbd6 21 i.c5 l;Iac8 22 i.a3 l;If7 23 i.b5 a6 24 i.d3! 'it>h8 25 'ii'f2, Clement-Romero, Matalascanas

1995. Even after losing a tempo on move 15, White has a small advantage.

c332) White allows Black's counterplay: 12 i.e3!? lbd6 13 'iVb3 f5 14 lbd2 (An interesting alternative was 14 lbg5!? with the idea of 14 .. .f4 15 lbe6!) 14 .. .f4 15 i.f2 g5 16 lbc4 g4, Ooi Chern Ee-Braga, Yerevan Olympiad 1996, and now White should have played 17 lbxd6 'ii'xd6 18 lbb5 with good pros­pects in this complicated posi­tion.

c333) 12 i.g5!? may trans­pose to variation b after 12 .. .f6 13 i.e3, but also sets a nice trap, since 12 ... 'iVb6+? is a mistake: 13 'it>hl f6 (l3 ... 'ii'xb2? 14lbb5! 'ii'M [14 .. .f6? 15 i.d2! and 16 i.c3±] 15 a3 'iVc5 16 l;Ic1 winning the exchange) 14 lba4! 'iVd8 15 i.h4 lbac7 16 'iVb3 lbd6 17 l;Iac1 (By playing ... ~d8-b6-d8 Black has lost two tempi and White is clearly bet­ter; the Black's next move al­lows a pleasing and decisive combination) 17 ... 'iVe7? 18 lbxe5!! i.h6! 19 lbc4! b5 (Or 19 ... i.xc1 20 lbxd6 'iVxd6 21 i.g3 'ii'd7 22 l;Ixc1 with a large advantage) 20 lbxd6 bxa4 21 lbxc8 l;Ifxc8, Lalic-Arakhamia, Staffordshire 1997. Now White could have obtained a winning ending after 22 d6+! axb3 23 dxe7lbe8 (23 ... i.xc1? 24 i.c4+ 'it>g7 25 i.xf6+ ~h6 26 l;If3 g5 27 i.f7 g4 28 l;If5 mating) 24 l;Ixc8 l;Ixc8 25 axb3.

Page 85: Beating King's Indian and Benoni

8 0-0 The most precise. Others: a) 8 h3 i.xf3 9 'iVxf3 eS!

(Weaker is 9 ... lLld7?! 10 i.e3 eS 11 dxeS dxeS 12 fS transposing to the game Braga-Comas - see note bl below) 10 fxeS lLld7 11 0-0 cS 12 lLldS cxd4 13 lLlf6+ ~h8 (13 ... i.xf6!?) 14 lLlxd7 ~xd7 IS exd6 'ii'xd6 16 i.f4 'We6 17 a3 lLlcs 18 b4 lLlxd3 19 ~xd3 1:!fe8 and Black has no problems, Weglarz-Kaminski, Gdansk 1994.

b) Ue.3 and now: bl) 8 ... lLld7?! 9 h3 ..txf3 10

if xf3 eS 11 dxeS! dxeS 12 fS lLldcS 13 i.e2;!;;. Braga-Comas, lbercaja open 1992, continued 13 ... gxfS 14 exfS e4 IS iVg3 ~h8 16 ndl lLld3+ 17 i.xd3 exd3 18 0-0 1:!g8 19 ~f3 with advantage to White.

b2) 8 ... eS?! 9 fxeS dxeS 10 dS lLlhS 11 O-O;!;; lLlf4?! 12 ..tc2 i.h6?! 13 iV d2 i.xf3 14 ~xf3 i.gS IS ~dl h6 16 h4! ..txh4 17 g3 iVgS 18 iVh2 'iVg4 19 ~f2 lLlb4 20 ..tbl+- Gretarsson-

Black Plays 6 .. . lLla6 83

Meesen, European Club Cup, Eupen 1994.

b3)Lc.5l The best way to exploit the

vulnerability of the bishop on e3.

9 dS e6 10 dxe6 Or 10 0-0 exdS 11 exdS !le8

12 i.d2 iVd7! 13 a3 lLlc7 with equal chances, Vaisser-Gallagh­er, Swiss Grand Prix 1990.

1O ... ..txe6! 10 .. .fxe6?!;!;; is worse.

After 1O ... i.xe6 the Black threat is 11...lLlg4 and 12 .. .fS. He has enough counterplay after 11 h3 'ifb6, so the critical line is 11 fS ..txfS ! 12 exfS :re8 13 ~d2 dS! 14 cxdS lLlxdS IS i.gS iVaS 16 iVb3 c4! with a very strong attack for the piece ( Gallagher).

8 lLld7 Instead the typical thrust

8 ... eS?! is premature after 9 fxeS and now:

a) 9 ... lLld7 10 iVe1! i.xf3 (lO ... cS? [lO ... dxeS 11 dS;!;;/±] 11 ..tgS! iVaS 12 iVh4 ..txf3 13 ~xf3 cxd4 [Or 13 ... dxeS 14 dS±] 14 lLldS with a decisive

Page 86: Beating King's Indian and Benoni

84 Black Plays 6 . ./Da6

attack, for example: 14 ... tbxe5 15 lih3 h5 16 tbe7+ ~h8 17 ~xh5+! gxh5 18 Itxh5+ ~h6 19 llxh6+ ~g7 20 tbf5+ ~g8 21 ~f6 tbg6 22 tbe7+ tbxe7 23 1ih8 mate, I. Belov) 11 gxf3 (Interesting is 11 lIxf3 c5 12 tbd5 cxd4 [12 ... dxe5 13 dxc5t] 13 exd6 with an unclear posi­tion. The game Gretarsson­Pedersen, Hallsdall open 1996, continued 13 ... 1ie8?! 14 b4!? tbe5 15 c5 f5 16 ~b2 tbxf3+ 17 gxf3 ~g5+ 18 ~hl ~e5 19 ~f2 with a difficult position for Black) 11...dxe5 12 d5, Hubert­Belov, Porz 1995. Here Belov proposed as an improvement 12 ... tbf6!? 13 ~e3 tbh5 14 ~hl tbf4 15 i.c2 c5t.

b) 9 ... dxe5 10 d5 is also good for White. Compared with the 7 ... e5 line we can see that the move ... ~g4 is a waste of time. This difference changes the evaluat~on in White's fa­vour. One example is 10 ... c6?! 11 ~e3 cxd5 12 cxd5 ~xf3?! 13 ~xf3 tbe8 14 tbb5± Irzhanov-Doornbos, Bratislava 1993.

9 i.e3 9 d5?! is ineffective: 9 ... c6 10

i.e3 tbac5 11 ~c2 ~b6 12 lIbl ~b4! 13 ~b3 tbxb3 14 ~xb3 a5+ Kahn-Panzer, Budapest 1993.

9 eS 10 fxeS cS!

The point! Now White is obliged to play 11 d5 and Black gains control of the e5-square.

Even so, the knight on a6 is a long way away from this key square, which allows White time to create problems for his opponent in the centre.

11 dS 12 ~e2

tbxeS tbxf3+!

Belov has tried several other options, but without much suc­cess:

a) 12 ... ~xf3 13 gxf3 f5 (13 ... tbc7!?t) 14 f4 tbn (Or 14 ... tbd7!?) 15 exf5 gxf5 16 ~hl ~e7 17 lIf3 with an ad­vantage for White, Petronic­Belov, Pravets 1989.

b) 12 ... ~d7 13 ~d2 tbc7 14 ~g5 ~e8 15 ~hl a6 16 a4 b6 17 ~el f6 18 ~d2 tbn 19 i.d3 b5 20 b3 b4 21 tbe2 ~e7 22 ~g3 1iae8 23 1iael tbe5 24 i.c2 and White has more space, Arkhipov-Belov, Moscow 1987.

13 ~xf3 i.xf3 14 ~xf3

After the exchange of White's light-squared bishop, 14 gxf3 is no longer dangerous.

14 ~e7

Page 87: Beating King's Indian and Benoni

Black does not have enough time to bring her knight to eS: 14 ... tDb8?! IS iLf4! (This also follows 14 ... tDb4) IS ... j.d4+ 16 ~hl f6 17 tDbS ..teS 18 j.xeS dxeS 19 'iYh3! with a clear ad­vantage for White.

15 j.f4 tDc7 Necessary to stop the un­

pleasant threat 16 tDbS. 16 'iYg3 ~ad8 17 ~hl!

Preparing 18 j.gS. The im­mediate 17 j.gS?! is ineffective because of 17 ... ..td4+! 18 ~hl f6.

17 j.d4! 18 ~ael f6!

Black had to stop the threat of 19 eS!

19 tDe2 iLe5 Of course not 19 ... ..txb2? 20

~bl iLeS 21 ~xb7±. 20 liJgl

The knight heads for a more active post on f3. Both sides have played well up until now and White has succeeded in keeping a small initiative.

Black Plays 6 ... liJa6 85

20 a6?! An unnecessary move. It

would have been better to ad-vance with 20 ... bS!? (Zsu. Pol-gar) or play 20 ... tDa6!? 21 tDf3 ..txf4 22 'iYxf4 liJb8, approach­ing the critical e5-square, with chances of equality.

21 liJf3 ..txf4 As shown by Zsuzsa Polgar,

21...b5? is bad because of 22 tDxe5 fxeS 23 iLh6!± (But not 23 ..tg5? :xf1+ 24 ~xf1 ~f8=).

22 'i'xf4 b5 23 b3 ~b8?

This allows a nice pawn break. White keeps a small ad­vantage after both 23 ... bxc4 24 bxc4 ~b8 25 ~b 1 lUe8 26 ~b3 and 23 ... b4 24 g4, preparing g4-g5.

24 b4! cxb4 25 c5 ~bd8

Or 2S ... dxcS 26 d6 'iYd7 27 dxc7 ~bc8 28 ~dl 'i'xc7 29 'iYxc7 ltxc7 30 ltd6±.

26 liJd4 Also possible was 26 c6 with

a stable advantage.

Page 88: Beating King's Indian and Benoni

86 Black Plays 6 ... lba6

26 27 lbc6 2S lbxdS 29 .:tel?!

dxc5 'it'd7 'it'xdS

A more straightforward path was 29l:idl c4 (29 ... 'iVd7 30 e5! lbxd5 31 ~f3 :t.d8 32 exf6 ~f7 33 lld2+-) 30 d6 lbe6 31 ~d2 a5 32 d7!±.

29 30 d6?!

30 !tcd1!? 30 31 ~d2

c4

lbe6 'it'd7?

31.. .a5! (Ftacnik) was neces­sary: 32 'iYd5 (Not 32 d7? :t.f7 33 ':'cdl lbf8=F) 32 ... 'iYd7 and Black has managed to blockade the white pawns.

32 'iVxb4

Now White is winning with-out any major problems.

32 lbg5 33 'iYc5 liteS 34 l:tce1! l:te6 35 e5 f5 36 a4! lbf7 37 axb5 axb5 38 'ilVc7! ~xc7

39 dxc7 l:lc6 Or 39 ... :t.e8 40 e6 lbd6 41

l:.dl and White wins (Zsu. Pol­gar).

40 41 42 43 44 45

e6 e7 e8'it'+ llxeS+ l:tbS ~gl

lhc7 lbd6 lbxe8 ~f7 c3 1-0

Game 13 Vaisser-Golubev

Biel open 1995

1 d4 lbf6 2 c4 g6 3 lbc3 i..g7 4 e4 d6 5 f4 0-0 6 lbf3 lba6

7 e5!? Other possibilities for White

on the 7th move (apart from 7 i..d3, which was considered in the previous game) are:

a) 7 c5?! (An interesting idea of Riedel's which may, how­ever, be dubious) 7 ... dxc5 8 d5 e6 9 i..xa6 bxa6 10 0-0 exd5 11 e5 lbe4 12 'iYxd5 'iYxd5 13 lbxd5 i..b7 14 lbe3 l:lad8 is slightly better for Black, Riede1-Held, Munich 1993.

b) 7 i..e3?! (In this system the bishop is often vulnerable on e3. Here Black can take ad­vantage of this by energetic play) 7 ... c5! 8 d5 e6! 9 dxe6 i..xe6! (An idea of Prie's). Compared to the line with 7 i..d3 we have a similar position but with the bishop placed on e3 instead of d3, a difference which favours Black. Black has

Page 89: Beating King's Indian and Benoni

excellent active play. c) 7 iLe2 eS!

and now: c1) 8 0-0 is hannless due to

8 ... exd4 9 liJxd4 liJcs 10 iLf3 ~e8 11 ~e1 ~g4!? 12 liJb3 ~xf3 13 'i'xf3 liJxb3 14 axb3 c6 IS iLe3 with pleasant equal­ity for Black.

c2)8..J::lill offers no more than equality 8 ... dxeS and now: c21) 9 liJxeS liJcs 10 iLf3 (10

'i' c2 ?! liJfxe4! 11 liJxe4 iLfS 12 iLd3 ..txe4 13 iLxe4 iLxeS 14 fxeS 'iYh4+ IS g3 'i'xe4++, Belov) 1O .. :~Wxd1+ 11 <;t>xd1 ~d8+ 12 <;t>c2 (12 <;t>e2?! ..te6 13 iLe3 liJfd7! 14liJdSliJxeS IS fxeS liJd7 16 iLgS iLxdS! 17 cxdS ~e8 18 iLf4 ..txeS 19 iLxeS liJxeS+ Gorelov-Be1ov, Moscow 1987) 12 ... liJfxe4 13 liJxe4 ~fS 14 ~e 1 iLxeS IS fxeS .l:Id4 16 b3 (16 <;t>c3 .l::rd3+ 17 <;t>b4 [17 <;t>c2=] 17 ... liJa6+ 18 'it>aS b6+ 19 <;t>xa6 iLc8+ 20 'it>bS ~d7+=) 16 ... liJxe4 17 'it>b2 liJc5 18 <;t>a3 liJd3 19 .l::rd 1 c6 20 g4 iLe6= Namgilov-Seoo,

Black Plays 6 ... liJa6 87

Rostovopen 1993. c22) 9 'i'xd8 lhd8 10 liJxeS An interesting alternative is

10 fxeS!? liJg4 11 iLf4 .l::re8 12 :.td 1 c6 13 h3 liJxeS 14 liJxeS ~xe5 IS ~xe5 .l::rxeS 16 1:.d8+ <;t>g7 17 0-0 l:tb8 18 b4 ~e6 19 .l:Ixb8 liJxb8 20 cS with ap­proximately equality as in the game Belakovskaya-Winslow, New York 1993.

lO ... liJcS 11 iLf3 Or 11 liJdS c6 12 lLle7+ ~f8

13 liJxc8 :.taxc8 14 i.e3 liJfxe4 IS 0-0 f6 16 lLlf3 fS 17 liJeS with equality, Chiburdanidze­Xie Jun, World Women's Ch., Manila 1991.

11...~e6 This move gives White chan­

ces to go wrong. Black immedi­ately forced a drawn position with 11...lLlfd7 12 lLlxd7 .i.xc3+ 13 bxc3 ..txd714 0-0 iLc6 IS ~e3 liJxe4 16 ..txe4 in Vera­Bass, Barcelona 1990.

In this critical position White has tried:

c221) 12 iLe3? liJfd7! (Mis­taken is 12 ... liJd3+? 13 liJxd3 ~xd3 14 ~f2;j; 14 ... iLxc4? IS

Page 90: Beating King's Indian and Benoni

88 Black Plays 6. JLJa6

eS!±) 13 0-0 (Or 13 llJxd7 llJd3+ 14 ~f1 ~xc4! and Black is better) 13 ... llJxeS 14 fxeS llJd3 IS llJdS ~xeS+ Sutter­Gallagher, Suhr 1992.

c222) 12 0-0 llJfd7! 13 llJxd7 .td4+! 14 ~hl ~xd7 IS llJdS c6 16 .te3! cxdS 17 .txd4 dxe4 18 .txcS exf3= Solovian-Be1ov, Podolsk 1990.

c223) 12 llJdS llJfd7! 13 llJxd7 (The interesting compli­cations which arise after 13 CtJxc7 llJxeS 14 fxeS llJd3+ IS ~f1 ~xc4 16 .tgS CtJxeS+ 17 .te2 f6 18 llJxa8 fxgS are in Black's favour) 13 ... ~xd7, and now White should play 14 0-0, with a likely transposition to variation c222. 14 ~e2?! is in­ferior due to 14 ... ~e8! IS eS c6+ Vaisser-Weindl, Mendrisio open 1989.

c3) The most popular choice, and key to the evaluation of 7 ~e2, is the line

8 fxeS dxeS 9 dS 9 llJxeS?! cS! 10 .te3 (10 dS

llJxe4 11 CtJxe4 ~xeS+, Belov) and now the game Komarov-VI. Georgiev, Benasque 1996, con­tinued 1O ... llJg4? l1llJxg4 cxd4 12 llJh6+! ~h8 13 .tf2 llJb4 14 llJdS CtJxdS IS exdS ~aS+ 16 ~f1 ~xh6 17 ~xd4+ with a clear advantage for White. In­stead Black should have seized the initiative by playing lO ... CtJb4!? or 1O ... cxd4 11 .txd4 llJb4!? with threats of 12 ... ~xd4! and 12 ... llJc2!

After 9 dS the position is

similar to the corresponding one in the line 7 .td3, but with some differences.

Black has two good possi­bilities:

c31) 9 ... CtJcS!? 10 .tgS (10 ~c2?! llJfxe4! 11llJxe4 .tfS 12 .td3 ~xe4 13 ~xe4 fS 14 .txfS gxfS+ Kavalerov-Kochiev, St Petersburg 1994) 1O ... h6! (Not lO ... aS? 11 CtJd2±) 11 ~xf6 (11 ~h4? gS 12 ~g3 CtJfxe4 13 CtJxe4 llJxe4 14 .txeS .txeS IS CtJxe5 c6! is in Black's favour) 11...~xf6 12 b4llJa6 13 ~bl c5 14 a3 ~b6 (Even world cham­pions make mistakes in the Four Pawns Attack! Against Lautier, Amsterdam 1995, Kasparov played 14 ... ~d7? here and would have had big problems after the correct 15 d6!) 15 ~ d2 ~d8 16 h4 hS 17 b5 CtJc7 18 ~e3 ~d6 190-0 and Black was at least equal in O. Rodriguez­Dorfman, Barcino 1994, though I believe that White's play could have been improved .

c32) 9 ... c6 10 0-0 Instead White can play 10

.tgS (10 dxc6 ~b6! with good

Page 91: Beating King's Indian and Benoni

counterplay or 10 ii.e3? 4Jg4) 1O ... h6 11 .th4 'ikb6 12 'i'd2 4Jc5 13 .txf6 .txf6 14 iVxh6 (14 l:!bl?! cxd5! 15 cxd5 J..g7 16 b4 4Ja6 17 4Ja4 'i'd6+ Bagaturov-Lukin, Frunze 1989) 14 ... 'i'xb2 15 l:!c1 .i.g7 16 'i'e3 f5 17 0-0 iVa3! with compli­cated play, Armstrong-Milligan, COIT. 1994.

1O ... cxd5 11 cxd5 iVb6+ 12 ~hl

At Linares 1995 Kaminski played 12 ... 4Je8?! against me here. Play continued: 13 4Jd2 (13 b4!?) 13 ... 4Jd6 14 4Jc4 4Jxc4 15 .txc4 'iYd4! 16 'iYxd4 (Possible are 16 'iYe2!?;l; or 16 ii.b5!?;l;) 16 ... exd4 174Jb5 iLd7 18 .tf4 .txb5 19 .txb5 4Jc5 20 l:i.fel d3. Now after 21 .l:iac1!? .te5 22 ii.d2 a6 23 l:!xc5 b6! White would have had a small advantage.

Six months later we met again in Bie1, where he un­leashed an improvement:

12 ... 4Jg4!? 13 iVel4Jb4!

A strong novelty! 144Ja4 'iYd6 Now the critical line IS 15

Black Plays 6.Jba6 89

h3!? 4Jc2 (15 ... 4Jf6? 16 'i'h4) 16 iVc3 4Jxal 17 hxg4 f5! 18 .i.e3! with a very complicated position, where White is proba­bly not better.

7 4Jd7 a) After 7 ... 4Je8?! the reply 8

h3! (8 c5!?) highlights the pas­sivity of Black's move and al­lows White to retain an advan­tage due to his strong centre.

b) Also 7 ... dxe5? cannot be recommended: 8 fxe5 4Je8 (8 ... 4Jd7 9 ii.f4) Leitao-Ivano­vic, Yerevan Olympiad 1996, and now 9 c5!?

c) 7 ... 4Jh5?! looks dubious but is not easy to refute. After 8 .te2 (8 ii.e3!?) 8 ... ii.h6!? White should avoid the over-optimistic 9 f5?! ii.xc1 10 'i'xcl ii.xf5 11 'iYh6 f6 12 g4 .i.xg4! 13 l:!gl .txf3 14 .txf3 4Jg7 15 .txb7 'iYb8! with an unclear position, Vokac-Babu1a, Lazne Bohdanec 1996. Instead White should play 9 g3 f6 10 0-0 with a small ad­vantage. For example: 10 ... dxe5 11 fxe5 .i..xc1 12 l:!xc1 .th3 13

Page 92: Beating King's Indian and Benoni

90 Black Plays 6 ... 0.a6

':'el;l; (Movsesian). 8 .i.e2!?

White has a number of other potential moves, but only one of them (8 h4!?) is playable:

a) 8 cS? dxcS 9 dS 0.b6 10 a3 (If 10 .te3 0.b4 White loses the dS-pawn for insufficient com­pensation) 1O ... e6!? (10 .. .f6!?) 11 .txa6 (11 .te3? exdS 12 .txa6 d4!) 11...bxa6 12 .te3 and now Black is better after either 12 ... exdS, 12 ... 0.xdS or 12 ... c4!?

b) 8 e6?! fxe6 9 0.gS 0.f6 10 'iWe2 'ii'd7 (Others are 1O ... eS!? or 10 ... ~h8!?) 11 .te3 (11 g3!?) 11...cS 12 0-0-0 and once again White does not have enough for the pawn, Sogaard­Kjeldsen, Odense 1993.

c) 8 'iWe2?! cS 9 dS 0.b6 10 0.e4 (No better is 10 h4 e6 11 hS exdS 12 hxg6 hxg6 13 cxdS 0.b4 14 a3 0.4xdS IS 0.e4 dxeS 16 fxeS .tg4 17 .th6 .txh6 18 ':'xh6 ~e7 19 0.egS 'iWxgS 20 0.xgS i.xe2 21 .txe2 ':'fe8 22 e6 f6+ Vaisser-Hebden, London 1991) 1O ... .tg4 11 0.egS (Or 11 0.f2 .txf3 12 gxf3 e6 13 dxe6 dxeS! 14 fxeS fS! IS h4 ':e8 16 f4 0.c7 17 hS 0.xe6 18 hxg6 hxg6 19 .i.g2 "iYd4!+ Zakhar­evich-V. Ivanov, Moscow 1991) 11...f6 (After 11...0.c7 the knight sacrifice 12 0.xh7 is not dangerous for Black: 12 ... .txf3 13 "iYxf3 ~xh7 14 i.d3 0.d7 IS h4 ':'h8 16 e6 fxe6 17 dxe6 0.f6 18 fS ~g8+ Rantanen-Sepp, Finnish Tellrn Ch 1(96) 12 h3

.txf3 13 0.xf3 e6! 14 dxe6 fxeS IS fxeS "iYe7+ was Glek-A. Kuzmin, USSR Team Ch., Podolsk 1990.

d) R.Ml1

and now: dl) 8 ... dxeS 9 dxeS?! (I like 9

fxe5!? cS 10 .tf4 cxd4 11 'iWxd4 0.b4 12 0-0-01 with an initiative for White) 9 ... 0.dcS 10 i.e3 i.fS?! (It was stronger to play 1O ... .tg4! 11 .te2 f6! with good counterplay) 11 lOd4 'iWd7, Vaisser-Avrukh, Moscow (rapidplay) 1992, and now White should have played 12 .te2! with an advantage, for example: 12 ... 0.b4?! 13 O-O! .td3 (13 .. JHd8 14 0.xfS 'ii'xfS IS 0.dS 0.xdS 16 cxdS;l;) 14 a3 .txe2 IS lOdxe2! lObd3 16 b4 0.e6 17 0.dS± or better 12 .. .f6 13 lOxfS "iYxdl +! (13 ... "iYxfS? 14 .txcS 0.xcs IS 'iWdS+ lOe6 16 g4+-) 14 ':'xdl gxfS IS e6! 0.xe6 16 .td3;l;.

d2) 8 ... c5 9 d5 White lost quickly in the

game O.Rodriguez-Magem, Li­nares Team Ch. 1991: 9 e6?! fxe6 10 h5 cxd4 11 lOe4? (It

Page 93: Beating King's Indian and Benoni

would have been better to play 11 t'bxd4) 11...t'bf6 12 t'beg5 h6 13 hxg6 hxg5 14 t'bxg5 e5 15 fxe5 'iY a5+ 0-1.

9".dxe5 10 h5! exf4 If 10 ... 'iY c7 then 11 hxg6

hxg6 (11.. .fxg6) 12 g3!?, pre­paring an attack along the h­file.

11 .1xf4 t'bf6 12 hxg6 Usually White makes this ex­

change automatically. Never­theless, 12 t'be5!? t'bxh5 (Or 12 ... t'bd7 13 t'bg4!) 13 ~xh5 gxh5 deserves further analysis. One sample variation: 14 .td3 .txe5! 15 'iYxh5! .txc3+ 16 bxc3 f5 17 'iYg5+ ~h8 18 .te5+ ~f6 19 d6! .td7 20 dxe7 'iixe7 21 'iixf6+ 'iYxf6 22 .txf6+ ~g8+.

12 ... fxg6 13 'iid2 .tfS 14 .Li3

If 14 O-O-O?! then 14 ... t'bg4!? is an unpleasant answer. Alter­natively 14 iLh6!? deserves attention but has yet to be tried in practical play.

14 ... .txd3 Vaisser-Petit, French Team

Ch. 1992, continued 14 ... t'bh5 15 .txf5 (15 .th2!?) 15 ... ~xf5 (l5 ... t'bxf4 16 .tg4) 16 .te3 b5?! 17 t'bxb5 ~b8 18 O-O! t'bg3 19 ~fel t'be4 20 'iYd3 t'bd6 21 t'bxa7! ~xb2 22 t'bc6 'iYd7 23 ~ab 1 and White obtained an advantage.

15 ~xd3 b5?! 15 ... e6!?, 15 ... t'bh5!? 16 t'bg5! 16 t'bxb5? t'bxd5!

Black Plays 6 ... t'ba6 91

16 ... 'iYb6 Not 16 ... t'bb4?! 17 'iie2

bxc4? 18 'iie6+ 'iith8 19 llxh7+! t'bxh7 20 'iYh3 winning.

The complicated position af­ter 16 ... 'iib6 arose in the game Kahn-G. Horvath, Budapest 1996. White could now have continued 17 cxb5!? t'bb4 18 'iYc4 t'bc2+ 19 'iitd2 t'bxal 20 d6+ ~h8 (Not 20 ... e6? which loses to 21 'iYxe6+ ~h8 22 t'bxh7!+-) 21 dxe7 with the better chances.

8 c5 9 exd6 exd6

Alternatively, 9 ... cxd4 10 t'bxd4 10 dxe7? gives Black a strong

initiative after 1O ... 'iYxe7 11 t'bxd4 t'bb6.

1O ... t'bb6!? After 10 ... exd6 11 t'bdb5

~h4+ 12 g3 'iYh3 13 .tfl ~e8+ 14 'iitf2 'iif5, Vokac-Spisak, Cappelle la Grande 1995, White could have taken a pawn: 15 t'bxd6!? 'iic5+ 16 'iitg2, when the threat of exchanging queens

Page 94: Beating King's Indian and Benoni

92 Black Plays 6 ... ttJa6

with ~ d5 keeps him out of trouble.

11 0-0 ~xd6 12 .te3

In this critical position Black has problems. For example:

a) 12 .. J:tdS?! 13 ttJdb5 ~xdl 14 ~axdl l::txdl 15 :Ixdl .te6 16 b3 ttJb4 17 ttJc7 :IcS IS ttJ3b5±.

b) 12 ... ~b4?! 13 a3 ~xb2 14 ttJcb5 :IdS 15 :If2 ttJc5 16 :Ibl ~a2 17 ~d3!± ~xf2+ IS .txf2 ttJxd3 19 ~xd3 ~f5 20 'iYb3 ~xbl 21 ~xbl ttJxc4 22 'iYb4± Knaak.

c) 12 ... ~c5 13 ttJe4 'iYc7 14 'iYb3! :IdS 15 :Iadl (15 ttJb5!? 'iYc6 16 ttJxa7 'iYxe4 17 .tf3 'iYxe3+ IS 'iYxe3 .td4 19 'iYxd4 :Ixd4 20 ttJxcS :IxcS 21 .txb7 is good for White) 15 ... .td7 16 c5! ttJxc5 (16 ... ttJd5?! 17 ~xd5 ~c6 IS 'iYc4 ~xe4 19 ttJe6! [Less clear is 19 ttJb5 'iY c6 20 ttJxa7 ~e6! 21 'iYxe6 fxe6 22 c6 ttJc7! 23 :IxdS+ :IxdS 24 ~b6 :Id2!] 19 ... fxe6 20 'iYxe6+ ~hS 21 'iYxe4 ttJxc5 22 'iYc4 ttJa6 23 'iYxc7 ttJxc7 24 ~f3!± Vokac­Kovalev, Ostrava 1993) 17 ttJxc5 'iYxc5 IS ttJb5 'iYfS 19

ttJxa7 ttJcS 20 ttJb5! (20 tbxcS? :IaxcS 21 ~b6 .te6 22 l';lxdS+ :IxdS 23 ~ a3=) and, thanks to the pinned bishop on d7, White is better.

d) 12 ... .te6 13 b3 :IfdS 14 ttJcbS 'iYbS (14 ... 'iYc5 15 'iYel! ~xd4? 16 'iYf2) IS 'iYel .td7 16 :Idl;;!;.

10 d5 This allows 10 ... ~xc3!? 11

.txc3 f5 with an approximately equal position (see the line 9 exd5 in Game 16).

White can try 10 0-0, waiting for 10 ... ttJf6 before playing d4-dS. If instead 10 ... :IeS then 11 fS can be considered.

The immediate 10 f5?! is premature, however, due to 1O ... cxd4 11 ttJbS d3! 12 'iYxd3 ttJdcS 13 'iYxd6 .txfS 14 0-0 :IeS IS :Iel :Ie6 with strong pressure for Black, Jagstaidt­Ad. David, Geneva 1996.

10 ttJc7 11 0-0 b5

Again 11.. . .txc3!? 12 bxc3 fS was possible.

Page 95: Beating King's Indian and Benoni

12 fS!? 12 cxb5!? was an alternative.

12 bxc4 13 fxg6

Or 13 ..tg5!? f6 14 ..tf4. 13 fxg6

This is more precise than 13 ... hxg6?!

14 ..tgS liJf6 Glek's 14 ... 'iYe8 is interest­

ing. White can then play 15 i.xc4 (15 'iYd2? liJb6! 16 ~ael 'i'd7+) 15 ... liJb6 16 .:tel 'i'd7 17 'fJ.e7 'iYg4 18· .:te4 with equality.

15 liJd2! h6? Golubev suggested that the

alternative 15 ... iLf5!? 16 liJxc4 if d7 might be an improvement. Now 17 J..f3!? is nothing spe­cial and Golubev stated that 17 lLle3 is harmless because of 17 ... lLle4! 18 liJxf5 lLlxc3(?). If we continue this variation, however, with 19 liJe7+! ~xe7 20 J..xe7 liJxdl 21 iLxf8 J..d4+ 22 'ithl 'fJ.xf8 23 'fJ.axdl we can see that White wins an ex­change without any compensa­tion. Instead of 18 ... tbxc3 Black should play 18 ... gxf5 19 tbxe4 fxe4 20 ~c2 ..td4+ 21 ~hl with an almost equal position.

16 iLh4 a6?! Here 16 ... ..tf5 17 tbxc4 ~d7

18 tbe3! liJe4 19 liJxf5 gxf5 is worse than on the previous move. White's bishop is not attacked and he can play 20 ~c2 'fJ.ae8 21 .:tael with a cer­tain advantage.

17 a4

Black Plays 6 ... tba6 93

Strategically Black's position is very difficult, so Golubev tries to generate some tactical resources.

17 gS!? 18 iLg3 tbfxdS 19 liJxdS liJxdS 20 iLxc4 'fJ.xfl +?

After 20 ... iLe6 21 ~b3 Itb8 22 Itxf8+! ~xf8 (22 ... 'itxf8? 23 ~f3+) 23 ii.xd5 'fJ.xb3 24 iLxe6+ ~h8 25 tbxb3 White's pieces are stronger than the black queen. The text move aims for an improved version of this continuation, but this fails tactically.

21 liJxfl If 21 'iYxfl then Black can

play 21...iLe6 22 'iYf3 ~f6!, indirectly protecting the knight (23 J..xd5 ~d4+). The same trick works after 22 'iYd3.

21 iLe6 22 ..txdS?!

Instead of the text move, it would have been more con­vincing to play 22 'ithl! ii.d4 23 ~f3 tbf4 24 J..xe6+ tbxe6

Page 96: Beating King's Indian and Benoni

94 Black Plays 6 ... ~a6

25 "i'd5 with a large advantage after both 25 ... "iWe8 26 l:el rJi;f7 27 iLxd6! and 25 ... r;;t>f7 26 l:tel Si.e5 27 ~d2.

Here 22 "iWb3 :b8 23 iLxd5 is evidently worse than it was in the note to Black's 20th move.

22 .id4+ 23 Si.f2

The three pieces are stronger than the queen after 23 'iVxd4 cxd4 24 iLxe6+ r;;t>g7 25 Si.d5! :b8 26 Si.f2, but the text move is more solid.

23 24 25 26

iLxd4 'i¥xd4 !1e1

Si.xd5 cxd4 .ie6 'iVe7!

Putting up the toughest re­sistance.

27 ~g3 !1f8

28 "iVe4?

A grave mistake. After 28 'i'e3! 11f6 (Or 28 .. .1~e8 29 'iVb6±) 29 ~h5 l:tg6 30 "iib6 White is clearly better.

28 'iYa7+ 29 rJi;h1 Si.b3 30 ~f5 rJi;h8!

White had considered only 30 ... 'iYf2? 31 ~e7+ rJi;h8 32 ~g6++-.

31 ~e7? Now the position becomes

equal. It was better to play 31 'iYd3. The rest of the game was played in mutual time trouble.

31 l:tf6 32 "iVd3 "iVf2 33 'iYc3 rJi;h7 34 h3 g4 35 hxg4 'iYh4+ 36 rJi;g1 "iVf2+

112-112

Page 97: Beating King's Indian and Benoni

5 The Pseudo Benko Gambit 7 ... b5

1 d4 4Jf6 2 c4 g6 3 4Jc3 iLg7 4 e4 d6 5 f4 0-0 6 4Jf3 c5 7 d5 b5

The idea of deploying the Benko Gambit structure against the Four Pawns Attack has both its pros and cons. Black's trump is that his opponent's position has been slightly exposed by e2-e4 and f2-f4. On the other hand, White has built a strong pawn centre without losing the right to castle and, additionally, has the constant threat of the break e4-e5. Black must keep a careful eye on this. Both the theory and the practice in this position promise slightly better chances for the White player.

Game 14 Nogueiras-Sax

Graz 1984

1 d4 4Jf6 2 c4 g6 3 4Jc3 iLg7 4 e4 d6 5 f4 0-0 6 4Jf3 c5 7 d5 b5

8 cxb5 The safest course of action.

After ~ Black has two choices: a) 8 .. ,tbfd7?! 9 4Jxb5! dxe5

10 iLd3 a6 11 4Jc3 ~c7 12 f5 and now: a 1) I cannot resist the temp­

tation to show one old game with a nice combination at the end:

12 ... 4Jb6? 13 g4 4J8d7 140-0 4Jf6 15 4Jg5 l:Ib8± 16 4Jge4 4Jbd7 17 "iYe2 4Jxe4 18 4Jxe4 4Jf6 19 iLe3 4Jxe4 20 iLxe4 iLd7 21 "iYf2 a5 22 .l:.ac1 .l:.bc8 23 "iYh4 "iY d6 24 1:[f3 .l:.b8 25 ~c2 .l:.b4

(see following diagram)

26 l:th3! h5 27 gxh5! gxf5 28 h6 iLf6 29 .l:.g2+ ~h8 (29 ... ~h7

Page 98: Beating King's Indian and Benoni

96 The Pseudo Benko Gambit 7 ... b5

30 'iWg3! fxe4 31 ~h5!+-) 30 'iWg3! ~g5 (Black has no good defence to White's threat 31 'iWg7+!) 31 'iWxg5 1-0 Vaisser­Loginov, Alma-Ata 1980.

a2) 12 .. .'~Jf6 13 'iWc2 gxf5 14 .txf5 e6 15 dxe6

15 ~d3!? h6 is interesting. 15 .. .fxe6 15 ... il..xe6, as recommended

by Taimanov, is hardly better: 16 .txe6 (Or 16 lL'lg5!?) 16 ... fxe6 17 0-0 lL'lc6 18 il..g5! lL'ld4 19 'iWdl ~ad8 20 ~el !;t, Wachsmuth-Jesch, COIT. 1989.

16 .td3lL'lc6 170-0 lL'lb4?! Or 17 ... lL'ld4! n. The encounter Ree-Ligterink,

Leeuwarden 1976, now contin­ued:

18 'iWe2 lL'lxd3 19 'iWxd3 il..b7 20 lL'lg5 h6 21 :xf6!

with a clear advantage to White.

b) 8 ... dxe5! 9 fxe5 lL'lg4 10 ~f4

After 10 cxb5 lL'lxe5 11 il..e2 lL'lbd7 12 0-0 a6 13 lL'lxe5 lL'lxe5 14 il..e3 axb5 15 il..xc5 b4! 16 lL'lb5, Strating-Vande Mortel, Amsterdam 1994, Black should have played 16 ... il..a6! and if 17 d6 then 17 ... 'iWa5! 18 dxe7 ~fc8 with the better chances.

10 ... lL'ld7 11 cxb5 Alternatively: 11 'iWe2 b4 12

lL'ldl 'iWc7 13 d6 exd6 14 exd6 'iWc6+; or 11 lL'lxb5 lL'ldxe5 12 lL'lxe5 lL'lxe5 13 'i'd2 e6 14 il..e2 exd5 15 cxd5 c4 16 lL'lc3 ~b8 with enough counterplay for Black as in the game Schuh­Steinbacher, German Bundes­liga 1987/88.

11. .. lL'ldxe5 12 h3 Another example went 12

lL'lxe5 lL'lxe5 13 il..e2 ii.f5 14 0-0 c4! 15 'iVd2 a6 16 a4 and White is certainly not better, Kai­danov-Efimov, Sochi 1980.

12 ... lL'lxf3+ 13 'iWxf3lL'lf6 13 ... lL'lh6!? deserves atten-

tion, e.g. 14 g4 e5 15 ii.e3 f5 with complicated play, Kaplan­Vukcevic, Hastings 1976.

14 il..c4 il..b7 15 0-0 lL'ld7 16 ~ael?! lL'lb6 17 'iVd3 ii.d4+ 18 .t.cl

(see following diagram)

Schiller-Mohr, USA 1984, and now 18 ... .i:.c8! would have

Page 99: Beating King's Indian and Benoni

The Pseudo Benko Gambit 7 ... b5 97

given Black an advantage.

8 a6

9 a4 The alternatives are: a) 9 e5?! is a little more justi­

fied than on the previous move, but it does not offer more than equality: 9 ... dxe5 10 fxe5 ctJg4 11 bxa6 (11 ~f4 ctJd7 12 e6? fxe6 13 dxe6 l::txf4 14 'i'd5 ctJe3 15 'i'xa8 ctJb6 16 'i'b8 lib4=F Kakageldiev-Acharya, Calcutta open 1994) 11.. .ctJxe5 12 ctJxe5 ~xe5 13 ~h6 lie8 14 ~b5 ~d7 15 ~xd7 'i'xd7 16 'i'd2 lixa6 17 0-0 and Black has no problems, Moskalenko-Lukin, Lvov 1984.

b) 9 bxa6?! 'i'a5!

and now: bl) 10 'i'd2?! ~xa6 11 e5

ctJfd7 12 e6 ctJb6!? 13 exf7+ lixf7 14 ctJg5 l::tf5 15 'i'e3, EI­bilia-Shirov, French Team Ch. 1993, when Black could have obtained a large advantage with 15 •.• ctJxd5! 16 ~e6+ ~h8 17 ~d2 .i.xc3! 18 bxc3ctJxf4.

b2) 10 ~d2?! ~xa6 (1O ... ctJbd7!? 11 'i'c2?! [11 ~e2 ~xa6 12 0-0] 11...~xa6 12 ctJb5 'i'b6 13 a4 c4! 14 .i.xc4 ctJg4 15 l::tfl ~fc8 16 lIc1 i.xb5 17 axb5 ctJe3 18 .i.xe3 'i'xe3+ 19 ~dl 'i'xf4 20 b3 i.h6=F Her­zog-Dzhindzichashvili, St Mar­tin 1991) 11 ~xa6 (Or 11 .i.e2!? 'i'b4!? 12 e5! dxe5 13 fxe5 ctJg4 14 i.xa6 ctJxa6 15 'i'e2 'i'xb2 16 .l::!.bl 'i'c2 17 O-O!, Kozul-Kochiev, Palma de Mallorca 1989, with an unclear position) 11...'i'xa6 12 'i'e2 'i'xe2+ 13 ~xe2 ctJfd7 with a good version of the Benko Gambit for Black. For example: 14 lihc1?! (14 ctJel!? ctJa6 15 ctJc2 lifb8 16 lihb 1 !? with the idea of ctJc2-e3-c4, Bangiev)

Page 100: Beating King's Indian and Benoni

98 The Pseudo Benko Gambit 7 ... b5

14 ... liJa6 IS liJdl fS! 16 liJgS fxe4 17 liJe6 llfb8+ Bangiev­Bologan, German Bundesliga 1993.

b3) 10 liJd2 iLxa6 11 i.xa6 liJxa6! 12 0-0 liJc7 13 ~e2 1:.fb8 14liJc4 ~a6 IS i.d2liJd7 16 b3 liJb6 17 l1.ac1 liJxc4 18 bxc4 ki.b4 with sufficient coun­terplay, Pr. Nikolic-Hoi, Esb­jerg 1982.

c) 9 ~b3!? (Threatening bS­b6) 9 ... ~b6 (9 ... axbS!? 10 .txbS i.a6; 9 ... e6!? 10 b6 ~e7) 10 a4! axbS 11 i.xbS, Piket­Shirov, Aruba (match) 1995, and now instead of 11...i.a6? 12 liJd2 .txbS 13 liJxbS liJa6, permitting 14 liJc4 ~b7 IS ~f3 liJb4 16 ~e2!± (Korchnoi), Black should have played 11...liJa6! 12 liJd2 liJb4 with compensation for the pawn.

d) After 9 i.d2 Black may transpose to the main line after 9 ... 'ii'b6 10 a4 axbS 11 i.xbS (11 axbS?! llxal 12 ~xal e6 13 dxe6 i.xe6 14 liJgS?! i.b3! is slightly better for Black, Arencibia-Tal, Manila Inter­zonal 1990).

Now we return to the main line 9 a4.

(see following diagram)

9 axbS Other possibilities for Black

in this position, which is critical for the 7 ... bS line:

a) 9 ... .tb7 10 i.e2! White is slightly better after

10 bxa6 liJxa6 11 Ji.c4 liJb4 12 0-0 liJd7 13 i.d2!? liJb6 14 b3 i.a6 IS .txa6 ki.xa6 16 ~e2, Vaisser-Brito, Las Palmas 1993.

10 ... axbS 11 i.xbS e6 12 0-0 exdS 13 exdSliJbd7 14 fS!

A thematic move. 14 ... liJb6 IS i.gS ~c7 16

fxg6 Also playable is 16 iLxf6

i.xf6 17 'i'd2! ± (But not 17 liJe4? .txb2! 18 :!bl .txdS!=t=).

16 ... hxg6 Not 16 .. .fxg6? 17 as±. 17 i.xf6 i.xf6 18 'J'd2+

Here are two examples from this position:

al) 18 ... Ji.a6 19 i.xa6 ki.xa6

Page 101: Beating King's Indian and Benoni

The Pseudo Benko Gambit 7. .. b5 99

20 iYf4 iYd8 21 as CDd7 22 CDe4 ~eS 23 iYh6 iLg7 24 iYh3 and White developed a decisive at­tack on the kingside, Piskov­Friegrich, Berlin 1990.

a2) 18 ... c4 19 'ii.thl! 'iYd8 20 as CDd7 21 CDe4 and White is winning, Kishnev-Vukic, Pula 1988.

b) 9 ... 'it'aS 10 iLd2 This is much more accurate

than 10 CDd2?! 'it'b4!? 11 Ita3 tbg4! 12 'it'f3 fS! with the ini­tiative.

10 ... 'it'b4 After 1O ... e6?! the simplest

continuation is 11 dxe6! (Not 11 eS?! dxeS 12 CDe4 'it'd8 13 4Jxf6+ iLxf6 14 dxeS iLxeS!? IS CDxeS 'it'h4+ 16 g3 'it'e4+ 17 'We2 'it'xhl) 11...iLxe6 12 iLe2 axbS 13 iLxbS;i;.

After 1O ... 'i'b4 White has three playable moves:

b 1) 11 'i'c2 c4 The most critical move. After

11. .. axbS (l1...iLd7!? 12 eS ~fS) 12 iLxbS iLa6? (The black hishop must control the fS­square. 12 ... iLd7!;i; was neces­.'iary, e.g. 13 eS? iLfS! 14 ~c1

CDe4) 13 eS! CDg4 (l3 ... dxeS changes nothing: 14 fxeS CDfd7 IS CDe4+-) 14 CDa2 iLxbS IS iLxb4 and White went on to win, Barsov-Dostan, Budapest open 1989.

l2....e5 Black has no real problems

after 12 as axbS 13 CDa4 'i'b3 14 'it'xb3 cxb3 IS tbb6 ~a7 16 eS dxeS 17 CDxeS, Lanchava­Van der Weide, Leeuwarden 1995, and now 17 ... i.a6!? is best.

12 ... iLf5! But not 12 ... dxeS? 13 fxeS

CDg4 (l3 ... iLfS?! 14 tba2!+-) 14 tbdl! 'iWcs IS 'i'xc4±.

13 tba2 'iics 14 'i'xc4 CDxdS

Compared with the variation covered in the previous com­ment, Black's queen is pro­tected and this move becomes possible. However, White re­tains slightly better chances af­ter

IS 'i'xcs dxcS 16 lIc1 axbS! This is much better than

16 ... CDd7?! 17 bxa6 Itfb8 18 iLc4 tbc7 19 b4! tbxa6 20 .tbS CDc7 21 iLc6 1.1a6 22 i.xd7

Page 102: Beating King's Indian and Benoni

100 The Pseudo Benko Gambit 7. .. b5

~xd7 23 a5± Blokh-Holthuis, COIT. 1994.

17.txbS+ Not 17 ~xc5?!lha4. b2) 11 'iWbll? is an interest­

ing try: b21) After 11...c4?! 12 as!

'iWcs (12 ... axbS?! 13 lLla4 c3 [Black cannot play 13 ... 'iWb3 as in the line with 11 ~c2 because of 14 :a3] 14 lLlxc3±) 13 b6 lLlg4 14 'iWcll ~xc3 (14 ... fS IS h3 'iWf2+ 16 ~d1 ~xc3 17 'iWxc3 fxe4 18 lLld4 e3 19 .tel e2+ 20 lbxe2+-) 15 'iWxc3 'iif2+ (1s ... lLlf2? 16 b4!+-) 16 ~d1 i.b7 (16 ... 'iic5 17 ~xc4 lbf2+ 18 ~e2 lLlxe4 19 'iWd4 'iixd4 20 lLlxd4±) 17 ':'a4! lbd7 18 h3 lbcs 19 ':'xc4 lLlf6 20 ~e 1 and White went on to win, Malinin­Sirota, COIT. 1992.

b22) Probably Black should have played 1l ... axb5!? 12 J..xb5 J..a6. At least 13 eS (which is winning with the queen on c2), doesn't work here: 13 ... dxeS 14 fxeS lLlg4 IS lLle4 'iib3, attacking the dS­pawn.

b3) 11 ~d3 c4 12 ~c2'iWcS 12 ... lLlfd7 13 :a2 is not very

promising for Black, but after the untried 12 .... f1xb2?! it is not easy for White to prove an ad­vantage. 13 ':'bl ~a3 14 b6! might well be the best way for White to continue, for example 14 ... ~b7 IS 0-0 'iicS+ 16 ~hl e6?! 17 dxe6 fxe6 18 lLlgS! 'iWc8 19 fS±.

11 ~e2 axb5

and now: b31) 14 lLlxbS? lLlxe4! IS

~xe4 (1S 'iWxe4 ~fS 16 'iWxe7 ~xc2+) lS ... 'iWxbS! 16 axbS ':'xal+ 17 ~f2 lixhl and Black is better, Balogh-Lechtynsky, Budapest 1986.

b32) 14 e5?! worked out well in the game Glek-Yanvarjov, USSR 1989, with Black's help: 14 ... dxeS? IS fxeS lLlfd7 16 i.e3 'iib4 17 ~d4 bxa4 18 0-0 lLlcS? 19 e6! .txd4+ 20 lLlxd4 f6 21 i.xg6!+-, but the move 14 ... lLlg4! would have been very annoying.

b33) 14 .te3! 'iWb4 IS 0-0 bxa4

IS ... 'iWxb2?! is refuted by 16 lLlxbS (Not 16 .td4? lLlxdS!+; 16 lia2? 'iWxc3 17 .td2 lbxd5! winning) 16 ... lbxe4! 17 lia3! ~fS (17 ... lLlc3? 18 ':'xc3! ~xc3 19 ':'b1+-) 18 lLlfd4 lLlg3 19 hxg3 ~xd4 20 lLlxd4! 'iWxa3 21 .txfS±.

16 e5 lLlg4 Avshalumov-Smirin, USSR

1986 saw White develop a very strong attack after 16 ... lLlhS 17 lLlg5! dxeS (Or 17 ... 'iWxb2?! 18 lbxa4 ':'xa4 19 ':'xa4 .tfS 20 ':'xc4 with a clear edge for White, Ehlvest-Kochiev, Lenin­grad 1984) 18 fxe5 ~xe5 19 lLlxh7! 'iWxb2 (19 ... .txh2+? 20 ~xh2 'iWd6+ 21 ~gl lLlg3 22 'iWf2+-) 20 lLlxf8 'iWxc3 21 i.xg6! fxg6 22 lLlxg6 i.f6 23 lLlxe7+ i.xe7 24 'iWxhS (24 ~f2!?) 24 ... 'iWxe3+ 2S ~hl.

17 i.d4

Page 103: Beating King's Indian and Benoni

The Pseudo Benko Gambit 7 ... b5 101

All White's pieces are well placed and he is clearly better. The game Glek-Sorin, Odessa 1989, continued 17 ... CLlh6 18 CLlxa4 iLf5 19 CLlb6 (19 e6!?; 19 iLc3!? 'i'b7 20 ii.xf5 CLlxf5 21 g4 CLlh6 22 CLld4, Glek) 19 ... ~a6 20 CLlxc4 iYb5! 21 ~ac I! dxe5 22 j,xe5! ii.xc2 (22 ... j,xe5 23 iYxe5 'i'xc4 24 ii.xf5) 23 ~xc2 iYxd5 24 ii.xg7 r/{;xg7 25 iYxe7±.

c)Le6

Here the possibility of trans­posing to the Benko Gambit with 10 b6 is not very convinc­ing: 1O ... iYxb6 11 a5 'i'b7 12 ii.c4 exd5 13 CLlxd5 iLe6 14 0-0 CLlxd5 15 exd5 j,d7, Banikas­Herraiz, Cala Galdana 1996.

Instead White has two prom­ising possibilities: 10 dxe6!? and 10 iLe2!?

c1) 10 dxe6!? j,xe6 1O .. .fxe6?! is met by 11 e5

dxe5 (l1...CLle8? 12 CLle4±) 12 CLlxe5;t.

11 j,e2 Not 11 e5?! dxe5 12 'ii'xd8

~xd8 13 fxe5 CLlg4 14 iLf4 CLld7! 15 j,e2 axb5 16 CLlxb5 CLlgxe5 17 CLlxe5 j,xe5! with an equal position, as in Karasev­Yuferov, USSR 1977.

11...axb5 12 iLxb5 CLla6 If instead 12 ... d5 (After

12 .. :i'c7 13 0-0 CLlc6, Vaisser­Gufeld, Sochi 1981, 14 f5! would have seized the initia­tive) 13 exd5 CLlxd5 14 CLlxd5 iLxd5 15 0-0 CLlc6 16 j,e3! iLxb2 (l6 ... j,xf3?! 17 ~xf3 CLld4 18 'i'f2±) 17 ~b 1 (The spectacular 17 iLxc5 iLxa 1 18 'ii'xal!? ~e8 19 ~dl 'i'a5! 20 ~xd5 'i'xb5 21 'i'h8+! r/{;xh8 22 axb5 is unclear) 17 ... iLd4 18 CLlxd4 cxd4 19 iLf2 'iif6 20 'i' d2 ~fd8 21 ~bc 1 White is slightly better. The game San Segundo-Magun, Linares 1986, continued 21...'i'e6?! 22 f5! 'i'd6?! 23 iLg3 'i'b4 24 'YWg5 and even after the accurate 24 ... iYe7! 25 'i'xe7 CLlxe7 24 ii.h4 ~a7 25 ~fdl iLe4 26 iLxe7! ~xe7 27 f6 ~e6 28 ~xd4! White has a large ad­vantage.

llQ.=O and now: c11) 13 ... CLlb4?! is less cir-

Page 104: Beating King's Indian and Benoni

102 The Pseudo Benko Gambit 7 ... b5

cumspect than 13 ... liJc7. After the typical blow 14 fS! gxfS IS exfS i..xfS 16 liJh4 both 16 ... i..c2 17 'iYf3 liJe4 18 liJfS liJxc3 19 bxc3 i..xfS 20 'iYxfS i..xc3 21 lIa3 i..d4+ 22 i..e3, Bleis-Zuse, Mannheim 1987, and 16 ... i..g4 17 i..e2 liJfd5 18 liJxdS i..xe2 19 'iYxe2 'iYxh4 20 liJe7+ ~h8 21 liJfS 'iYf6 22 lIa31, Lalic-Zakic, Yugoslavia 1986, give White a strong at­tack.

el2) 13 ... 'iYb6 14 fS! gxfS IS liJgS!? (IS exf5 also deserves attention: IS ... i..xfS 16 liJh4 c4+ 17 ~hl i..d3 18 lIf4 dS 19 liJf5, lashvili-Gavrilov, Mos­cow 1990) IS ... liJc7 16 exfS i..c8 17 i..f4 liJxbS 18 liJxbS h6 19 liJh3 lId8 20 lIa3 with the better chances for White, Sem­kov-Peev, Plovdiv 1988.

el3) 13 ... liJc7 14 iLd3 'iYb8 I prefer White's position after

14 ... dS?! IS eS liJd7 16 liJgS!? 'iYe717f51

ll[S1 with a better posItIon for

White. In the game Ivanov­Kupr(>irl,il, Tpnin,!p(l lQ~N

White missed this move and Black achieved full compensa­tion for the pawn after 15 e5?! dxeS 16 liJxeS liJfdS 18 'iYf3 liJb4 19 i..e4 lIa6 20 'iYe2 ~e8 21 i..e3 f6 22 liJc4 f5.

c2) 10 i..e2!? An idea originated by Uhl­

mann and developed by Piskov. 10 ... axbS 10 ... exdS 11 eS normally just

represents a transposition of moves after 11...dxeS 12 fxe5 liJg4 13 'iYxdS 'iYxdS 14 liJxdS axbS 15 i.xbS. Sooner or later Black must play ... a6xbS in this line.

11 i.xb5 exd5 A strange miniature occurred

in Michaelsen-Binzenhoefer, Berlin 1993: 11...i..a6 12 i..xa6 liJxa6 13 dxe6 fxe6 14 0-0 dS IS eS liJe8 16 liJgS~ 'iYe7? 17 liJxdS 1-0. ~ This is the point! 12 ... dxeS Or 12 ... liJe8?! 13 liJxdS (13

'iYxdS liJc7 14 'iYxd6? 'iYxd6 IS exd6 i.xc3+ 16 bxc3 liJxbS-+; 13 i..xe8 lIxe8 14 'iYxdS~) 13 ... i..b7 14 i..c4 liJc6 IS i..e31 dxe5 16 i..xcs liJd6 17 liJe7+! liJxe7 18 'iYxd6 i.xf3 19 gxf3 liJfS 20 'iYxd8 lIfxd8 21 b31 and White is clearly better, Piskov­Savon, Norilsk 1987.

13 fxeS liJg4 Not 13 ... liJe4?1 (13 ... liJe8? 14

'iYxdS±) 14 'iYxdS liJxc3 (If 14 ... 'iYxdS IS liJxdS the black knhrht is harllv placed on e4) IS

Page 105: Beating King's Indian and Benoni

The Pseudo Benko Gambit 7 ... b5 103

'ilHxd8 ~xd8 16 bxc3 ii.a6 17 ii.gS! ~c8 (17 ... i.xeS 18 tLlxeS ~dS 19 i.f6 i.xbS 20 c4 ~d6 21 cxbS ~xf6 22 tLlf3±, Glek and Piskov. 22 ... ~f4 doesn't help: 23 a-a! ~fxa4? 24 ~xa4 ~xa4 2S ~d 1! +-) 18 0-0 tLlc6 19 ii.f6!± Piskov-Vasyukov, Moscow 1987.

14'ilHxdS 14 .tgS!? is complicated but

seems to favour White, e.g. 14 ... 'ilHaS IS 'ilHxdS i.e6 16 'ilHe4 tLld7 17 i.f4 (17 a-a? tLldxeS 18 tLlxeS i.xeS ! +) 17 ... tLlb6 (l7 ... 'ilHb4!?) 18 0-0 'ilHb4 19 'ilHxb4 (19 h3!? 'ilHxb2 20 ~fc1) 19 ... cxb4 20 tLle4 tLldS 21 i.g3 tLlde3 22 ~fc1 b3!, El­bilia-Kaabi, Maghreb Ch. 1994.

14 ... 'ilHxdS 14 ... 'ilHaS?! is dubious. After

IS h3! .te6 16'ilHe4 tLlh6 17 g4 White is clearly better.

IS tLlxdS

White has a small but secure advantage after both:

c2l) IS ... tLlxeS 16 tLlxeS i.xeS 17 i.h6 i.g7 (17 ... ~d8 18 a-a-a!) 18 tLle7+ (18 i.xg7!? rJ:;xf!.7 19 0-0-0) 18 ... rJ:;h8 19

.txg7+ rJ:;xg7 20 tLlxc8 ~xc8 21 rJ:;d2 tLlc6 22 rJ:;c3 tLld4 23 ~hel; and

c22) IS ... .td7 16 tLlb6 ~a7 17 tLlxd7 tLlxd7 18 .i.f4 tLlgxeS 19 tLlxeS tLlxeS 20 0-0-0 c4 21 .txeS .i.xeS 22 ~dS, Destre­becq-Cortes, French Team Ch. 1991. .

c23) Probably the best de­fence for Black was found in the very first game with this variation: IS ... .tb7!? 16 tLlb6 lta7 17 tLlc4, Uhlmann-Votru­ba, Czechoslovakia 1978, and now Black should have played 17 ... tLlc6 18 0-0 tLlcxeS 19 tLlfxeS tLlxeS 20 tLlxeS .i.xeS 21 .te3i.

10 i.xb5 i.a6 Black cannot be satisfied

with his position after 1O ... tLla6?! 11 0-0 tLlb4 12 eS! tLle8 13 'ilHe2 (Or 13 .1c4 im­mediately) 13 ... tLlc7 14 i.c4 .i.b7 IS ~dl l:ta7 16 i.e3 'i'a8 17 'i'd2 ~d8 18 fS!± Banikas­Kamer, Cala Galdana 1996.

11 ~d2!

Page 106: Beating King's Indian and Benoni

104 The Pseudo Benko Gambit Z .. b5

This is more precise than 11 lib 1 ?! iL.xb5 12 axb5 lbbd7 13 0-0 lbb6 14 'iVe2 'tIYc8 15 c;f;thl e6 16 dxe6 'tIYxe6 with enough counterplay for the pawn, Vais­ser-S. Garcia, Bayamo 1985.

The evaluation of 11 'tIYe2!? depends on the unclear position arising after 11...lbxe4! 12 'tIYxe4 i.xc3+ 13 bxc3 iL.xb5 14 iL.b2! In the game 1. Horvath­Kiss, Zalakaros 1988, White succeeded in developing a strong attack after 14 ... iL.a6?! (Better is 14 ... iL.d7! 15 c4 'tIYb6) 15 f5 lbd7 16 fxg6 hxg6 17 c4 'tIY a5+ 18 'it>f2 liab8 19 lihb 1 life8 20 'tIYh4 f6 21 'tIYh6 lbf8 22 lbh4 i.xc4 23 lbxg6 lbh7 24 lbxe7+! lixe7 25 iL.xf6! and went on to win.

11 iL.xbS Instead: a) The line which Hoi played

against Peicheva (Copenhagen 1989) has not found a follow­ing: 11...e6?! 12 dxe6 fxe6 13 0-0 d5 14 e5 lbe4 15 iL.e3 lbxc3 16 bxc3 lbd7 with an advantage for White. The game continued 17 lbg5 'tIYe7 18 'tIY g4 lif5 19 'tIYh3 h6 20 lbf3 lbf8 21 'tIYg3 lif7 22 ~fbl 'tIYc7 23 'tIYf2 lbd7 24 iL.xd7! 'tIYxd7 25 'tIYc2 and White won a second pawn.

b) After 11... 'tIYb6 12 0-0 lbe8, Galioto-Nepeina, Cap­pelle la Grande, White has the interesting try 13 a5!? 'tIYb7 14 iL.xa6 followed by 15lba4;!;.

12 axbS lixa1 L\ft('~ 1'1 0hrf7 n 0-0 llx!'ll

14 'tIYxal we transpose to the next comment. It is less precise to play 13 ':'xa8, e.g. 13 ... 'tIYxa8 14 0-0 lbb6 15 'tIYb3?! (15 'iVe2!? lbe8 16 g4!?t) 15 ... lbe8 16 iL.e3lbc7 17lbel? f5!+ Lut­skan-Krakops, Latvian Ch. 1993.

13 'tIYxa1 'tIYb6 14 0-0

14 lbe8?! My first experience with the

logical move 14 ... lbbd7!? was not successful:

a) 15 'iVa6?! (the queen does nothing on a6) 15 ... lib8 16 lial lbe8 17 'tIYxb6 lbxb6 18 lia7 'it>f8 19 i.e 1 lbc8 20 lid7 lbb6 l/z-lh Vaisser-Lanka, Cappelle la Grande 1994.

b) White did better in the following game: 15 'tIYel! (Pre­paring 16 'iVe2 or 16 'tIYh4) 15 ... e6 (l5 ... 'tIYb7!?) 16 dxe6 (Or 16 'tIYh4!? exd5 17 e5) 16 .. .fxe6 17 e5 lbh5 18 exd6 'iVxd6 19 g3 with an advantage to White, Vaisser-Nataf, French Team C:h. 1996. The game con-

Page 107: Beating King's Indian and Benoni

The Pseudo Benko Gambit 7 ... b5 105

tinued 19 ... e5 20 fxe5 CDxe5 21 CDxe5 .i.xe5 22 l::i.xf8+ ~xf8 23 .i.h6+! ~e7?!, and now the simplest way to play was 24 .i.f4! (24 ... CDxf4? 25 gxf4 'i'd4+ 26 ~g2+-).

15 'i'e1! CDc7 16 'i'h4!

Another natural plan for White was shown in the game Lazarev-Yuferov, Naberezhnie Chelni 1988: 16 'i'e2!? CDd7 (Or 16 ... 'i'b7!?) 17 e5 'i'b7 18 c6 fxe6 19 dxe6 CDb6 20 CDg5 with the initiative.

16 CDxb5 Black's position is difficult.

The choice is not easy: 16 ... l::i.e8 17 e5! and 16 ... e6 17 f5! exf5

18 exf5 ttJd7 19 ttJg5 both give White a strong attack.

17 ttJxb5 'i'xb5 18 'i'xe7

18 c4? The decisive mistake. It was

necessary to play 18 ... .i.xb2! 19 'i'xd6 (19 e5? is weak. In the game Michaelsen-Kunsztowicz, Hamburg 1987, Black contin­ued 19 ... dxe5! 20 fxe5 l::i.e8 21 'i'h4 l::i.xe5!=t) 19 ... c4 and the passed c-pawn offers Black some counterplay.

19 .i.c3 20 'i'xd6 21 tiJe5 22 l::i.a1 23 'i'c6

l::i.e8 l::i.xe4 'i'e8 tiJd7 1-0

Page 108: Beating King's Indian and Benoni

6 Other Systems for Black

In this chapter we consider other possibilities in the Four Pawns Attack from the King's Indian move order. The material is presented in two games. In Game 15, Vaisser-Krasenkov, we study sidelines without ... c7-c5, while in Game 16, Vaisser­Kr. Georgiev, we look at side­lines with ... c7-c5.

Game 15 Vaisser-Krasenkov

Paris 1990

1 d4 lbf6 2 c4 g6 3 tbc3 J..g7 4 e4 d6

Overly-brave experiments in this opening can usually be punished, e.g. 4 ... 0-05 f4 c6?! 6 e5! tbe8 7 h4! with a strong attack.

5 f4

(see following diagram)

5 0-0 The immediate 5 ... J..g4 is

interestinp-. Now the Mtempt to

deliver a direct refutation is dangerous for White:

a) 6 'i'b3?! ctJc6! 7 J..e3 e5 8 dxe5 dxe5 9 'i'xb7 J..d7 with a strong initiative.

b) If White tries to transpose to the game continuation 5 ... 0-0 6 ctJf3 J..g4 7 ii.e2 by playing 6 lbf3, Black can try 6 ... ctJc6 7 d5 (7 ii.e3?! e5! 8 fxe5 dxe5 9 d5 ctJd4 10 J..xd4 exd4 11 'i'xd4 0-0 gives a lot of play for a pawn) 7 ... J..xf3 8 gxf3 ctJb8 with complicated play.

c) 6 ii.e2!? looks best, e.g. 6 ... ii.xe2 7 'i'xe2 (7 ctJgxe2 is also possible) and now:

c 1) 7 ... 0-0 8 tbf3 c5 9 dxc5!

Page 109: Beating King's Indian and Benoni

(An unclear position arises after 9 d5 e6 10 e5!? exd5! 11 exf6 lIe8 12 fxg7 1:lxe2+ 13 lLlxe2) 9 ... dxc5 (9 ... 'i'a5? is incorrect: 10 cxd6 lLlxe4 11 'ilixe4 .Jtxc3+ 12 Wfl! .Jtf6 13 dxe 7 I:re8 14 'i'xb7 lLla6 15 .Jtd2 'i'b6 16 'i'xb6 axb6 with a big advan­tage for White) 10 e5 lLlfd7 11 0-0 lLlc6 12 lIdl with pressure for White.

c2) 7 . ..liJc6 8 lLlf3 e5 (8 ... lLld7?! 9 e5!) 9 dxe5 dxe5 10 fxe5 lLlg4 11 .Jtg5 'i'c8 (ll...'i'b8!?) 12 lLld5 0-0 (The capture 12 ... ctJcxe5? does not work due to 13 h3, when now both 13 ... lLlxf3+ 14 'i'xf3 lLle5 15 'i'g3 and 13 ... h6 14 .Jtf4 g5 15 .Jtg3 ctJxf3+ 16 'i'xf3 ctJe5 17 'iWc3 are bad for Black) 13 0-0 I:re8 14 .Jtf6 ctJcxe5!? 15 lLle7+ lixe7 16 .Jtxe7 'iie6 and Black has insufficient compen­sation for the exchange.

6 ctJf3 a) If White wants to get the

most flexible version of the line 6 lLlf3 c5 7 d5 e6 8 dxe6 fxe6, he can play 6 ..td3: 6 ... c5 7 d5 c6 8 dxe6 (8 lLlge2 exd5 is not attractive for White; after 9 cxd5 'i'b6 or 9 exd5 lLlh5 Black has a good position) 8 .. .fxe6 (8 ... .Jtxe6?! 9 f5! is in White's favour) 9 lLlge2. Yuneev-Smir­in, Daugavpils 1989, continued 9 ... lLlc6 10 0-0 a6 11 a4 b6 12 4:Jg3 ctJd7 13 'i' g4 lLld4 14 ..te3 "fie7 15 I:rael i.b7 16 h4 Wh8 I 7 h5 g5! with complicated play.

Other Systems for Black 107

After 6 iLd3 Black can also play 6 .. .':t:Ja6 7 tlJf3 transposing to the line 6 tlJf3 lLla6 7 ..td3 or try the interesting 6 ... tlJc6!? 7 tlJf3 ..tg4 8 e5! (8 .Jte3 e5!) 8 ... dxe5!? 9 dxe5 lLlb4 10 ..tbl 'i'xdl + 11 tlJxdl i.xf3 12 gxf3 ctJh5 with the idea of ... f7-f5 and ... ..th6.

b) After 6 ..te2 the simplest for Black, depending on his taste, is to transpose to the line 6 lLlf3 tlJa6 7 iLe2 with 6 ... tlJa6 7 ctJf3 or to the line 6 lLlf3 c5 7 d5 b5 with 6 ... c5 7 d5 b5. In the latter line the position of the bishop on e2 instead of the knight on f3 usually favours Black. For example, after 8 cxb5 a6 9 a4?! axb5 10 ..txb5 ..ta6 Black wins an important tempo compared to the 'normal' variation. In the case of 9 bxa6 'iWa5 10 ..td2 ..txa6 11 lLlf3 ..txe2 12 'i'xe2 'i'a6! Black gets a Benko-like position with full compensation for the pawn, Welling-Hoeksema, Eindhoven 1988.

6 ..tg4 Let us consider other moves

in this position apart from 6 ... c5 and 6 ... ctJa6, which are studied elsewhere in this book:

a) 6 ... e5?! 7 dxe5!? (White can also play 7 fxe5 dxe5 8 d5 tlJbd7 9 ..td3 with a position similar to that in the line 6 ... lLla6 7 i.d3 e5) 7 ... dxe5 8 'iWxd8 lIxd8 9 lLlxe5! (Less at­tractive is 9 fxe5 lLlfd7! [After 9 ... lLlg4? 10 ..tg5 I:re8 11 lLld5

Page 110: Beating King's Indian and Benoni

108 Other Systems for Black

l2Ja6 12 c5! is bad for Black] 10 l2JdS l2Ja6 11 ..tgS 1:.1e8 120-0-0 [ 12 l2Jf6+ ? ! .txf6 13 exf6 l2JacS+] 12 ... l2JxeS 13 cSt? [13 l2Jf6+?! ..txf6 14 ..txf6 l2Jg4+] 13 ... ..tg4 with mutual chances) 9 l2JxeS and now none of Black's three possibilities are enough to give Black equality:

a 1) 9 ... l2Jxe4 10 l2Jxe4 f6 11 l2Jxf6+ ..txf6 12 .td2 l2Jc6 13 ii.c3 ..tfS 14 ~f2 l2Jd4 IS 1:.1dl l2Je6, was the game Lingnau­Finke, RLNS 1989, and now White could have kept an ad­vantage after 16 1:.1xd8+ 1:.1xd8 17 ~e31:.1dl 18 g3.

a2) 9 ... 1:.1e8 10 ..td3 l2Jfd7! (1O ... l2Jxe4? loses to 11 i.xe4 f6 12 .tdS+ ~h8 [12 ... ~f8 13 b3 l2Jd7 14 ..ta3+ cS IS 0-0+-] 13 .tf7 1:.1f8 14 l2Jxg6+) 11 l2Jxd7 i.xc3+ 12 bxc3 l2Jxd7 13 i.e3 l2Jf6 14 eS l2Jg4 IS ..td2 f6 16 h3 l2Jh6 17 0-0 fxeS, Stoy­Kuntzig, RLNS 1988, and after 18 fxeS! l2Jf7 19 ii.f4 White could have achieved a small plus (l9 ... l2JxeS? 20 1:.1ael 1:.1d8 211:.1xeS1:.1xd3221:.1e8+±).

a3) 9 ... l2Ja6 10 ii.e3 reaches a position similar to that after 6 ... l2Ja6 7 i.e2 es 8 dxeS dxes 9 l2JxeS, except that instead of ii.e2 White has played ii.e3. But now the cS-square is controlled and this difference changes the evaluation of the position in

. White's favour. Two examples: a31) 1O ... ..te6 11 ..te2 l2Jb4

12 1:.1c1 l2Jd7, Metge-Rogers, Auckland 1992. and now the

simple 13 l2Jxd7 ~xd7 14 ~f2 would have assured White a plus. For example: 14 ... lbd3+?! IS ..txd3 lixd3 16 l2JdS! c6 17 ~e2± or 14 ... liad8 IS ~hdl! 1:.1xdl 16 lixdl 1:.1xdl 17 lbxdl l2Jxa2 18 ii.xa7±.

a32) 1O ... b6 11 ii.e2 ii.b7 12 ii.f3 l2Jb4 13 ~e2l2Jc2 14 I!adl l2Jxe3 IS ~xd8+ lixd8 16 ~xe3 lie8 171:.1dl ..tf8 18l2JdS± Gab­riel-Steinbacher, Dresden open 1994.

b) After 6 ... l2Jbd7?! White has two ways to assume the ad­vantage:

bl) 7 ..te2 eS 8 dxeS dxeS 9 fxeS l2Jg4 10 .tg5 'iWe8 11 l2JdS l2JdxeS 12 'iWd2!? (12 l2Jxc7? l2Jxf3+ 13 gxf3 'iWe5 14 l2Jxa8 'iWxgS IS fxg4 'iWh4+ 16 ~f1 lid8 17 'iWb3 ii.xg4 18 ii.xg4 'iWxg4 gives Black a strong at­tack). The game Konikowski­Rechel, Bundesliga 1989/90, continued 12 .. .f6? (Better was 12 ... 'iWd7) and now White could have obtained a clear edge with 13 l2Jxc7! 'iWe7 (or 13 ... 'iWc6 14 l2Jxa8 fxg5 IS 'iWdS+±) 14 l2Jxa8 fxgS IS lid 1 ..te6 16 'i'd6.

b2) 7 eSt? l2Je8 8 cS! (It is necessary to stop 8 ... cS) 8 ... c6 9 ..te3 b6 (No better is 9 ... dxcS 10 dxcS ~aS 11 a3) 10 cxd6 exd6 11 ..tc4 bS 12 ..tb3 l2Jb6 13 0-0 with advantage to White, Skem­bris-Kalesis, Greece 1994.

c) 6 ... a6. The main idea of this move is to prepare ... b7-bS, as for example in the case of:

Page 111: Beating King's Indian and Benoni

c1) 7 ~e2 cS 8 dS bS! c2) After 7 ~d3 Black plays

7 ... ~g4 (Not 7 ... cS? 8 dxcS! dxcS 9 eS and the weakness of the b6-square guarantees an advantage for White. If 7 ... c6?! 8 eS!? l2Jfd7 9 ~e2 bS 10 cxbS axbS 11 h4 b4 12 l2Je4 .i.a6 13 hS .i.xd3 14 ~xd3 with a dan­gerous attack for White, o. Rodriguez-Gallego, Spanish Ch. 1993) 8 ~e3 l2Jfd7 attacking White's centre. Still the compli­cated position arising after the sequence 9 h3 (Also possible is 9 .i.e2!? .i.h6 10 ~d2) 9 ... .i.xf3 10 iVxf3 l2Jc6 11 'iVf2 eS 12 dxeS dxeS 13 fS l2Jd4 14 0-0-0 (14 1:i.dl!?) 14 ... bS IS g4 c6 16 gS f6 17 h4 must be better for White, Avshalumov-Loginov, Budapest 1990.

c3) A strange but perfectly playable move is 7 a3!?, pre­venting 7 ... cS? because of 8 dxcS! dxcS (8 ... ~aS? 9 b4) 9 ~xd8 1:i.xd8 10 eS±. 7 ... l2Jbd7?! is also no good: 8 eS! l2Je8 9 cS c6 10 .i.e3 as in variation b2 above. Black should play 7 ... .i.g4 reaching a position very similar to that in the main game.

d) 6 ... l2Jfd7 7 .i.e3 eS 8 fxeS dxeS 9 dS as 10 .i.e2 l2Ja6. Now instead of the modest continua­tion 11 0-0 l2JacS 12 b3 h6 13 .. 3 'iVe7 14 1:i.bl fS IS exfS gxfS 16 b4 axb4 17 axb4 l2Ja6 18 cS with complicated play, although White's chances might still be hetter, Vaisser-Hausrath, Baden Baden (rapidplay) 1995, I

Other Systems for Black 109

should have continued with the precise 11 a3!? 'iVe7 121:i.bl.

7 .i.e2 Alternatively: a) 7 .i.e3!? is not bad either:

7 ... l2Jfd7 8 h3!? ~xf3 9 'iVxf3 eS (If 9 ... l2Jc6?! 10 eS! l2Jb4 11 O-O-O!±) 10 dxeS (10 fxeS? cS!) 10 ... dxeS 11 fS l2Jc6 (Or per­haps 11.. . .i.h6!? 12 .i.f2 [12 O-O-O!?] 12 ... l2Jc6 13 1:i.dl;t) 12 0-0-0 l2Jd4 13 'iVf2 c6 14 g4 iV as IS gS 1:i.fd8 16 h4 l2Jcs 17 ~b I! with advantage to White, Glek-Damljanovic, Belgrade 1988.

b) 7 h3?! is premature, how­ever: 7 ... .i.xf3 8 'iVxf3 l2Jc6 9 .i.e3 eS! 10 dxeS dxeS 11 fS l2JhS! (An important difference compared to the previous note -a black knight comes to f4) 12 g4 l2Jf4 13 1:i.dl 'iVe7 14 ~f2 1:i.fd8 IS l2JdS l2JxdS 16 exdS l2Jd4 and Black seized the ini­tiative, Ruban-Krasenkov, Pod­olsk 1990.

7 8 .i.e3

l2Jfd7 l2Jc6?!

Page 112: Beating King's Indian and Benoni

110 Other Systems for Black

In the game Moskalenko­Ennenkov, Wijk aan Zee II 1992, Black played 8 ... e5?! 9 fxe5 dxe5 10 d5 .txf3 11 .i.xf3 .i.h6 12 ~f2! 'Wie7 13 0-0 a5 14 a3 lLla6 15 l:.bl b6 16 b4! axb4 17 axb4 with a clear advantage to White. 8 ... .i.h6 deserves at­tention, preparing 9 ... e5.

9 dS Also interesting is 9 e5!? e6

10 0-0 lLle7 11 h3 ~xf3 12 ~xf3 ith6 13 'Wiel! lLlf5 14 l:.dl c6?! (Better is 14 ... lLlxe3!? 15 'Wixe3 l:.b8) 15 itc1 'Wib6 16 'Wif2 dxe5 17 dxe5 l:.fd8 18 lLle4 'Wixf2+ 19 'iitxf2, Petronic-Mar­kovic, Nis 1994, and White's space advantage was trans­fonned into something more tangible.

9 .txf3 10 itxf3 lLlaS 11 'Wid3

11 .i.e2 with the idea of 11...c6 12 itd4 is not bad either.

11 e6! 12 l:.c1!

It is less precise to play 12

O-O?! cxd5 13 exd5 lLlc5!, while Black has good compensation for the piece after 12 b4? cxd5! 13 bxa5 'Wixa5 14 l:.c1 dxc4.

12 exdS 13 exd5 l:.e8 14 b3 lLlc5 15 'Wid 1

Now the queenside is well protected and White is better thanks to the two bishops and the bad position of Black's knight on a5.

15 e6 More precise was 15 ... e5.

16 0-0 exd5 17 .i.xd5 lLlc6

At last Black's knight re­enters the game, but now the d6-pawn cannot be defended.

18 lLlb5 'iith8 19 lLlxd6 was a threat.

19 ~f3 lLlb4 20 lLlxd6 lLled3

White's position is winning, so Black tries to create compli­cations.

21 lLlxe8 'Wixe8 22 'Wid2!

Page 113: Beating King's Indian and Benoni

The simplest solution. 22 lbxc1 23 'ii'xb4 lbxa2 24 'ii'd2! 'ii' e6

24 ... ttJc3 25 ~d4 does not save the knight.

25 jLd5 26 SLd4 27 'ii'xd4+ 28 .l::i.al 29 'ii'f2 30 h3

'iWa6 SLxd4+ ~g8 .l::i.e8 'ii'a3

The knight is lost. Black re­signed some moves later.

Game 16 Vaisser-Kr. Georgiev French Team Ch. 1996

1 d4 lbf6 2 c4 g6 3 lbc3 jLg7 4 c4 d6 5 f4 0-0 6lbf3 c5

7 d5 White can also play: a) 7 dxc5 'ii'a5 7 ... dxc5 8 'iWxd8 .l::i.xd8 9 e5

favours White. After 7 ... 'ii'a5 White can choose between:

a 1) If White is facing a stronger opponent and is happy to make a draw, he can try 8 .~d2!? 'ii'xc5 9 b4 and now Black can choose between a draw by repetition after:

all) 9 ... 'ii'xb4 10 ttJa4 (The active 10 lbd5?! is dubious be­cause of 1O ... 'ii'a3 11 SLb4 'ii'a6 12 c5 lbxd5! 13 SLxa6 ttJxb4 14 c5 lb8a6+) 1O ... 'ii'a3 11 SLel ~b4+ 12 SLd2; or the continua­tion

a12) 9 ... 'ii'b6 10 SLd3 (The

Other Systems for Black 111

complications after 10 e5?! were in Black's favour in Ham­douchi-Ehlvest, Lucem 1989: 1O ... dxe5 11 fxe5 lbg4 12 ike2 'ii' e6 13 ttJdS ttJxe5 14 0-0-0 ttJbc6+) 1O ... SLg4 11 ~b 1 lbc6 12 h3 .txf3 13 'ii'xf3 e5 14 SLe3 ikd8 15 f5 a5 16 b5 lbb4 17 SLgS h6 with equality, Dorf­man-Sznapik, Warsaw 1983. a2)~

and now: a21) 8 ... lbfd7!? is Van der

Wiel's favourite move, but it now has the reputation of being very dubious. The game Gre­tarsson-Van der Wiel, Leeu­warden open 1995, is widely regarded as being the final refutation: 9 cxd6! ~xc3+ 10 bxc3 'i!fxc3+ 11 'i!fd2 'ii'xal 12 dxe7 lIe8 13 e5 ttJc6 14 0-0 lbd4? 15 lbg5! (An improve­ment on the original game Ljubojevic-Van der Wiel, Wijk aan Zee 1986, where after IS .tb2? ttJxf3+ 16 gxf3 'i!fxa2 17 f5 Black could have stopped White's attack by playing 17 .. Jhe7! 18 f6 .l::i.xe5 19 'ii'h6 .l::i.gS! 20 '\t>h1 ttJxf6-+) Is ... lbc5

Page 114: Beating King's Indian and Benoni

112 Other Systems for Black

16 i.a3 ttJcb3 17 'i'f2! 'i'c3 18 'i'h4 h5 19 ttJe4 'i'xd3 20 ttJf6+ 'it>h8 21 ttJxe8 winning. A little­known correspondence game Szczepaniec-Stypka, corr. 1989, could change this conclusion. Instead of 14 ... ttJd4? Black tried 14 ... ttJc5! and won quickly after 15 .tb2? 'i'xa2 16 f5 i.xf5! 17 i.xf5 ttJe6 18 i.xe6 fxe6 19 'i'h6 ':xe7. The critical posi­tion, and one that needs practi­cal tests, arises after 15 .ta3! ttJb3! 16 ':xal ttJxd2 17 ttJxd2 ttJxe7 18 ttJe4 ':d8 19 ttJd6. White has good compensation for the exchange.

The courageous sacrifice 9 cxd6! is considered to be best, as the alternatives do not com­pare favourably, for example: 9 i.d2?1 (9 0-0?1 i.xc31? 10 bxc3 'i'xc3) 9 ... ttJxc5 10 .tc2 'i'b4 (10 ... ttJc6 11 a3?! 'i'a61 12 ttJb5 .tg4 13 ':bl ttJd4 14 ttJbxd4 i.xd4, Gorbatov-Bologan, Nov­gorod open 1995, is also in Black's favour) 11 ttJd5 (11 i.b3?1 ~b6 12 'i'e2 ttJxb3 13 ttJd5? 'i'a6-+ Johannessen-Tal, Reykjavik 1964) 11...'i'xb2 12 ':bl 'i'xa2 13 ttJc7 i.g4 14 ttJxa8 ttJba6+.

a22) 8 ... 'i'xc5 9 'i'e2 ttJc6 10 i.e3 'i'a5

An alternative is 1O ... 'i'h5 11 h3 ttJg4 (11...ttJe8 12 ':c1 e5 13 ttJd5 exf4 14 i.xf4 ttJe5 15 O-Oj; Cifuentes-Herraiz, Benasque 1996) 12 i.d2 ttJd4 (l2 ... ttJb4?! 13 ttJd51 ttJxd5 14 cxd5 i.xb2 15 ':bl i.a3 16 'i'fl ttJf6 17

g4± Szily-Gereben, Hungary 1953) 13 'i'fl ttJxf3+ 14 ~xf3 .td4 15 'it>e2j;.

110-0.tg4 A very interesting alternative

is 11...ttJg4!? 12 i.d2 ~b6+ (12 ... ttJd4 13 ttJxd4 .txd4+ 14 'it>hl 'i'h5 15 h3 ttJf6 16 'i'f3 'i'xf3 17 ':xf3j; H. Gretarsson­A. Gretarsson, Icelandic Team Ch. 1995) 13 'it>hl 'i'xb2 14 ttJb51 (Stronger than 14 ':bl) 14 ... i.d7 15 e5! dxe5 16 ':fbl e4 17 ':xb2 exd3 18 'i'fl i.xb2 19 ':e1 ':ad8 20 h3 ttJh6 21 'i'xd3 and the complications were resolved with a slight edge for White, Leitao-Shaked, Cala Galdana 1996.

12 ':ac1 ttJd7 13 'i'f2 i.xf3 13 ... ttJb4 ? 1 instead led to a

strong White initiative after 14 i.bl lhc8 15 a3 ttJa6 16 ttJd2 .te6 17 f5! in the game Gabriel­Uhlmann, German Bundesliga 1995.

14 gxf3 ttJc5 15 i.b 1 Now White's natural plan is

to attack on the kingside using the half-open g-file.

15 ... ttJa4!

Page 115: Beating King's Indian and Benoni

15 ... ~b4?! is very risky: 16 .I1fdl ~xc4 17 tbd5 ~a4 18 e5! b6 19 b4 tbxb4 20 tbxe7+ ~h8 21 exd6± Cifuentes-Van W ely, Amsterdam 1995.

16 tbdl !? White has nothing special in

the position after after 16 tbxa4 ~xa4 17 .I1fdl (Or 17 b3 ~a3 18 c5!? dxc5 19 i£,xc5 ~xc5! 20 .I1xc5 i£,d4= Topalov-Kaspa­rov, Linares 1994) 17 ... I:rac8 18 b3 ~a5 19 .I1d5 ~c7 20 .I1cdl b6 21 a3 .I1fd8 22 h4 e6 23 .I1g5 ~e7, Topalov-Dolmatov, Bur­gas 1995.

16 ... ~h5?! 16 ... tbc5!? is possible, ready

to repeat moves after 17 tbc3 tba4.

17 b3 tbc5 18 tbc3 tbe6 19 tiJd5 f5 20 exf5 gxf5 21 b4 I:rf7 22 ~hl b6 23 I:rgl ~h8 24 J:.g3

White has a dangerous initia­tive, Leitao-Arduman, Yerevan Olympiad 1996.

b) It is worth saying a few words about a line that has al­most fallen into disuse nowa­days: 7 i£,e2 cxd4 8 tbxd4 tbc6 9 ~e3 (If 9 ttJc2 ttJd7 10 0-0 tbc5 11 i£,f3 f5 with counter­play). Now the simplest way to equalise may be 9 ... e5!? 10 tbxc6 bxc6 11 fxe5 dxe5 120-0 (Or 12 ~c5 .I1e8 13 ~xd8 .I1xd8 14 0-0 .I1d2 15 .I1ad 1 .I1xd 1 16 ~xdl i£,e6 17 ~d6 tbg4= Uhl­mann-Fisher, Varna 1962) 12 ... ~c7 13 ~el, Benko-Gligo­ric, Los Angeles 1963, and now 13 ... i£,e6 14 ~h4 ttJd7=.

Other Systemsfor Black 113

7 e6 7 ... b5 is seen in Game 14. Right now 7 ... a6 can be met

not only by the natural 8 a4, but also by the energetic 8 i£,d3!? b5 9 e5! dxe5 (After 9 ... ttJe8!? 10 0-0 ttJc7 11 h3!? looks good, preventing 11.. .i£,g4; while in­stead 9 ... tbfd7!? 10 e6!? fxe6 11 ttJg5 ttJf6 is unclear) 10 fxe5 ttJg4 11 ~e2 ttJd7 12 ~f4 ~c7?! (Better is 12 ... b4!? first) 13 0-0 tbgxe5 (l3 ... tbdxe5 14 ~hl) 14 ~ael tbxf3+ 15 .I1xf3 ~b6 16 ~xe7;j;. Now the con­tinuation 16 ... bxc4?! 17 i£,xc4 ~xb2 18 .I1bl ~c2 19 i£,b3 'ilif5 20 i£,d6 ~g4 21 ttJe4 gives White a strong attack, Korch­noi-Gheorghiu, Vienna 1986.

8 i£,e2 8 i£,d3 exd5 9 cxd5 is not

dangerous for Black. The sim-plest is 9 ... ~g4 (9 ... b5; 9 ... c4; 9 ... .I1e8; 9 ... iVb6) 10 0-0 a6 11 a4 ttJbd7 transposing to an im­proved version of Chapter 2. If now Black can play ... c5-c4, ... ttJc5 and ... tbfd7 he will have

Page 116: Beating King's Indian and Benoni

114 Other Systems for Black

a good position, while if White prevents this plan by playing 12 ~c2 then Black gets active play with 12 ... ~xf3 13 :'xf3 :'c8 14 b3 tDg4!?

After 8 dxe6 fxe6 (8 ... ..txe6 is met by 9 ..td3, intending f4-f5) White can develop his bish­op on either e2 or d3, but in both cases Black has some pro­mising possibilities. The sim­plest way to equality is the uni­versal manoeuvre ... e6-e5:

a) 9 ..td3 e5!? 10 fxe5 (10 f5? is bad: 1O ... gxf5 11 exf5 dS! 12 ..tc2 [12 cxd5 e4!=t], Yuneev-I. Zaitsev, Chigorin Memorial 1994, and now Black should have played 12 ... d4 13 tDe4 ~xf5=t; while 10 O-O!? exf4 11 ~xf4 tDg4!? 12 ..tg5 ~f6 is acceptable for Black) 10 ... dxe5 11 0-0 tDc6. The position is equal, but look how White was outplayed in the game Dok­hoian-Smirin, Sverdlovsk 1987: 12 ~g5 h6 13 ~h4 ~d6 14 tDdS g5 IS ~el ..tg4 16 ~c3 tDd4 17 ~xd4 tDxd5! 18 exd5 exd4 19 ~e4 bS! 20 ~c2?! bxc4 21 ~xc4 ~c8! 22 :'fc1 d3! 23 ~xd3 ~xb2=t.

b) 9 ~e2 tDc6 100-0 e5!? 11 fxe5 dxe5 12 ~xd8 :'xd8 13 ~g5 Zlf8=.

8 :'e8 Black also has a) 8 ... bS when 9 cxbS a6 (or

9 ... exdS 10 eS) 10 a4 exd5 11 eS transposes into one of lines of the 7 ... bS system which fa­vours White.

b) 8 ... tDa6 can be met by 9 e5 dxe5 10 fxe5 tDg4 11 ..tgS.

c) 8 ... exdS and now: c1) 9 cxd5 corresponds to the

main lines considered earlier in the book (Chapters 1-3). c2)~ This variation, attributed to

Gunderam, is currently in crisis because of:

9 ... tDe4! 10 tDxd5 No better is 10 cxd5 tDxc3 11

bxc3 tDd7 120-0 (Or 12 e6 fxe6 13 dxe6 tDb6 14 0-0 ..txe6 15 tDg5, I. Ivanov-Hernandez, St John 1988, and now Black should have played 15 ... i.d5!=t and if 16 :'bl?! ~xc3 17 ~e3 ~xa2-+) 12 ... dxe5 13 ~b3 e4 14 tDg5 tDb6 15 :'dl ~d7 16 a4 ~e8 17 ~b5 h6=t Arencibia­Martin del Campo, Santa Clara 1990.

1O ... tDc6! 11 ~c2 Or 11 ~d3 f5 12 exf6 tDxf6

13 0-0 tDxd5 14 cxdS tDd4 15 tDg5 ~f5! 16 Whl ~d7 17 ~e3 ~xd3 18 ~xd3 ~f5 19 &tadl, as in Kantorik-Balogh, Slova­kian Team Ch. 1995, and now

Page 117: Beating King's Indian and Benoni

19 ... lbc2! 20 ii.c1 (Not 20 lbe6? 'fixd3 21 ~xd3 c4!-+) 20 ... 'ilfxd3 21 llxd3 lbb4 would have maintained Black's ad­vantage.

11...fS 12 0-0 dxeS 13 fxe5. lbxeS 14 ii.f4 lbc6

14 ... 11e8!? also deserves at­tention, e.g. IS lladl b6 with advantage to Black.

IS l'ladl gS 16 J.c7 'ilfe8 17 ii.d3 h6 18 ,l:tde1 'ilfhS 19 .txe4 fxe4 20 'ilfxe4 ius 21 'ilfe2 g4 22lbd2

This position arose in the game Reidel-Heinatz, Kecske­met 1989. Black could have won a pawn by 22 ... llae8 23 ~dl ii.xb2.

It was in this variation that I played a memorable game ag­ainst Garry Kasparov in 1981. After 9 eS?! lbg4?! 10 cxdS dxeS 11 h3 e4 12 hxg4 exf3 13 gxf3 (This was a novelty at the time) 13 ... ~e8 14 fS 'i'Hb6? (Six years later Kalinin discovered an improvement: 14 ... gxfS! IS J.h6 J.xc3+! 16 bxc3 fxg4, but White is still okay after 17 'ilfd2 J.fS 18 0-0-0 'fib6 19 J.d3 .txd3 20 'iVxd3 lbd7 [+ Kaplun] 21 'ii/S! lbf8! [21...lbeS? 22 d6! lle6 23 J.gS+-] 22 'ilfxg4+ lbg6 23 'fihS) IS J.h6! 'ilfxb2 16 .txg7 1;xg7 17 f6+? (White could have secured a large ad­vantage after 17 ~c I! gxfS 18 llc2) 17 ... 1;g8! 18 'fic1! 'ilfb4!! 19 1;f1? (19 'ilfd2!=) 19 ... lbd7 20 J.bS 'iVd4! Black was clearly better because the planned 21

Other Systems for Black 115

'ilfh6? is refuted by 21...lbxf6! 22 .txe8 'ilfxc3 23 1;g2 .txg4! 24 ii.xf7+ 1;xf7 2S fxg4 ~e8-+. Some moves later Kasparov won the game. c3)~

This line is much quieter than 9 cxdS and not very popular nowadays. Black has a choice of solid continuations which give him comfortable equality.

9 ... l'le8 Black can try an interesting

blockade idea 'it la Nimzow­itsch': 9 ... lbhS!? 10 0-0 (10 .td2?! ..td4! 11 lbxd4 cxd4 12 lbbS a6! 13 lbxd4 'ilfh4++ as in the game Pribyl-Vukic, Kapfen­berg 1970) 10 ... J.xc3 11 bxc3 fS! (A necessary move. If 11...lbg7?! then 12 fS! ..txfS 13 J.f4 with a strong initiative for the pawn) 12 lbgS lbg7 13 ..tf3 lbd7 14 J.d2 lbf6 IS ..tel J.d7 16 .th4 'i'Hc7 with equality, Danner-Vukic, Austrian Team Ch. 1991.

10 0-0 .tfS 11 J.d3 A complicated and approxi­

mately equal position arises after 11 lbh4 !? lbe4! 12 lbxfS

Page 118: Beating King's Indian and Benoni

116 Other Systems for Black

gxfS 13 liJxe4 fxe4 14 iLe3 ~xb2 IS ~bl 'ilVf6 16 'ilVb3 ~d4 17 iLxd4 'ilVxd4+ 18 ~hl b6, Antoshin-Boleslavsky, Lenin­grad 19S6, 19 'ilVa4!? ~f8!

11. " 'iIV d7 12 liJh4 White's king is in constant

danger after 12 h3 liJa6 13 a3 (Or 13 g4?! ..txd3 14 'ilVxd3 liJb4 IS 'ilVdl hS! 16 gS liJe4 with an edge to Black) 13 ... liJc7 14 g4 (The game Peng-J. Pol­gar, Novi Sad Womens' Olym­piad 1990, saw 14 'ilVc2 bS! IS cxbS [IS b3!?] IS ... liJfxdS 16 liJxdS liJxdS 17 ~xfS gxfS 18 ~bl ~e4! with better chances for Black) 14 ... ~xg4!? IS hxg4 'ilVxg4+ 16 ~h2 'ilVhS+ 17 ~g2 ~g4+ 18 ~h2 bS! 19 ~gl ~hS+ 20 ~g3? (20 ~g2) 20 ... bxc4 21 ..txc4 ~e7 22 'ilVd3 ~h6! 23 ~g2 (23 iLd2 i.xf4!+ 24 ~xf4 ~h3-+) 23 ... ~ae8. The game Conquest-Mestel, Hast­ings 1986, finished with a pleas­ant combination: 24 ii.d2? (24 liJgS!?) 24 ... ii.xf4! 2S ..txf4 "i¥g4+ 26 ~g3 l:te3 27 ~f1 liJhS 28 ~h2 l;'Ixf3 29 ~h3 "i¥xc40-1.

12 ... ii.xd3 The unclear positIon which

arises after 12 ... liJe4 13 liJxfS gxfS 14 ..txe4 fxe4 IS fS i.d4+ 16 ~hl e3 17 ~g4+ ~h8 18 liJe2 ii.f6 19 ~f3, Glek-Smirin, Minsk 1986, would not be to everybody's taste.

13 'ilVxd3 liJg4 14 liJf3 liJa6 IS a3 liJh6 16 ..td2 liJc7 17 :E!:Jel 'iYf5 lR v.wxf5 4lxf5

with equality, Danner-Uhl­mann, Graz 1991. If now 19 b4, Black has the powerful 19 ... bS! in reserve.

9 e5!?

White can also play 9 dxe6. With the black rook on e8 this line is more attractive for White than the immediate 8 dxe6. Even Gunderam's variation 9 0-0 dxeS 10 eS becomes play­able.

9 10 fxe5 11 ~g5

dxe5 liJg4

11 ~f4 resembles the line 8 ... exdS 9 cxdS ~e8 10 eS dxeS 11 fxeS liJg4 12 i.f4 liJxeS 13 0-0 (Game 4).

11 'ilVa5 More active is 11...'ilVb6 12

'ilVd2 (12 d6?! liJxeS 13 liJbS ii.d7 14 liJc7 liJbc6 IS 0-0 liJd4 16liJxa8 ~xa8 17 ~e7liJec6 18 liJgS fS and Black has more than enough play for the ex­change, Bagaturov-Avshalu­mov, Yurmala 1982) 12 ... liJxeS 13 0-0 (13 liJxeS! ? ii.xeS 14

Page 119: Beating King's Indian and Benoni

0-0) 13 . ..tZJbd7 14 liadl (14 ~hl !?) 14 ... ttJxf3+ 15 .txf3, Padevsky-Bilek, Bad Pyrmont 1970. Now instead of 15 ... ttJf6? 16 d6! ttJd7 17 ttJb5 with ad­vantage to White, Black should have played 15 ... ttJe5! with an unclear position.

12 0-0 ttJd7?! It was better to play 12 ... exd5

13 cxd5 ttJxe5 14ttJxe5 lixe5. 13 ttJb5

Exploiting the offside posi­tion of the black queen, White's knight heads for d6. However, the more precise route was via e4, in order to keep open the option of playing ttJf6+.

13 ttJdxe5 14 ttJxe5 ttJxe5 15 ttJd6 !IfS 16 iLe7 .td7

17 ttJe4 17 ttJxb7 would have been

very unclear after 17 ... 'iYb6 18 iLxf8 (18 'iYb3 lifc8; 18 dxe6? iLxe6 19 iLxf8 iLxf8 20 'iYb3 'iYc7!=t) 18 ... .txf8! 19 'tWb3 exd5 (19 ... lib8? 20 ttJa5!) 20

Other Systems/or Black 117

cxd5 c4+ 21 'iYxb6 axb6 22 lif6.

17 f5! 18 it.xfS lixfS

We often meet this type of position in the Four Pawns At­tack. The strong bishops and extra pawn provide almost full compensation for the sacrificed exchange.

19 dxe6 i.xe6 20 ttJg5 it.d7 21 i.f3 ttJxf3+ 22 'iYxf3 'iYd2

Interesting was 22 ... i..c6!? 23 'i'h3 h6 24 ttJe6 !Ie8 25 ttJxg7 ~xg7 (Or 25 ... lie2 26 ttJxf5 lixg2+ 27 'iYxg2 i.xg2 28 ttJe7+ ~g7 29 ~xg2 'iYd2+ 30 ~hl±) 26 ~d3. White is better, but the strong bishop on c6 makes victory highly problem­atic.

23 24 25 26 27

'iYxb7! 'i'd5+ cxd5 liabl ~hl

it.c8 'i'xd5 i.xb2 iLd4+ it.e3!

It is necessary to diminish the

Page 120: Beating King's Indian and Benoni

11 8 Other Systems for Black

activity of White's rooks. White wins after 27 ... .i.a6? 28 ~fe1 ~d8 29 ttJe6 (Or 29 ~e7!?) 29 ... ~xd5 30 ttJc7 ~d6 31 ~e7 ~f8 32 ~xh7 ~g8 33 ~e7 ~f8 34 lIe6.

28 ttJe6 Or 28 ttJf3!? .i.a6 29 lIfe 1 f4

30 ttJe5. 28 29 dxe6 30 ~b7

.i.xe6 e4 lIe8?

The decisive mistake. Black should have played 30 ... lIe8! 31 lId1 c3 32 ~c7 .i.d2 33 ~b1 ~f8 34 ~g1 .i.e3+! (34 .. Jhe6?

35 lIb8+ ~e8 36 ~xe8+ ~xe8 37 ~f2±) 35 ~f1 i.b6! with real chances of a draw.

31 lId1! e3 32 ~dd7 i.d4

32 ... c2? loses immediately to 33 lIg7+ ~h8 34 I:txh7+ ~g8 35 lIbg7+ ~f8 36 e7+ ~e8 37 ~g8+ .

33 34 35

~de7 g3 ~d7 1-0

~e8 h5 iH6

Here Black's flag fell, but after 36 ~xa7 the result would have been clear in any case.

Page 121: Beating King's Indian and Benoni

7 Taimanov System

After the introductory moves of the Modem Benoni, 1 d4 tDf6 2 c4 cS 3 dS e6 4 tDc3 exdS 5 cxdS d6 6 e4 g6 7 f4 ~g7 we reach the diagrammed position. Here 8 tDf3 leads to the Four Pawns Attack systems examined in the earlier chap­ters. White can also opt for two alternative moves: 8 .1i.b5+ (the Taimanov System) and 8 e5 (the Mikenas Attack). The first of these lines is dealt with in this chapter and is particularly dangerous for Black, while the second one (covered in Chapter 8) is almost forgotten and could provide an unpleasant surprise

for your opponent. First we shall consider Tai­

manov's 8 .1i.b5+.

8 .tbS+ introduces the ag­gressive Taimanov system. This bishop check aims to disturb the smooth development of Black's pieces and is rather unpleasant for Black. In fact, some of the leading Benoni players nowa­days play 3 ... c5 only after 1 d4 tDf6 2 c4 e6 3 tDf3 in order to avoid this variation. The critical move to challenge White's ap­proach is 8 ... tDbd7. The evalua­tion of this extremely sharp line has oscillated for many years between \ unclear , and 'White is

Page 122: Beating King's Indian and Benoni

120 Taimanov System

better'. After the impressive game Sokolov-Topalov, Wijk aan Zee 1996, (Game 17) how­ever, ,Black has hardly dared repeat this line, even though everything is not entirely clear. In the notes to this game you will also find 8 ... ~d7 and other deviations from the main move order. The alternative 8 .. .'~Jfd7 has been considered slightly better for White for a long time now and this is examined in Game 18.

Game 17 Sokolov-Topalov Wijk aan Zee 1996

1 d4 ltJf6 2 c4 e6 3 ltJc3 cS 4 dS exdS 5 cxdS d6

One way of avoiding the Four Pawns Attack is S .. :~e7. Un­fortunately this is not the best place for the queen in several other systems of the Benoni.

6 e4 g6 7 f4 iLg7

Other possibilities are not very attractive for Black:

a) 7 .. :~e7?! does nothing to combat White's build up: 8 ltJf3! iLg4 (8 ... ltJxe4? 9 'iVa4+) 9 h3 iLxf3 10 'iVxf3 iLg7 11 iLd3 0-0 12 0-0 ltJbd7 13 iLd2 is clearly better for White (ECD).

b) 7 ... a6?! 8 eS and now: b1) 8 ... ltJfd7 9 ltJf3 ~g7 10

ltJe4 dxeS 11 ltJd6+. We now have one of the principal posi­tions of the Mikenas Attack

with an extra tempo for White (because of 7 ... a6?!), and this changes the evaluation in White's favour. Look at the mmlature Guseinov-Zaitchik, Volgodonsk 1983: 11...~f8 (Or 11... ~e7 12 ltJxc8+ irxc8 13 iLe2) 12 ~e2!? irc7 (12 ... exf4 13 iLxf4 iff6 14 ~g3 ifxb2 IS ~b1±, Kapengut) 13 ltJxc8 irxc8 14 fxeS ltJxeS IS ltJxeS ~xeS 16 iLh6+ iLg7 17 'ii'd2 iffS 18 ifc3 iLxh6 19 ifxh8+ ~e7 20 d6+ ~xd6 21 ~d1+ ~c6 22 ~f1 1-0.

b2) The game Nutu-S.Lupu, Romania 1993, saw 8 ... ife7? 9 ltJf3 .i.g7 (9 ... iLg4 10 iLe2 ltJfd7 11 exd6 'ii'xd6 12 ltJgS! ~xe2 13 ifxe2+ ife7 14 ltJce4±) and now White could have obtained a big advantage after 10 h3 ltJhS 11 ltJe4.

c) 7 ... iLg4?! 8 'ii'a4+! .i.d7 9 ifb3 if c7 10 ltJf3 iLg7 11 eS±.

d)7 ... ltJbd7?! 8 ltJf3 ~g7 9 eS dxeS 10 fxeS ltJg4 11 e6 fxe6 12 dxe6 ife7 13 ltJdS! 'ii'xe6+ 14 ~e2 ifd6 IS iLgS iLd4 16 ~c4±.

8 iLbS+ ltJbd7?! The alternative 8 ... iLd7? has proved unsuccessful. Af-

ter ~ Black has: a) Not many players are

ready to suffer after 9 ... dxeS 10 fxeS ire7 11 ltJf3 (11 ife2!?) 11...0-0 12 0-0 ltJg4 13 iLf4 (13 iLgS!?) 13 ... ltJxeS 14 .i.xeS!

Page 123: Beating King's Indian and Benoni

~xe5 15 lIel f6 16 ~c4! ri;g7, Lau-Perenyi, Budapest 1981, and now the simplest was 17 tDxe5 fxe5 18 'ii'e2 with an evi­dent advantage.

b) 9 ... tDh5 offers more resis­tance. 10 tDf3 and then:

bI) After 1O ... dxe5 White was clearly better in the game O'Kelly-Diaz del Corral, Ma­drid 1957: 11 fxe5 0-0 12 ..ixd7! and now:

b11) After 12 ... tDxd7 Black loses a knight: 13 g4 tDxe5 14 gxh5 tDxf3+ 15 'ii'xf3 lIe8+ 16 ri;d1± Vaisman-Stein, Odessa 1958.

In 1993 the English interna­tional master Andrew Martin tried to revive this variation. He proposed here 16 ... lIe5 17 h6! ..ih8 (17 ... lIf5 18 'ii'xf5! gxf5 19 lIgl±) hoping that the threat of 18 ... lIf5 or quiet counter­attack ... 'ii'd7, ... b5-b4 would give Black good counterplay. White has nothing to be afraid of; two pawns instead of one for the bishop do not make Black's position any more enviable. The simplest solution is 18 ~f4! !If5 19 'ii'e4 'i'd7 (Other moves are no better: 19 .. :ilih4 20 :IeI!; 19 ... b5 20 d6! 'ii'e8 21 lie 1 'ii'xe4 22 :Ixe4 g5 23 i.g3 lifl + 24 lIe2! !Ixa1 25 d7+-; 19 ... i.xc3 20 bxc3 'i'xd5+ [20 ... lixd5+ 21 ri;c2+-] 21 'ii'xd5 !Ixd5+ 22 ri;c2 f6 23 h4+-) 20 lie1 i.xc3 (20 ... b5 21 d6! !Id8 22 ri;c2 ~xc3 23 bxc3 b4 24 a4! looks winning for

Taimanov System 121

White) 21 bxc3 !Ixd5+ 22 ri;c2 !Id8 23 !Ie2 and, since 23 ... .i:.d4 does not work: 24 cxd4 'ii'a4+ 25 ri;b2 'i'b5+ 26 ri;c1 'i'c4+ 27 'ii'c2 'ii'xd4 28 i.e5+-, White has a practically winning posi­tion.

b12) 12 ... 'i'xd7 13 0-0 'i'f5 14 'i'a4 i.xe5 15 tDxe5 'ii'xe5 16 i.h6lid8 17 liae1 'i'd4+ 18 'ii'xd4 cxd4 19 tDb5 tDa6 and now 20 d6! would have secured White's advantage.

b2) 10 ... 0-0 11 i.xd7 'i'xd7 12 0-0 tDa6 13 !Ie 1 lIae8 14 a3 b6 15 b3 tDc7 16 !Ia2 lId8 17 g3, when Black can hardly cre­ate sufficient counterplay with the knight out of business on h5, but after 17 ... dxe5?! 18 fxe5 h6 19 d6 tDe6 20 tDd5 White is much better, Hertneck-Vlaho­poulos, Katerini 1993.

9 e5! Without this key move the

check 8 i.b5+ would represent a loss of time, as Black has de­veloped his queen's knight to the natural d7 -square.

9 dxe5 The immediate 9 ... tDh5? is

bad because of 10 e6 iYh4+ (l0 ... fxe6 11 dxe6 'ii'h4+ 12 g3 tDxg3 13 lLlf3 iYh3 14 'iixd6!+­as in the game Schakis-Farron, Crei11975) and now both 11 g3 tDxg3 12 tDf3 'iih3 13 !IgI! i.xc3+ (13 ... tDf5 14 exd7+ i.xd7 15 'ii'e2+ ~d8 16 i.xd7 ri;xd7 17 'ii g2) 14 bxc3 tDe4 15 'Vii d3! and 11 ~fl i.d4 12 'ii e I! are winning for White.

Page 124: Beating King's Indian and Benoni

122 Taimanov System

10 fxe5 ctJh5 Of course not 1O ... 1i'e7? 11

1i'e2+-.

11 e6 White also has the quieter

move 11 ctJf3!? at his disposal. Practical play has seen: 11.. .0-0 12 iLg5 (12 g4? ctJxe5 13 gxh5 iLg4 14 J.e2 ctJxf3+ 15 iLxf3 ~e8+ 16 ~f1 iLh3+ 17 i.g2 1i'f6+ 18 ~gl1i'h4-+) and:

a) 12 ... f6 13 iLh4 ctJxe5 14 ctJxe5 "Jiie7 15 0-0 "Jiixe5 16 lIel 1i'd4+ 17 i.f21i'xdl 18 :'axdl;J;; Gustavsson-Kountz, Dortmund 1989.

b) 12 ... 'tWa5 13 0-0 a6?! (Interesting is 13 ... ctJxe5!?) 14 i.xd7 iLxd7 15 d6± Hauschild­Voigt, Dortmund 1992.

c) 12 ... 1i'b6 13 1i'e2 a6 (After 13 ... ctJxe5?! 14 ctJxe5 iLxe5 as in Grooten-Hendriks, Enschede 1992, White should have played simply 15 1i'xe5 f6 16 iLxf6 ctJxf6 [16 .. .lhf6 17 O-O-O±] 17 0-0 i.f5 18 d6 :'ad8 19 g4±) 14 iLa4 (Not so good is 14 iLc4?! 4Jxe5 15 ctJxe5 l'1e8 16 0-0

:'xe5 17 1i'd2 1i'b4 ! ,+, as in Cherepkov-Katishonok, Lenin­grad 1990, but quite playable is 14 iLxd7!? i.xd7 15 d6 c4 [15 .. .f6!? 16 ctJd5 "Jiib5] 16 'iYd2 ii..c6 17 0-0-0 i.xf3 18 gxf3 iLxe5 19 d7 with a strong ini­tiative, Wells-Hodgson, British Ch., Southport 1983) 14 ... 'iYb4 IS O-O-O! bS 16 iLc2 ctJb6 17 a3 1i'a5 18 ctJd2±, as in Petursson­Muller, San Bernardino 1989.

11 1i'h4+ Weaker is 11. "fxe6? 12 dxe6 0-0 Or 12 ... "Jiih4+? 13 g3 iLxc3+

14 bxc3 Vi'e4+ IS 'i'e2 'tWxhl 16 exd7+ ~d8 17 i.gS+ ~c7 18 0-0-0 winning.

13 ctJf3!

and now: a) 13 ... ctJdf6 14 1i'xd8 ':xd8

15 e7 l'1d6 16 ctJg5! i.e6 170-0 a6 18 ctJxe6 axbS 19 ctJc7 l'1c8 20 ctJ3xbS is winning for White (Nei).

b) 13 ... l'1xf3 14 Vi'xf3 and now:

bl) 14 ... iLxc3+ 15 bxc3 ctJe5 16 1i'e4 1i'f6 17 e7! 1i'xe7 18 0-0 ii..fS 19 1i'dS+ ~Q.7 20

Page 125: Beating King's Indian and Benoni

~gS! 'iVxgS 21 'iVxeS+± Kap­engut.

b2) 14 ... lbeS?! IS 'iVdS 'iVh4+ 16 g3 'iVe7 17 ~gS .i.f6 (l7 ... 'iVxe6 18 'ti'd8+ .i.f8 19 0-0+-) 18 ~xf6 'iVxf6 19 l'lfl 'ti'xe6 20 'iVd8+ 1-0 Wessman­Melgosa, Tunja 1989.

c) 13 ... .i.d4 and then: c1) 14 'i'b3?! is very unclear,

14 ... 'iVe7 IS .i.e3 lbeS 16lbxeS .i.xe3 17 lbdS .i.f2+ (After the alternative 17 ... 'ti'h4+? 18 g3 .i.f2+ 19 ~e2 'ti'e4+ 20 lbe3 'iVxeS 21 e7+ .i.e6 22 exf8'i'+ I'txf8 23 "iYxe6+ White was better in Mestel-Hodgson, Brit­ish Ch., Southport 1983) 18 ~e2 'iVxe6 19 lbc7 'Wxb3 20 axb3 .i.d4! 21 lbxa8 .i.xeS 22 I'txa7 ~g4+ with enough com­pensation for the exchange -Martin.

c2) 14 exd7! ~xd7 IS .i.gS 'iVe8+ (lS ... 'iVb6? 16 'i'b3+!+-) 16 .i.e2 lbf4 17 .i.xf4 .l:f.xf4 18 lbdS! lixf3 (the only move) 19 gxf3 'i'eS 20 lbc3 and White's enormous material advantage assures him of victory, Simon­cini-Caruso, corr. 1989.

12 g3 12 ~d2!? is very complicated

- 12 ... fxe6! 13 dxe6 and now: a) 13 ... O-O? is bad for Black.

14 exd7 ~xd7 IS .i.xd7 liad8 (lS ... I'tf2+ does not help: 16 lbge2 lid8 17 'ti'a4! .i.xc3+ 18 'itxc3 'ilif6+ 19 'itc2 l'lxe2+ 20 ~bl+-) 16 ~c2 .l:tf2+ 17 Wbl .i.xc3 18 'i' g4! .i.f6 19 lbh3 l'lxd7 20 lbxf2 1-0 Alexandrov-

Taimanov System 123

Wojtkiewicz, Wisla 1992. b) 13 ... .i.xc3+! 14 bxc3 (14

~xc3? 'ilib4+ IS ~c2 'ilixbS 16 exd7+ ~xd7 17 'i'el + 'itf7 hands the advantage to Black) 14 ... 0-0 15 lbf3! (IS exd7? .i.xd7 16 .i.xd7? ~f2+ 17 lbe2 .l:td8 18 ~c2 'i'e4+ 19 ~b3 .u.xe2 20 .i.gS ':xd7 is losing for White) IS ... 'fif2+ 16 'fie2 'i'xe2+ 17 ~xe2 lte8 18 ~f2 lixe6 19 .i.c4±. Probably the play of both sides can be im­proved. It is worth noting that very few players of the white side choose this risky line.

12 lbxg3 13 hxg3

Weaker are: a) 13 lbf3? i.xc3+ 14 bxc3

'jVe4+ 15 ~f2 lbxhl + 16 'i'xhl fxe6 17 dxe6 O-O! 18 exd7 .i.xd7 19 i.xd7, as in Little­wood-Hartoch, London 1984, and now 19 ... l:tad8! 20 .i.b5 c4 21 i.e3 %:td3 22 ltel g5! 23 h3 hS! with a strong attack.

b) 13 exd7+?! i.xd7 14 .i.xd7+ Wxd7 IS 'i'a4+ 'ti'xa4 16 lbxa4 lbxhl 17 lbf3 .l:f.he8+ 18 ~fl i.d4 19 'itg2 .tIe2+ 20 ~xhl ~d6 with a better ending for Black.

13 'ti'xhl Bad is 13 ... 'i'xg3+? 14 ~d2

.i.xc3+ 15 bxc3 'ilig2+ 16 'i'e2 'i'xdS+ 17 ~c2 'ilVxe6 18 'ti'xe6+ fxe6 19 i.h6 with a de­cisive advantage for White, Fecht-Betker, corr. 1989 .

14 ~e3! Another option for White is

Page 126: Beating King's Indian and Benoni

124 Taimanov System

14 exd7+ ii..xd7 and now:

a) After 15 ~e2+? Black has: al) 15 ... ~d8 16 .tg5+! f6 17

0-0-0 lte8 18 ~f1 fxg5 19 ii..xd7 ~xd7 20 ~b5+ was play­ed in Savchenko-Sandler, Bel­gorod 1989, and here Savchen­ko gives the variation: 20 ... ~c7 (the only move) 21 d6+ ~d8 22 tbd5 ~ g2 23 tbc7 ~xb2+ as finally leading to equality. This analysis is innacurate, however, as 22 ... ~g2?! can be improved by 22 ... b6!?, 23 tbc7? by the move 23 ~xb7 and finally 23 ... 'iVxb2+? by the decisive 23 ... l!e4!-+.

a2) 15 ... ~f8! 16 iLe3 .l:.e8! 17 ~xd7 (17 ~d2? proposed by some commentators as good for White is refuted by 16 ... ii.h6! 18 tbdl "VJk xdS+-+) 17 ... ltxe3 18 ~xe3 ii..d4 19 ~f3 'iVxgl+ 20 ~f1 ~xg3+ 21 ~d2 ~g7 is clearly better for Black, Gil­Kuczynski, Sharjan 1985.

b) 15 ii..xd7+ ~xd7 and then: bl) After 16 ~a4+ Black can

play 16 ... ~d8 (16 ... ~c8?! 17 ii..e3 ii..xc3+ 18 bxc3 'ii xd5 19 ltdl 'iic6, Rabinovich-Cordy,

European Junior Ch. 1993, 20 ~c4!n) 17 ii..g5+ f6 18 .li.e3 .l:.e8 19 ~d2 iLh6 20 ii.f4 with complicated play, Helmers­Obers, Netherlands 1995.

b2) 16 ~g4+ f5 (l6 ... ~d8? gave White a powerful attack in the game Zaja-Kutuzovic, Cro­atia 1995: 17 iLg5+ f6 18 O-O-O! Ite8 19 tbge2 'iih2 20 tbf4! fxg5 21 tbe6+ :xe6 22 dxe6+ 1-0) 17 ~a4+ ~c8 18 iLe3 iLxc3+ (After 18 ... i.h6?! White quickly developed a winning attack in the game Burgess­Anderson, London 1985: 19 ii..xc5 'ii g2 20 ~ c4 :e8+ 21 tbge2 ~xg3+ 22 ~dl ~g4 23 ii..d4+ ~d7 24 d6 :ac8 25 'iif7+ ~xd6 and now 26 tbb5++-) 19 ... bxc3 ~xd5 20 :dl ~c6 and now:

b21) The endgame after 21 ~xc6+ bxc6 22 ii..xc5 is equal. Sandler-Gobleja, USSR 1988, continued 22 ... :e8+ 23 ~f2 ltb8 24 :d2 a6 25 tbf3 ~b5 26 .1i..e3 :d5 27 :b2 ~b5=.

b22) After 21 ~c4 b6 (Or 21...:e8 22 ~f2 b5 23 ~xc5 ~xc5 24 .li.xc5 ~c7 25 tbf3 ltad8 26 .li.d4 which was slight­ly better for White, Comas­Gallego, Spain 1991) White had a certain initiative that was not converted into anything serious in the game Comas-Martin, Catalonia Team Ch. 1996: 22 'iia6+ ~b7 23 ~a4 ~d8 24 :xd8+ ~xd8 25 i.g5+ ~c7 26 .li.f4+ Wd8 27 ~f2 ~e4 28 ~b3 ~d7 29 ~b5+ We7 30 tbf3 Ite8

Page 127: Beating King's Indian and Benoni

31lbe5=. 14 ~xc3+!

Weaker is 14 ... 0-0?! (Not 14 .. .fxe6? which loses to 15 dxe6 0-0 16 exd7 iLxd7 17 iLxd7 ~ad8 18 'i'b3+ 'it'h8 19 0-0-0+-) 15 exd7 iLxd7 16 iLxd7 ~ae8! 17 iLxe8 ~xe8 reaching a position much ex­plored by English players in the mid-1980s:

a) White achieved only a small edge after 18 'it'd2 iLxc3+ 19 bxc3 'i'xd5+ 20 'it'c2 'i'e4+ 21 'i'd3 'i'xe3 22 ~f1!, Crouch­Martin, England 1985

b) Better is 18 'i'e2! ~d4 (18 ... iLh6!? 19 ~f1) when:

b 1 ) 19 0-0-0 ':xe3. This po­sition was reached in the game Littlewood-Norwood, Comm­onwealth Ch. 1985. Now in­stead of 20 'i'c4?! a6 White could have played 20 'i'c2! with a clear advantage. For example: 20 ... ~xg3 21 lbge2 'i'h6+ 22 ~bl ~g2 23 'i'a4!± (Konikow­ski).

b2) 19 ~d2!? b5 20 ~dl b4 21 iLxd4 ~xe2+ 22 lbgxe2 bxc3+ 23 il.xc3 'i'xd5+ 24 'it'c1 also gave an advantage to White in Flear-Norwood, London 1984.

15 bxc3 a6 15 .. jWe4?! does not promise Black an

easy life: 16 'i'f3 'i'xf3 17lbxf3 Less convincing is 17 exd7+

j,xd7 18 j,xd7+ ~xd7 19 lbxf3, as in Kouatly-Schmitt-

Taimanov System 125

diel, Augsburg 1989, because of 19 ... 'it'd6!? 20 c4 f6 and Black can hold the position.

17 .. .fxe6 18 dxe6 O-O?! More resilient is 18 ... a6 19

exd7+ iLxd7 20 j,xd7+ 'it'xd7 21 .i..xc5. White has an extra pawn compared to a similar ending in Comas-Gallego. The game Urban-Ciemniak, Polish Ch. 1993, continued 21...~c6 22 .i..e3 ~ae8 23 ~d2 ~e4 24 lbd4+ 'it'd5 25 'it'd3 ~c8 26 lbc2 ~ce8 27 ~bl with an ad­vantage for White.

19 j,h6!! Only this fantastic move con­

firms White's superiority in this position. In the game Lautier­Ivanchuk, Monaco (rapid) 1995, White played 19 ~e2?, and af­ter 19 ... lbf6 20 e7 IH7 21 iLxc5 il.g4 22 ~dllbd5 23 ~d3 4:Jxe7 24 il.c4 lbc6 25 ~f2 ~e8 26 iLxf7+ 'it'xf7 27 lbd4 4:Je5 Black gained an advantage. Was Ivanchuk bluffing or had he found an improvement for Black?

19".~xf3 19 ... ~e8 changes nothing,

Page 128: Beating King's Indian and Benoni

126 Taimanov System

e.g. 200-0-0 tbf6 (20 .. Jhe6 21 ..tc4 tbb6 22 tbg5!+-) 21 ..txe8 ..txe6 22 ..ta4 ~xa2 23 tbe5±.

20 l:td I! ~xc3 Or 20 ... tbf8 21 e7 .td7 22

exf8if+ ~axf8 23 ..tc4+ which is winning for White.

21 exd7 .txd7 22 ~xd7 a6 23.tfl !

This winning move was sug­gested by Kalinin after his cor­respondence game against Ko­nev (1991): 23 ~g7+? ~h8 24 ~xb7 axb5 25 ~g7+ ~g8 26 ..txc3 b4 27 ~e5;j;;.

23 ... ~f8 24 ~g7+ ~h8 25 ~xb7

and White wins. 16 exd7+ ~xd7 17 ..txd7+ ~xd7

18 ifb3! A fresh idea that was intro­

duced in this game. White's plan - to keep the queens on and to play for the attack - is very strong. Before this game White had always gone into various fonns of the endgame:

a) 18 iff3 ifxf3 19 tbxf3

(similar to Kouatly-Schmittdiel in the note to Black's 15th move) 19 ... ~he8 (19 ... ~d6!? deserves attention, e.g. 20 c4 f6 with the idea of meeting 21 a4 b5 22 tbd2 with 22 ... ~ae8!) 20 ~f2 when Black has tried:

al) 20 ... ~e4 21 ..txc5 ~ae8 22 ~e1 ~xel 23 tbxel b6! 24 ~xb6 ~d6 25 tbc2 ~xd5, Thuesen-De Finnian, Farum 1993, and now White should have played 26 tbb4+ ~e4 27 a4 with at least equal chances.

a2) 20 ... ~d6?! is too slow. 21 c4 f6 was played in Illescas­Topalov, Alcobendas 1994, and now according to Illescas 22 a4! was strong, e.g. 22 ... b5 23 tbd2±.

b) 18 if a4+ b5 and then: bl) 19 ifg4+ f5 20 iff3 ifxf3

21 tbxf3 ~he8 22 ~f2 ~e4 23 tbg5 ~c4 24 tbxh7 ~h8 25 ~hl ~c8 26 d6 with a complicated ending, as in Remlinger-Brown­scombe, San Mateo 1994.

b2) 19 'i'f4 ~he8 20 0-0-0 'i'e4 21 ~xc5? (Perhaps 21 ifxe4!?) 21...ifxf4+ 22 gxf4 ~ac8 23 .td4 ~e4 24 tbf3 (Not 24 tbh3? ~xd4!-+) 24 .. Jhf4+ Remlinger-Shabalov, Las Vegas 1993.

c) Also deserving attention is 18 ifg4+ f5 19 iff3 (19 ifg5 promises nothing after 19 ... "ife4 200-0-0 h6) 19 ... "ifxf3 20 tbxf3

18 bS 19 0-0-0

After 19 ~xc5 ~ac8 is not so good for Black due to 20 .tf2

Page 129: Beating King's Indian and Benoni

lIheS+ 21 'it'fl lIc4 (Stohl) and then 22 a4! A more promising alternative is 19 ... 'ii'g2!?

19 ~he8 It is not easy to defend

against the threat of penetration by White's queen into the black camp:

a) 19 .. J:tacS?! 20 d6! [20 'i'a3!?] 20 ... c4 21 'ii'c2 ~heS 22 ~f2 f5 23 ttJf3±, Sokolov.

b) 19 ... c4?! 20 "i1Vb4 .l::!.hcS 21 ~g5±.

c) 19 ... 'ii'g2 20 d6! (20 ~a3!?) 20 ... c4 21 "i1Vc2 followed by 22 'ii'f2 with a clear advan­tage.

d) 19 .. J:thcS?! 20 il.xe5

20 ~ae8?! A critical alternative try is

20 ... 'ii'g2!? Black's previous move (l9 ... ~eS) allows him to meet 20 d6 with 20 .. J:Le6, while this queen move prevents the development of White's knight and covers the important c2-square. After 21 d6 ~e6 White has:

Taimanov System 127

a) The direct 22 ttJh3? did not work in the game Cuevas Rodriguez-Eriksson, Yerevan Women's Olympiad 1996: 22 ... 'ii'xh3 23 'ii'd5 ~bS 24 il.a7 "i1Vh6+ 25 'itb1 'ii'fS with ad­vantage to Black.

b) White should try 22 iLb4!? preparing 23 a4. At least in the case of 22 ... 'ii'xg3 23 'ii'd5 ~e3+ 24 'itb2 'ii'e4 25 ~c5 'ilVc6 26 "i1Vf2! White's queen and knight find freedom and can organise a strong at­tack.

21 il.d4 'ii'g2 After 21...~c4 White's knight

enters the game with unpleasant consequences: 22 ttJe2 'fIf3 23 ttJf4 is clearly better for White. 21.. . .l::!.e4, threatening 22 ... ~xd4, can be met by the quiet 22 'itb 1.

22 'ii'a3 As a consequence of the

move 20 .. J::tacS?! the a6-pawn is unprotected.

22 "i1Vxg3 After 22 .. J~aS both 23 c4!

and 23 'fIc5! are very strong. 23 'ii'xa6 ':xe3+

Faced with a bad position, Topalov tries a tactical trick. 23 ... 'i'g5+ does not help as 24 'it'c2 'i' g2+ 25 ~d2 'i'fl 26 ttJe2+-.

24 'itb2?! 24 'iitbl! ~b3+ 25 'ital!

would have won more quickly. 24 !tee8

Black could have given some more checks with 24 ... "i1Vg2+ 25 'iitxc3 ~cS+ 26 'itd3 but the re-

Page 130: Beating King's Indian and Benoni

128 Taimanov System

sult would have been the same. 2S 'iixbS+ ~d6 26 ~al 'iia3 27 ~b2 'iics 28 ~a6+ ~d7 29 'iia4+ 1-0

Game 18 Yuneev-Kostometov St Petersburg 1995

1 d4 tDf6 2 c4 cS 3 dS e6 4 tDc3 exdS S cxdS d6 6 e4 g6 7 f4 i.g78 i..bS+

8 tDfd7

If the conclusions of the pre­vious game are borne out by future practice, it follows that this is the only playable move after 8 i..b5+. White can be happy, however, as the harmony of Black's development has been disrupted. When the knight comes back from d7 to the natural square f6, we get a kind of Four Pawns Attack with an extra tempo for White.

For the purposes of complet-

ing White's repertoire we will analyse only one good line, 9 a4, although 9 i..d3, 9 i..e2 and 9 tDf3 are all not bad either.

9 a4 0-0 Black can also choose instead

9 ... 'iih4+ or 9 ... tDa6. The move 9 ... a6 will transpose to other lines, e.g. 10 i..e2 0-0 11 tDf3 is examined later in the game and 10 ~e2 'i'Vh4+ in line a below.

a) 9 .. JWh4+ 10 g3 Black probably has fewer

problems after 10 ~fl. White quickly obtained an advantage in the game Litinskaya-Fisch­dick, Baden Baden 1980: 10 ... 0-0 11 tDf3 'iie7 12 ~f2 tDa6? 13 ~el tDb4 14 ~gl a6 15 i.f1 ~d8 16 i.e3 'iif8 17 i.f2, but after 12 ... i..xc3! 13 bxc3 tDf6 Black is okay.

Kapengut suggests that Black should try and prevent White from 'castling by hand' by 10 ... a6! 11 tDf3 'iid8 12 i..d3 tDf6 13 h3 tDh5 14 ~f2 c4! 15 i..xc4 'i'Vb6+ 'with enough counterplay'. This needs further analysis - instead of 13 h3, 13 e5! looks strong.

After 10 g3 Black can try: al) 10 ... 'iid8 11 tDf3 0-0 12

0-0 a6 (l2 ... tDf6?! 13 e5! dxe5 14 fxe5 tDg4 15 ii'e2 a6 16 i..c4 ~e8 17 e6!± Finegold-Jaulin, Paris 1989) 13 i..e2!? (13 i..c4!? deserves attention, e.g. 13 ... tDb6 14 i..e2 i..g4 15 tDg5! i.xe2 16 'i'Vxe2 'i'Ve7 17 a5 tDc8 18 iLd2 tDd7 19 ~ael;j:; Olaf­sson-Psakhis, Moscow 1989)

Page 131: Beating King's Indian and Benoni

13 ... .l:.e8 (13 .. .'=t:'lf6?! 14 e5!) and now White has a pleasant choice between:

all) 14 .l:.e1 tDf8 (White has more than enough play for the pawn after 14 ... ~xc3?! 15 bxc3 .l:.xe4 16 c4 tDf6 17 i.b2, Ark­hipov-Sax, Hungary 1984) 15 i.f1 ~g4 16 h3 i.xf3 17 'i'xf3 tDbd7 18 i.d2 "fic7 19 b3 with a small but stable White advan­tage, Kouatly-Tringov, France­Bulgaria 1985.

a12)14 'it>g2 tDf8 15 e5! (15 h3!?) 15 ... i.g4?! 16 tDg5! ~xe2 17 ~xe2 f6 18 tDge4 dxe5 19 f5!± Savchenko-Pigusov, Aal­borg 1992.

a2) 1O ... ~cl 11 tDf30-0 11...~xc3+?! is very danger­

ous: 12 bxc3 ~xe4+ 13 'it>f2 0-0 14 .l:.e 1 (Interesting is 14 tZJg5 ~e7 15 i.d3 c4 16 i.xc4 tZJc5 17 .l:.el ~c7 18 i.f1 and White is better, Platonov-Berelovich, Russia 1996) 14 ... ~f5 15 i.f1! ~f6 16 ~a2 c4 17 i.e3 tZJc5 18 ~d4± Midoux-Caruso, Cannes 1995. 120-0~ The move 12 ... a6 creates

problems with the development of the queen's knight, e.g. 13 i.e2 (Also good for White is 13 j,d3 .l:.e8 14 'it>g2 tZJf8, Ser­gienko-Parkanyi, Nagy kanizsa 1993. Now instead of 15 h3 tDbd7 16 j,d2 .l:.b8 17 .l:.b1;!; White could opt for the attack­ing 15 f5!?) 13 ... tZJf6 14 e5 tZJe8 15 .l:.el! (Less clear is 15 e6)

Taimanov System 129

15 ... i.g4 16 e6 ~xf3?! 17 i.xf3 fxe6 18 dxe6 ~xc3 19 bxc3 tZJc6 20 .l:.b 1 with a large White advantage, Levitt -Zamansky, Groningen 1990.

After 12 ... tZJa6, White can choose to prepare e4-e5 or play it immediately:

a21) 13 e5!? dxe5 (Weaker is 13 ... tZJb4?! 14 tZJe4! tZJb6 [14 ... dxe5 15 d6 ~d8 16 fxe5 tZJc6 17 ~g5± ~b6 18 tZJf6+ 'it>h8? 19 tZJd5+- Pecenka-Nun, COIT. 1987] 15 tZJxd6 tZJ6xd5 16 i.d2 ~g4 17 ~b3 .l:.ad8 18 .tc4 tZJb6 19 ~xb4 cxb4 20 ~xf7+! .l:.xf7 21 tZJg5 with a big advantage for White, Bagi­rov-Malaniuk, Baku 1983) 14 d6 'i'd8 (l4 ... 'i'e6? 15 tZJg5 'i'f5 16 tZJd5+-) 15 tZJd5! e4 16 tZJg5 (16 tZJe5!?) 16 ... .td4+ 17 ~e3! .txe3+ 18 tZJxe3 tZJf6 19 f5 with the initiative (Kapengut).

a22) 131:!.e1!

13 ... tZJb4 14 e5!? White could also keep a

small advantage without major complications after the quiet 14

Page 132: Beating King's Indian and Benoni

130 Taimanov System

.i.fl. 14".a6 15 i.fl dxe5 16 d6

'i'e8 17 fxe5 b6?! It is better to continue with

17 ... ltJxe5 !? 18 ltJxe5 i.xe5 19 ltJe4! 'i'c6 (19 ... i.d4+? 20 'i'xd4+-) 20 i.g2 ~h8 21 .i.h6!, but White still has an initiative.

lRJii!? fxe6 19 .tc4 Black is in a very dangerous

position, as in Tal-V elimirovic, Moscow Interzonal 1982. Now, even after the best move 19 ... ~h8! 20 !Ixe6 'i'f7 21 .i.f4 i.b7 22 ltJg5 'i'f5 23 !Ie7 White's threats are still strong­er.

b) Black's attempt to imme­diately develop the queen's knight and attack White's bish­op on b5 before it can retreat to fl seems logical but fails tacti­cally:

9. "lba6 10 ltJo ltJb4?! Or 1O ... ltJc7 11 0-0 ltJxb5?

(White has a strong initiative similar to that in Kasparov­Nunn after 11...a6?! 12 .txd7 i.xd7 13 f5! 0-0 14 i.g5, Baumbach-Danner, corr. 1985. Better is 11...0-0 transposing to variation a) 12 ltJxb5! ltJb8 (Al­ternatively 12 ... ltJb6? 13 a5+-; 12 ... 'i'b6? 13 ltJd2+-) 13 f5 (13 e5!?±) 13 ... a6 14 ltJc3 0-0 15 i.g5 'i'b6?! 16 'i'd2± Trosti­anecky-Sliapkin, corr. 1988

11 0-0 a6?! 11...0-0 was necessary, trans­

posing to the note to Black's 11th move.

12.txd7+! Black's knight on b4 is not

well placed in this line and even 12 ..te2 is not bad for White, but the text is more energetic.

12 ... .txd7 13 f5! 0-0 Black does not have much

joy after 13 ... gxf5 14 .tg5 f6 15 .tf4 'i'c7 16 ltJd2 0-0-0 17 ltJc4! or 13 ... c4 14 .tg5 'i'b6+ 15 ~h1 ltJd3 16 f6 .tf8 17 a5! ltJf2+ 18 ~xf2 'iYxf2 19 ltJa4 (Kasparov) .

14.tg5 Kasparov-Nunn, Lucerne Ol­

ympiad 1982. Even after the best move 14 ... ..td4+! 15 ~h1 f6 16 .th6 !Ie8 White has a dangerous initiative.

10 ltJf3 ltJa6 Alternatively: a) White created a decisive

attack after the SUSpICiOUS looking 1O ... ~e8 11 0-O..txc3?! 12 bxc3 ~xe4 13 ltJg5 (13 .td3!? ~e8 14 c4) 13 ... ~e7 14 f5 ltJe5 15 f6 l::tc7 16 'i'el± b6? 17 ltJxh 7! V aisser -Chess Genius Aubervilliers (rapid) 1994.

b) The knight's return to f6 let White get a clear advantage

Page 133: Beating King's Indian and Benoni

after 1O .. .tiJf6?! 11 0-0 (11 e5!?) 11.. .ii..g4 (The inclusion of the moves 11...a6 12 .lte2 does not help Black, e.g. 12 ... .ltg4 13 e5 .ltxf3 14 .ltxf3 dxe5 15 fxe5 ttJfd7 16 e6 ttJe5 17 .ltg4± Semkov-Popov, Bul­garian Ch. 1980) 12 .l:tel ttJbd7 13 e5 dxe5 14 fxe5 ttJh5 15 ii..xd7! "iVxd7 16 h3 ii..xf3 17 "iVxf3, Vaisser-Schalkx, Ostend 1992.

c) 1O ... a6?! 11 Ae2 and now: c1) 11...f5 trying to justify

8 ... ttJfd7 weakens the e6-square and can be met by 12 ttJg5!?, 12 exf5 or even 12 0-0.

c2) Black is far from equality after 11..."iVc7 12 0-0 c4 and now:

c21) In Chekhov-Berelovich, Russia 1996, White played 13 .lte3 with success: 13 ... ttJc5 14 .ltxc4 ttJxe4 15 ttJxe4 "iVxc4 16 ttJxd6 "iVb4 17 ttJxc8 .l:txc8 18 f5±.

c22) More complicated but also good for White was 13 ttJd2 b5 14 axb5 ttJb6 15 ~hl ttJ8d7 16 e5! dxe5 17 ttJde4 .ltb7 18 bxa6 .l:txa6 19 ~xa6 .ltxa6 20 f5!± Li Zunian-Sax, Bie1 Interzonal 1985.

c3) 11.. .lIe 8 120-0 ttJf8 12 ... ttJf6 transposes into the

main line of the Four Pawns Attack with an extra tempo for White. After 13 e5 dxe5 14 fxe5 ttJg4 15 JLg5 f6 (15 .. :i1Vb6 16 a5!±) 16 exf6 ..txf6 17 'tid2

Taimanov System 131

ii..f5 18 h3 White is clearly bet­ter.

13 e5!? ttJbd7 No better is 13 ... ii..g4 14

ttJg5! ii..xe2 15 "iVxe2 dxe5 16 f5 e4 17 ttJgxe4 with advantage to White.

14 ttJg5! dxe5 15 f5 A typical attacking manoeu­

vre. 15.)2Jf6 16 g4! b5! 17 axb5

c4 18 Ae3 h6 19 ~ge4 ~xe4 20 ttJxe4 ii..b7 21 ii..xc4 axb5 22 ':'xa8 'i'xa8 23 i.b3

with a stable White advan­tage, Petursson-Perenyi, Saint John 1988;

11 0-0 ttJc7 After 11... ttJb4?!, the position

of Black's knight on b4 looks pleasant but is far from the main battle:

(see following diagram)

12 .l:te I! (Preparing e4-e5 and ii..f1 at the same time. 12 f5!? also deserves attention, e.g. 12 ... ttJe5 13 ttJxe5 iLxe5 14 ii..f4 i.d4+ 15 ~hl a6 16iLd3;!; Pein-Ivanka. Budapest 1990)

Page 134: Beating King's Indian and Benoni

132 Taimanov System

12 ... a6 13 Jtfl. Now White qui­etly prepares e4-e5 by .1c1-e3-f2-g3 and, if necessary .1c4. Black practically has no coun­terplay and must await the exe­cution of White's plan:

a) Here are convmcmg ex­amples from recent practice after 13 ... lbf6?! 14 h3!:

al) 14 ... h5 IS .1e3 1:ie8 16 Jtf2 b6 17 .1h4 'i'c7 18 .1c4 lbh7 19 eS j.fS 201:ic1 Jth621 Jtg3 'i'd7 22 e6!± Matamoros­Hernandez, Capablanca Memo­rial 1996.

a2) 14 ... b6 IS .1e3 .1b7 16 'i'd2 'i'c7 17 .1c4lbd7 18 .1f2 1:lae8 19 .1g3 .1d4+ 20 ~hl lbf6 211:iacl lbhS 22 .1h2 Jtg7 23 e5± Dao-Kristensen, Am­sterdam 1996.

b) 13 .. ..l:Ie8 14 h3! and now: bl) Passive defence leads to

an unpleasant position: 14 .. Jlb8 IS .1e3 lbf6 (IS ... b5 16 axb5 axbS 17 lbxbS lbf6 18 e5 dxeS 19 JtxcS±) 16 j.f2 lbhS 17 g3 lbf6 18 'i'd2 lbd7 19 g4 b6 20 1:lac1 .1b7 21 .1c4 Jta8 22 .:tcdl 'i'c7 23 .1g3± Tataev­Blodstein, Voskresensk 1993.

b2) Over-ambitious activity, however, was also quickly pun­ished in the game Maximenko­Kotsur, Azov 1991: 14 ... lbf6 15 .1c4 lbd7 16 Jte3 h6 17 .1f2 gS?! 18 eS! dxeS 19 fxgS hxgS 20 lbe4 lbb6 21 lbfxgS! Jtf5 22 b3 with a large White edge.

b3) The game Komarov­S.Kovacevic, Massy Open 1993 saw interesting complications ending up in White's favour: 14 .. .fS IS eS! dxeS 16 d6 e4 17 lbgS 1:if8 18 lbe6 'i'f6 19 lbxg7! ~xg7 20 .1e3 b6 21 a5 bS 22 lbdS lbxdS 23 'i'xdS 1:ia7 24 b4!±.

12 Jtc4 12 .1xd7!? deserves attention

e.g. 12 ... .1xd7 13 f5 gxf5 14 .1gS f6 IS ii.f4, as in Djukic­Mihajlovic, Budva 1996, while 12 .1d3 is another promising move.

12 1:ie8?! It is dangerous to leave the

f7-pawn unprotected while the white rook is still on fl. Thanks to this factor, the central break-

Page 135: Beating King's Indian and Benoni

through e4-eS becomes danger­ous.

Instead: a) After 12 ... a6 13 .:reI (13 eS

dxeS 14 d6 lZJe6 IS .txe6 fxe6 is not dangerous for Black) Black has a number of moves:

a1) After 13 ... bS?! 14 axbS lZJb6 IS .tfl axbS 16 ~xa8 lZJbxa8 17 lZJxbS lZJxbS 18 .txbS ~b6 19 ..tc6 .tb7 20 ..txb7 ~xb7 21 eS lZJc7, Skem­bris-Bellon, Genova 1989, Black has insufficient compen­sation for the pawn.

a2) 13 ... .:rb8 14 eS (14 as?! bS IS axb6 lZJxb6 16 .tfl=) 14 ... bS is better than its bad reputation but still White can hope for some advantage after IS axbS (Black has a strong po­sition after IS e6?! bxc4 16 e7 ~e8 17 fS lZJf6) lS ... lZJb6, Si­lakov-Kristol, USSR 1972, and now instead of 16 e6? fxe6 17 dxe6 ~e7 18 ..tfl axbS+ White should have played 16 exd6! (16 .tfl?! gives nothing after 16 ... axbS 17 lZJxbS lZJxbS 18 ..txbS dxeS 19 .tc6 .tb7) 16 .. :~xd6 (Or 16 ... lZJxbS 17

Taimanov System 133

..txbS axbS 18 lZJxbS;;!;; while 16 ... lZJe8 17 ..tfl axbS 18 lZJeS!;;!;) 17 lZJeS! lZJxc4 (17 ... axbS? 18lZJe4±) 18lZJxc4.

a3) White won quickly after 13 ... ~e8 14 eS dxeS? IS d6 lZJe6 16 fxeS lZJd4 17 ..tgS! Flear-Oei, Mondorf 1991. Criti­cal for this line is the compli­cated position after 14 ... lZJb6 IS .ta2..tg4.

a4) 13 ... lZJb6!? deserves at­tention and is similar to line b, e.g. 14 ..ta2 .tg4 IS h3 ..txf3 16 "iYxf3 as in Chachere-Blees, Krumbach 1991.

b) 12 ... lZJb6 13 .ta2 .tg4 14 h3..txf3 IS ~xf3. White seized the initiative in Ielen-Skembris, Cannes 1995, after lS ... lZJd7?! 16 eS! dxeS 17 fS lZJe8 18 lZJe4 c4 19 ~h1 l:tc8 20 .tgS f6 21 ..te3 a6 22l:tac1 bS 23 b3. More precise is lS ... a6!? (Psakhis).

13 eS!?

13 dxeS Black also has problems after

13 ... lZJb6 14 ..ta2 ..tg4 (Or 14 ... aS IS lZJe4! dxeS 16 fxeS

Page 136: Beating King's Indian and Benoni

134 Taimanov System

CDbxd5 17 it.g5 'iY d7 18 CDd6±) 15 h3 it.xf3 16 'iixf3. For ex­ample: 16 ... 'iid7 17 a5 CDc8 18 CDe4 dxe5 19 fxe5 .ll.xe5 (19 ... .:.xe5 20 CDxc5) 20 it.h6! 'tWf5 (20 ... ~d4+ 21 ~hl f5 22 d6+ ~h8 23 dxc7± ':'xe4?? 24 'iYxe4) 21 d6! it.d4+ 22 ~hl ':'xe4 23 'iixe4! 'iixe4 24 ':'xf7 winning.

14 CDg5 e4?! After 14 ... CDb6 15 it.a2 e4

(l5 ... h6 16 CDxf7 ~xf7 17 a5 CDd7 18 fxe5+ ~g8 19 e6±) 16 d6 CDe6 17 a5 CDd7 18 CDcxe4 threatening 19 f5 White has a strong attack according to Se.Ivanov and Yuneev.

15 d6 CDe6

16 f5! it.d4+ It is not easy for Black to find

a defence now.

17 ~hl CDxg5 18 fxg6 hxg6

I would be surprised if Black can survive after 18 ... ~g7 19 gxf7 ':'e5 20 CDb5.

19 it.xg5 'iixg5 Or 19 ... it.f6 20 ':'xf6! CDxf6

21 CDd5 CDg4 22 it.xd8 CDf2+ 23 ~gl CDxdl 24 CDf6+ ~f8 25 it.e7+ winning.

20 it.xf7 + ~g7 20 ... ~f8 loses more quickly

after 21 i.xe8+ ~xe8 22 CDxe4 'i'd5 (22 ... 'iWe5 23 'i'b3) 23 'i'g4CDf8 24 ':'ael!

21 i.xe8 CDe5 22 lLlxe4 'i"h4 23 lLlxc5! .ll.f5

24 ':'xf5! ':'xe8 24 ... gxf5 25 'i"xd4+-.

25 d7 lLlxd7 26 g3 1-0

Page 137: Beating King's Indian and Benoni

8 Mikenas Attack

1 d4 lOf6 2 c4 c5 3 d5 e6 4 lOc3 exd5 5 cxd5 d6 6 e4 g6 7 f4 iLg78 e5

In this chapter we discuss the Mikenas Attack, in which White pushes e4-e5 without any of the preparation that we see in other variations of the Four Pawns Attack. This overly­aggressive system was popular at the end of the 1950s and the first half of the 1970s, but nowadays it has been practically abandoned in favour of the Taimanov variation and the main line of the Four Pawns Attack. As you will see below, . m~nv of thp. okl thp.oretical that

have been passed from book to book are either incorrect or in­complete. I believe that there are still many more improve­ments to be found in this sys­tem. The variations below are mostly given from Black's point of view.

Game 19 Meszaros-Stefanov

Satu Mare 1987

1 d4 lOf6 2 c4 c5 3 d5 e6 4 lOc3 exd5 5 cxd5 d6 6 e4 g6 7 f4 iLg78 e5

8 lOfd7 This is more solid than 8 ... dxe5?! 9 fxe5 lOfd7 10 e6 which gives White a danger-

ous initiative. Let us look at the main line of this variation:

1O ... fxe6 11 dxe6 'iYe7! lZ lOd5! 'l'xe6+ 13 'jj'eZ! 'iYxeZ+ 14 .txeZ 0-0

After 14 ... .te5 15 lOf3 lOf6 16 iLc4 lOxd5 17 iLxd5 iLf6 18 0-0 lOc6 19 .tg5 White has a strong attack even without the Clllp.p.n<;. on thp. hoanl T .t1l1tiaTI-

Page 138: Beating King's Indian and Benoni

136 Mikenas Attack

Norwood, Lvov 1986. 15 CDc7 CDc6 16 CDxa8 CDb4 In this critical position in­

stead of giving back a rook with 17 CDf3?! CDc2+ 18 ~dl CDxal 19 ..tc4+ ~h8 20 l':te1 a6! 21 i.e6 CDe5! with complications favourable to Black, Shereshev­sky-Semeniuk, USSR 1974, White should have played 17 ~d1. Certainly, Black has a strong initiative, but it is a moot point whether this is sufficient compensation for a whole rook. For example: after 17 ... CDeS 18 i.d2 CDbd3 (or 18 ... l':td8 19 a3 CDbd3 20 ..txd3 CDxd3 21 ~e2 ..tg4+ 22 CDf3) 19 ..txd3 CDxd3 20 CDf3 CDf2+ 21 ~e2 CDxhl 22 l!xhl (Konikowski) Black does not have enough. Probably the best try for Black is 17 ... CDf6!? threatening 18 ... CDe4.

9 CDbS The alternatives are: a) After 9 e6?! fxe6 10 dxe6

CDb6! 11 CDe4 (Not 11 fS? 'ilHh4+) 11...0-0 12 'ilHxd6 "iVxd6 13 CDxd6 ~xe6 14 CDxb7 CDa4 Black has more than enough compensation for a pawn. The game Mileika-Elkon, Riga 19S9, continued IS CDf3 CDd7 16 i.bS CDxb2 17 l':tbl l':tab8! with better chances for Black.

b) In the case of 9 exd6?! 0-0 10 CDf3 CDf6 11 ~e2 CDe8 12 0-0 CDxd6 with equality we can see a difference compared to the similar line 8 CDf3 0-0 9 i.e2 l':te8 10 eS CDfd7?! 11 exd6 (see page <)) - Black's kniQht com-

fortably uses the free square e8. c) 9 CDf3 0-0 10 i..e2 dxeS 11

0-0 does not promise much: c1) After 11...a6?! 12 a4 exf4

13 i.xf4 CDf6 14 h3 CDbd7 IS d6 CDhS 16 ~h2 White had strong pressure in Kristiansen-Holm, Denmark 1977.

c2) However, better is 11.. .e4 12 CDxe4 CDf6 13 CDc3, and now instead of 13 ... ~g4?! 14 h3 ..txf3 IS i..xf3 'ilHb6 16 i..e3;!;; Nogueiras-Griinfeld, Zagreb In­terzonal 1987, Black should have played 13 ... CDe8, trans­posing to the solid position of variation b minus one, practi­cally insignificant tempo.

d) 9 CDe4 is just a transposi­tion of moves after 9 ... dxeS 10 CDd6+.

9 dxeS 10 CDd6+ ~e7

The correct square for the king. After 1O ... ~f8?! 11 CDf3! h6 12 i..e2, both 12 .. .fS 13 fxeS CDxeS 14 CDxc8 'ilHxc8 IS 0-0, Gliksman-Kosansky, Yugosla­via 1972, and 12 ... CDa6 13 0-0 e4 14 CDeS!? i.xeS IS fxeS CDxeS 16 CDxe4 ~g7 17 i..f4, Bozinovic-Sindik, Makarska 1994, were in White's favour.

11 CDxc8+ Other moves cannot be rec­

ommended: a) 11 fxeS?! CDxeS 12 CDxc8+

"iVxc8 13 d6+ ~f8 14 CDf3 "iVe6 IS CDxeS i..xeS 16 i.e2 ~g7 17 0-0 CDc6 with a clear plus to Black, Kavalek-Trapl, Czecho­!':lovll kill 1 q6~

Page 139: Beating King's Indian and Benoni

b) 11 ttJb5?! is too slow due to 11.. .~e8! 12 d6+ ~f8 and Black has completed his cas­tling by hand with an active position. Now Black stands better after both:

b 1) 13 .1i.e2 ttJc6 14 ttJf3 ttJd4 150-0 ttJxe2+ 16 ~xe2 exf4 17 ~dl ~e4 18 ttJc7 ~b8 19 ~d5 ttJf6 20 ~xc5 b6, Anikaev­Gorelov, Moscow 1981; and

b2) 13 ttJc7?! exf4+ 14 i.e2 (After 14 ttJxe8?! ~xe8+ 15 .1i.e2 ttJe5 16 .1i.xf4 ttJbc6, as in the game Smimov-Kapengut, Minsk 1979, White had the ad­ditional problem of developing his kingside) 14 ... ttJc6 15 ttJxe8 ~xe8 16 ttJf3 ttJd4 17 ttJxd4 .1i.xd4 18 .1i.xf4 ttJe5, Taylor­Donnelly, carr. 1990, and here the natural 19 ~ d2 .1i.f5 20 0-0-0 ~a4! would have given Black a strong attack.

11 ~xc8 12 ttJf3

12 d6+?! prevents Black from castling by hand but allows Black the c6-square for his knight and that factor seems more important here. After 12 ... ~f8 13 ttJf3 e4 (13 ... ttJc6!? is not bad either) 14 ttJg5 (14 ttJe5 ttJc6!) 14 ... h6 White is practically forced to sacrifice a knight with 15 ttJxf7 (After 15 ttJxe4 ~e8 16 ~e2 ttJc6, Par­tos-Holm, Skopje Olympiad 1972, White's position is very disagreeable) 15 ... ~xf7 16 .1i.c4+ ~f8 17 f5 i.d4!? (Black must be precise here, as demon-

Mikenas Attack 137

strated in the spectacular game Maffeo-Pastor, USA 1973: 17 ... g5?! 18 ~d5 ttJe5? 19 f6! .1i.xf6 20 0-0 ~g7 21 ~xf6! ~xf6 22 i.xg5+ with a very dangerous attack) 18 fxg6 ~g7 19 .1i.e6 ~ d8 20 ~b3 ~b6 21 ~g3 ttJf6 and White's attack is over, Sulava-Namgilov, Buda­pest 1990.

12 ~e8 As we shall see, the main line

of this variation eventually leads to equality. One of Black's possible ways of play­ing for a win is to try 12 ... e4!? 13 ttJg5 ttJb6 (It is amazing how many authors simply repeat Rajkovic's old analysis without verification: 13 ... ~e8? 14 d6+ ~f8 15 ttJxf7 ttJc6 16 ttJg5 h6 17 i.c4 hxg5 18 fxg5±. In fact after 15 ttJxf7? Black can sur­vive with 15 ... c4!, but 15 ~d5! wins for White: 15 ... ~g8 16 ~xf7+ ~h8 17 .1i.c4 and Black has no good defence against the threat 18 'iVg8+!) 14 d6+ (14 ~b3?! ~f5! 15 d6+ ~f8 16 g4 ~ d7 is good for Black) 14 ... ~f8 15 a4! (Not 15 .1i.c4? ttJxc4 16 'iV d5 ttJxd6 17 'iVxd6+ ~g8 180-0 .1i.d4+ 19 ~hl 'iYf5 20 g4 'iVd7 21 ttJxe4 'iVc6 22 ~el ~g7 with a decisive ad­vantage for Black, Kerr-Povah, London 1976) 15 ... h6 16 a5 hxg5 17 axb6 a6 18 'iVd5 'iVc6 19 .1i.c4 'iVxd5 20 .1i.xd5 ttJd7 21 .1i.xb7 ~b8 22 .1i.c6 ttJxb6 with an unclear endgame, Kooiman­Povah, London 1976.

Page 140: Beating King's Indian and Benoni

138 Mikenas Attack

13 SLc4 This is a critical moment for

White. Besides the text move he can play 13 fxeS or 13 SLe2.

a) After 13 fxeS Black can playa more or less

drawing variation, 13 ... ctJxeS, or take a risk trying to play for a win with 13 ... ~f8!?:

a1) 13 ... ctJxeS 14 .1bS ctJbd7 IS ctJxeS

IS 0-0 is just a transposition of moves after lS ... ~f8 16 ctJxeS. However, in this line lS ... a6? is a mistake due to 16 ctJxeS .1xeS 17 d6+! ~f8 (Or 17 ... ~xd6 18 Ile1+!) 18 'i!idS SLd4+ 19 ~h1 ctJf6 20 .1h6+ ~g8 21 Ilxf6! SLxf6 22 d7 and White wins.

After IS ctJxeS Black has a choice:

all) After lS ... ~xeS? 160-0 White stands better, as shown by the following practical ex­amples:

alII) 16 ... 'i!ic7 is met by 17 'i!ig4! ~d6?! 18 lIxf7 ~d4+ 19

~h1 ctJeS 20 lIf6+ ~xdS 21 'i!ie2 c4, Shereshevsky-Kapen­gut, Minsk 1974. Now 22 ~f4 would have decided the game.

al12) 16 ... ~f8 17 'i!if3 fS 18 g4 .1d4+ (Or 18 ... a6 19 gxfS! axbS 20 fxg6+ with a strong attack) 19 ~h1 ctJeS 20 'i!ig2 lId8 21 gxfS ~g7 22 .1gS lId6 23 Ilae 1 with a clear advantage for White, Rajkovic-Planinc, Maidanpek 1976;

al13) 16 ... c4 17 d6+ ~f8 18 .1h6+! (Black managed to sur­vive after 18 'i!idS 'i!icS+ 19 'iixcs ctJxcS 20 .1xe8 ~xe8 21 lIe 1 ctJd3 22 .1f4 ctJxe1 23 1:l:xe 1 f6 24 .1xeS fxeS 2S 1:l:xeS+ ~d7 26 1:l:e7+ ~xd6 27 Ilxb7 as in A. Petrosian-Kapen­gut, USSR 1975) 18 ... ~g8 19 'i!idS 'iicS+ 20 'iixcs ctJxcS 21 ~xe8 llxe8 and now not Kap­engut's 22 !!ae1?, spoiling White's advantage because of 22 ... ctJd3!, but 22 l:t.fe1!±, with the idea of 22 ... ctJd3 23 .1gS!

a12) lS ... Wf8! A necessity. 16 0-0 l:t.xeS 17 .tf4

In this position Black had se-

Page 141: Beating King's Indian and Benoni

rious problems until Kapengut's impressive discovery 17 ... c4! was revealed in 1975:

a121) 17 ... .l:Ie4?! 18 'i'f3 f5 19 .l:Iael ii.d4+ 20 Wh1 .l:Ixe1 (20 ... lbf6!?) 21 .l:Ixe1 lbf6, Ka­gan-Artishevsky, Minsk 1975, and now White could have played 22 d6! lbe4 23 d7! with a big advantage.

a122) 17 ... a6?! is met by 18 ii.xd7 'i'xd7 19 .txe5 .txe5 20 ~f3 (This is stronger than 20 'i'b3?! Wg7 21 J:tae1 f6 22 Wh1 b5 with equality, Shereshevsky­Savon, USSR 1975) 20 ... Wg7 21 d6! with a clear plus for White.

a123) 17 ... 1:H5 18 .td6+ (Or 18 g4 .td4+ 19 Wh1 .l:If6!) 18 ... Wg8 19 .l:Ixf5 gxf5 20 .txd7 (20 'i'f3!? deserves atten­tion, trying to use the bishop pair) 20 ... 'i'xd7 21 .txc5 .txb2 22 .l:Ibl .ll.e5 23 ~d2 f6 with equality, Kutin-Lobron, Yugo­slavia 1980.

a124) 17 ... c4t And now: a1241) 18 ii.xe5?! lbxe5 19

Wh1 'i'c5 20 .ta4 .l:Id8 and Black's chances are better.

a1242) 18 .txd7?! 'iHc5+ 19 Wh1 .l:Ixd5 20 'i'g4 (20 .td2 'i'd6!; 20 'i'a4 b5!) 20 ... f5 21 'i'h3 (21 .ll.e6? does not work due to 21...fxg4 22 .td6+ We8 23 .txc5 .l:Ixc5 24 .l:Iael .l:Ie5 25 .l:Ixe5 .txe5 26 .l:Ie11Id8! with a clear advantage for Black, Kap­engut) 21.. . .:t.xd7 22 'i'xh7 cJ;;f7 and Black keeps his extra pawn

Mikenas Attack 139

with the advantage, Yuferov­Kapengut, Minsk 1976.

a1243) 18 'i'd4! The only move that allows to

White to equalise. 18 ... .l:If5 Also possible is 18 ... .l:Ih5 19

'i'xc4 lbb6 20 'i'b4+ Wg8 21 .l:Iac1 (21 d6? as! 22 'i'b3 'i'c5+ 23 ii.e3 'i'xd6=t=) 21...'i'f8 22 'i'b3 J:txd5 and the position is not far from equality, Astolfi-S. Kovacevic, Cannes 1989. 19~ Not 19 .th6? .txh6 20 .txd7

.tg7! 21 'i'xg7+ Wxg7 22

.txc8 .l:Ixfl + 23 Wxfl .l:Ixc8 winning for Black, as in Legky­Shvedchikov, USSR 1978.

19.)2Jb6 Or 19 ... 'i'xc4 20 ii.xc4 ii.xb2

21 .l:Iadl lbe5 22 .th6+ We7 23 .l:Ixf5 gxf5 24 d6+ and a draw was agreed, Shakarov-Schmul­enson, COIT. 1976.

20 'ifxc8+ .l:Ixc8 21 .td6+ Wg8 22 J:txfS gxf5 23 .l:td1 Itd8 24 .ll.c5 l:txd5

with a drawn position in Yuneev-Korsuns10J, Alma-Ata 1980.

a2) 13 .. .'~f8!? Here White has tried: a21) 14 ii.b5?! a6 (l4 ... c4!?)

15 .ll.xd7 (Or 15 .ll.a4 b5) 15 ... lbxd7 16 e6 fxe6 17 0-0 Wg8 18 Wh1 exd5 19 'i'xd5+ Wh8 20 lbg5 lbe5 21 .te3 'ifc6 22 'i'xc6 lbxc6 23 .txc5 Wg8 with a better ending for Black, N. Garcia-Danailov, Alcoben­das open 1994.

Page 142: Beating King's Indian and Benoni

140 Mikenas Attack

a22) 14 e6 fxe6 15 .te2 Black gained a substantial

advantage in the game Vukovic­T. Petrosian, Bor 1980, after 15 d6?! ~g8 16 j,c4 ltJc6 17 0-0 ltJb6 18 j,b3 ltJd4 19ltJg5 Vi'c6 20 Vi'g4 ltJxb3 21 axb3 Vi'xd6 22 j,f4 Vi'd7. However, 15 dxe6!? deserves attention.

15 ... exd5 160-0

The game Sobek-Hardicsay, Ostrava 1979, saw 16 ... ltJf6? 17 ltJg5 Vi'c6 (Not 17 ... ~g8? 18 Ii.xf6! j,xf6 19 Vi'xd5+ ~h8 20 ltJf7+ ~g7 21 ..th6+ ~g8 22 ltJd6+ Vi'e6 23 ..tc4 winning or 17 ... ~e7? 18 j,b5 ltJc6 19 j,e3 d4 20 Vi'b3 with a strong attack) and now instead of 18 a4 c4 19 ltJxh7+ ~g8 20 ltJxf6+ ..txf6 21 Ii.xf6 Vi'xf6 22 Vi'xd5+ Vi'f7 23 Vi'xf7+ ~xf7 24 .txc4+ ~g7 25 b4 ltJc6 26 ..tb2+ ~h6 with equality, as occurred in the game, Hardicsay gives the vari­ation 18 ..tb5!? Vi'xb5 19 ltJxh7+ ~g8 (l9 ... ~e7? 20 ..tg5 ltJbd7 21 Vi'xd5+-) 20 ltJxf6+ ..txf6 21 'i'xd5+ ~g7 22 ..tg5 ltJd7 23 Ii.xf6 ltJxf6 24 ..txf6+ ~xf6 25 Ii.fl + ~g7 26 Vi'f7+

with a perpetual check. How­ever, the last of these lines can be improved on move 21 by 21 ~xf6 cJi;g7 22 ~d6! and Black is in trouble. For example: 22 ... 'i'e2? 23 Vi'xe2! Ii.xe2 24 Ii.d8 ~f7 25 ..tg5+-; 22 ... d4? 23 'i'd2! Ii.h8 24 'i'g5 Vi'e8 25 'i'f6+ ~h7 26 Ii.e6+-; and 22 ... 'i'b4 23 iLg5±.

So, can we conclude that 13 ... ~f8 has been refuted? No! In the position of the last dia­gram Black can play 16 ... cJi;g8! 17 Vi'xd5+ ~h8 and it is not evident how White can prove compensation for the pawn. For instance, 18 ltJg5 ltJe5 19 j,b5 ltJbc6 20 Vi'xc5 h6 21 ltJe4 ltJf3+ and 18 ..tb5 ltJc6 19 ..tf4 ltJb6 20 Vi'xc5 ltJd4! are clearly better for Black. So it seems that White's last chance in this line rests with 15 dxe6!?

b) 13 j,e2!? This is better than its reputa-

tion. After 13 ... ~f8 140-0 e4 14 ... c4!? is also interesting. 15 t2Jg5 h6?! 16 f5! hxg5 17

fxg6 The game F. Mainson-Podzi­

elny, Groningen 1974, now con­tinued 17 ... t2Jf6 18 ..txg5 Vi'd8 19 'i'b3 ltJbd7 20 ..tb5 Ii.e5 21 Vi' g3 Ii.xg5 22 'i'xg5 t2Jb6 23 Ii.f5 e3 24 Ii.afl t2Jbxd5 25 iLc4 e2 26 ..txe2 'i'd6 27 ~c4±. However, Larsen's improve­ment here:

17 ... t2Je5 has been quoted by every

Page 143: Beating King's Indian and Benoni

author as practically a refutation of this line. The reality is not as simple.

After 18 gx.f7 White has good compensation for a piece. Black has three ways of meeting the threat to his rook, but only one of these is playable:

bl) IB .. .'~Jxf7? 19 .i.hS 'fJ.e7 20 ~xgS 'fJ.d7 21 'i' g4 ttJa6 22 iLf6+-.

b2) IB ... 'fJ.dB? 19 .i.xgS 'fJ.d6 20 'i'a4! c4 (Black's alterna­tives are unattractive: 20 ... 'fJ.g6 21 ~hS! 'i'd7 22 'i'xe4! 'fJ.xgS 23 'i'h7 ttJa6 24 'i' gB+ rJite7 2S d6+!+-; 20 ... ttJxf7 is met by 21 'i'xe4 'i'eB 22 'i'xeB+ rJitxeB 23 iLhS winning; and 20 ... ttJbd7 21 'i'xe4 'fJ.g6 22 'i'h4 'fJ.xgS [Not 22 ... ttJxf7? 23 iLhS ttJdeS 24 iLxg6 ttJxg6 2S 'i'h7 ttJgeS 26 d6+-] 23 'i'xgS 'i'dB±) 21 'i'a3 'i'd7 (21...'i'c7 22 'i'h3!) 22 iLhS ttJa6 (Or 22 ... ttJd3 23 ~f4! ttJxf4 24 'fJ.xf4 'i' e7 2S 'fJ.xe4+-) 23 'i'e3! 'fJ.xdS 24 'i'xe4 and Black has no defence against 2S 'i'h7, e.g. 24 ... 'i'e6 2S ~7 ttJf3+ 26 'fJ.xf3 iLd4+ 27 ~e3! and White wins.

Mikenas Attack 141

b3) The only move is IB ... 'fJ.e7. Now White has sev­eral interesting possibilities:

b31) 19 iLgS 'fJ.xf7 20 'i'c2! with compensation. For exam­ple: 20 ... ttJbd7 21 'i'xe4 rJitgB (21...'i'c7 22 'i'h7!) 22 iLhS 'fJ.xfl + 23 'fJ.xfl 'i'c7 24 iLe7! and White is at least equal.

b32) 19 .i.hS c4! 20 d6 (20 iLxgS 'i'cS+ 21 rJithl ttJbd7! results in an unclear position) 20 ... 'i'cS+ 21 .i.e3! (An idea of Destrebecq's) 21...'i'xe3+ 22 rJithl 'fJ.d7?! (Black is also suf­fering after 22 ... 'i'cS?! 23 b4! 'i'b6 24 dxe7+ rJitxe7 2S fB'i'+! i.xfB 26 'i'dS±. It is therefore better to play 22 ... ttJbc6! 23 'i'dS 'i'd4!) 23 'i'dS ttJbc6 24 'i'e6 ttJe7 2S dxe7+ 'fJ.xe7 26 'i'fS 'i'd4 27 'fJ.adl 'i'xb2 2B 'fJ.f2 'i'xf2 (2B ... 'i'c3? 29 'iVh7 ttJxf7 30 'fJ.afl +-) 29 'i'xf2 ttJxf7 (29 ... ttJd3 30 'i'fS e3 31 'i'h7 'fJ.xf7 32 iLxf7 e2 33 'fJ.el!±) 30 'i'cs 'fJ.aeB 31 'fJ.fl rJitgB 32 'i'xc4 ~fB 33 g3 with advantage to White. These long variations can probably be im­proved but they illustrate the strength of the pawn on f7 per­fectly.

Returning to move IS. In­stead of IS ... h6?! a better choice is IS ... .i.d4+ 16 rJithl ttJf6 (16 ... h6?! allows a strong at­tack, as in the game F. Mein­sohn-Wittmann, Imperia 1973: 17 fS! hxgS [17 ... gxfS!?] IB fxg6 f6 19 .i.xgS rJitg7 20 i.g4! 'i' c7 [20 .. .fxgS 21 'fJ.f7 + rJitxg6

Page 144: Beating King's Indian and Benoni

142 Mikenas Attack

22 'i'£1 ~f6 23 i.hS+! 'iii.txhS 24 'i'fS mating] 21 'i'b3 'i'b6 22 ~h6+! and White soon won) 17 fS! 'iii.tg7 18 fxg6 hxg6 19 'i'b3 ll'lbd7 20 d6 ~f8 21 ~f4 bS!? with complicated play, Destrebecq-Michalski, corr. 1984.

13 ~f8 Another option is 13 ... ll'lb6!?

14.1i.bS (After 14 d6+?! ~f8 IS i.bS ll'lc6 16 0-0 ~g8 17 fxeS ~xeS 18 ll'lxeS ~xeS 19 i.f4 1:.dS White has insufficient play for a pawn) 14 ... 1:.d8 IS 0-0 1:.xdS 16 'i' e 1 ~f8 17 fxeS ll'lc6 (If 17 ... c4?! 18 ll'lgS! with a dangerous attack) 18 i.xc6 'i'xc6 19 'i'h4 ~g8 20 ll'lgS hS 21 ll'lxf7 1:.f8 22 ll'lh6+ ! (22 i.h6?, Mikenas-Suetin, Yere­van 1962, could have been punished by 22 ..• ~xj7! 23 ~xf7 i.xh6) 22 ... rJith7 23 ll'lf7 and Black has nothing better than 23 ... rJitg8 with a repetition.

14 0-0 e4 After 14 ... ll'lb6 IS i.bS l:td8

16 d6!? (Instead 16 fxeS trans­poses to the game Mikenas­Suetin) 16 ... a6 (16 ... exf4 17 i.xf4 ll'lc6 18 i.e3±, Bole­slavsky) 17 ~e2 (17 fxeS axbS 18 ll'lgS deserves further analy­sis) 17 ... e4 (Nunn) 18 ll'lgS ~d4+ 19 rJithl the position is very unclear. Now 19 .. .fS?! can be met by 20 'i'b3 c4 21 'i'h3 hS (21...'i'd7 22 i.e3) 22 i.e3! ll'lc6 23 .1i.xhS! ~xd6 24 .1i.xg6!±.

15 ll'lg5

Black was able to fend off White's attack in the game Ma1kotsi-Dzjordzjesku, corr. 1971, after IS ll'leS ll'lb6 16 i.bS ll'l8d7 (16 ... ~xeS!?) 17 ll'lxd7+ ll'lxd7 18 fS i.d4+ 19 'iii.thi a6 20 fxg6 hxg6 21 'i'g4 ll'leS 22 'i'h4 axbS 23 .1i.gS ll'lf3! 24 gxf3 e3=F.

15 ll'lb6 In the case of IS ... h6? White

should not play 16ll'lxf7? .1i.d4+ 17 rJithl rJitxf7 18 fS ll'lf6! 19 fxg6+ 'iii.tg7 20 ~xf6 i.xf6 21 'i'hS e3 winning, Gigerl-Griin­feId, Groningen 1974, but, as proposed by Yuneev, 16 fS! hxgS 17 fxg6ll'leS 18 d6!±.

16 ~b5 ~d8 16 ... i.d4+ 17 ~hl h6?! 18

fS! proved too dangerous for Black in Yuneev-Var1amov, Leningrad 1982. That game continued 18 ... hxgS 19 fxg6 f6 20 i.xgS ~g7 21 l:txf6! ~xf6? (21...~h8!) 22 'i'hS .1i.xgS 23 1:.£1! l:th8 24 1:.f7+ ~g8 2S 1:.h7 and White won.

17 ll'lxh7+?!

Page 145: Beating King's Indian and Benoni

It was better to play 17 d6!? transposing to the note with 14 ... tDb6 above.

17 ~g8 18 tDg5 c4! 19 'iNel

As so often in this system, White sacrifices a piece.

19 'iNc5+ 20 Jl.e3 'iNxb5 21 f5! Ihd5 22 fxg6

Perhaps 22 'iNh4!? gxf5. 22 ~xg5?

And as so often, Black de­fends badly against this attack. Here he underestimates the ex­tremely dangerous pawn on f7. It was necessary to play 22 ... fxg6 23 'iNh4 (Neither 23 tDe6 nor 23 tDxe4 is of any help to White) 23 ... ~xg5 24 Jl.xg5 tDbd7, when White's attack dis­appears but Black retains his material advantage.

23 gxf7+ ~fS 24 a4! 'iNa5

If 24 ... 'iNe5 25 'iNb4+. 25 'iNh4 ~f5?

Mikenas Attack 143

A losing move. 25 ... ~d5 26 'iNh7 (The amazing 26 'iNh8+?! .i.xh8 27 ~h6+ ~e7 28 f8'iN+ ~d7 is inferior; Black's chan­ces look better) 26 ... tDbd7 27 'iNg8+ cJ;;e7 28 'iNxg7 ~f8 would have kept Black in the game.

26 'iNh7 Less convincing is 26 ~xf5?!

'iNxf5 27 ~fl 'iNxfl + 28 cJ;;xfl±. 26 ~xf1+

If 26 ... ~xf7 27 ~xf7+ ~xf7 28 ~fl + mating.

27 ~xf1 28 iVg8+ 29 'iNxg7 30 ~dl?

tD8d7 ~e7 ~fS

White could have won with 30 Jl.g5+! cJ;;d6 31 'iNh6+! ~c7 32 .i.e7. Now it is Black's turn to take command.

30 31 'iNd4 32 iVd6 33 as 34 'iNd2 35 36 37 38 39 40 41 42 43 44 45 46 47 48

bxc3 'iNxd7 ~el ~hl ~bl 'iNc8+ iVxb7 iVb3+ g3 'iNb8+ l:txb8+ ~e8 ~e4 ~xg4+ 0-1

'iNh5! ~xf7! ~g8 ~f6 c3! tDc4 tDxe3 'iNc5 'iNxc3 'iNe5 I:lfS tDg4 ~f7 e3 'iNxb8 ~g7 ~f2 e2 ~f6

Page 146: Beating King's Indian and Benoni

Index of Variations

Main Line (1 d4 tiJf6 2 c4 g6 3 tiJc3 i.g7 4 e4 d6 5 f4 0-0 6 tiJf3 c5 7 d5 e6 8 i.e2 exd5 9 cxd5 or 1 d4 tiJf6 2 c4 c5 3 d5 e6 4 tiJc3 exd5 5 cxd5 d6 6 e4 g6 7 f4 i.g7 8 tiJf3 0-0 9 i..e2) 9.oo~e8

10 'iVc2 8 100-08 10 eS dxeS (lOoo.others 8) 11 fxeS tiJg4

12 e6 32 120-0 (12 i.f4 tiJxeS 130-0 transposes)

12oo.tiJxeS (12oo.i.fS 33) 13 i.f4 33 12 i..gS 'iVaS 9

12oo.'iVb6 130-0 (13 others 10) 13oo.tiJxeS (13 ... others 12) 14 tiJxeS 14 14 d6 21

12oo.f628 10 tiJd2 39

9 ... i.g4 10 0-0 (10 others 61) 10oo.tiJbd7 (lO ... others 57) 11 ~e1 (11 h3 62; 11 others 63) 11...~e8 (l1...tiJe8 60) 12 h3 i.xf3 13 i.xf3 'iYaS 48

13 ... ~b8 49 13 ... a6 14 a4 c4 (14 ... others 49) 54 13 ... ~c8 52 13 ... c452

9 ... bS (9 ... others 69) 10 eS 10 .. .t2Jfd7 71 lO ... dxeS 73

King's Indian move order (1 d4 tiJf6 2 c4 g6 3 tiJc3 i.g7 4 e4 d6 5 f4) S ... O-O 6 tiJf3 tiJa6

7 i.d3 80; 7 eS 86; 7 cS 86; 7 i.e3 86; 7 i.e2 86 S ... O-O (S ... others 106) 6 tiJf3 (6 i..d3 107; 6 i.e2 107) 6 ... cS (6 ... i.g4 107) 7 dS (7 dxcS 111; 7 i..e2 113) 7oo.bS (7 ... a6 113; 7 ... e6 8 i.e2 ~e8 113) 8 cxbS (8 others 95) 95

Modern Benoni move order (1 d4 tiJf6 2 c4 c5 3 d5 e6 4 tiJc3 exd5 5 cxd5 d6 6 e4 g6 7 f4) 7 ... i..g7 (7 ... others 120) 8 i.b5+ (8 eS 135) 8 ... i..d7 120

R .. ~hP 1?(): R U\frl7 Tn

Page 147: Beating King's Indian and Benoni
Page 148: Beating King's Indian and Benoni