BATSFORD CHESS OPENING GUIDES

A comprehensive guide to a highly popular opening

The King's Gambit is the most daring and dangerous opening. White throws caution to
the wind, and Black must know what he is doing to avoid early defeat. The King's
Gambit was all the rage in the 19th century, but has an enduring popularity throughout
the chess world. Remarkably, it is once more the focus of top grandmaster attention,
for example when Nigel Short played it three times in a row against the world's best at
the Madrid 1997 tournament.

• A classic opening which has come right back into vogue thanks to the efforts of
(amongst others) top English players Nigel Short and Michael Adams

• A dangerous weapon at all levels of chess

Part of the Batsford Chess Opening Guides series, which provide a rapid
understanding of fashionable openings through the use of model
games and clear explanations

Neil McDonald is a young English Grandmaster who has a string of

International tournament victories to his name. He is the author of
the successful Batsford books Winning with the Kalashnikov and
(with Andrew Harley) Mastering the French.

OTHER BATSFORD CHESS OPENING GUIDES:

THE BUDAPEST GAMBIT THE QUEEN’S GAMBIT ACCEPTED

Bogdan Lalic Chris Ward
Up-to-date coverage of a A lively explanation of an important
dangerous gambit. opening.
0 7134 8456 X 0 7134 8467 5

THE COMPLETE NAJDORF: THE SPANISH EXCHANGE

MODERN LINES Andrew Kinsman
John Nunn & Joe Gallagher An instructive survey of a Bobby Fischer
Definitive coverage of Fischer and favourite.
Kasparov's favourite opening. 0 7134 8471 3
0 7134 8218 4

For further information about Batsford £14.99 / US $21.95/ Can $32.95

chess books, please write to:

B T Batsford ISBN 0-7134-8451-9

The Chrysalis Building
Bramley Road
London W10 6SP
 Batsford Chess Opening Guides

The King's Gambit

Neil McDonald

B.T.Batsford Ltd, London

First published 1998
Reprinted 2000, 2001, 2003

Copywright © 1998 Neil McDonald

ISBN 0 7134 8451 9

British Library Cataloguing-in-Publication Data.

A catalogue record for this book is
available from the British Library.

AU rights reserved. No part of this book may be

reproduced, by any means, without prior permission
of the pubhsher.

Typeset by John Nunn and Petra Fink

and printed in Great Britain by
Creative Print and Design (Wales), Ebbw Vale
for the publishers
B T Batsford Ltd
The Chrysahs Building
Bramley Road
London WIO 6SP

lit ot Chrysalis Books Gk pic

Distributed in the United States and Canada by Sterling Publishing Co., 387 Park
Avenue South, New York, NY10016, USA

Chess set used in cover photograph courtesy of the

London Chess Centre

A BATSFORD CHESS BOOK

 CONTENTS

Bibliography 8
Introduction 9

Part One. King’s Gambit Accepted (2...exf4)

1 Fischer Defence (3 ^f3 d6) 12
2 Kieseritzky Gambit (3 ^f3 g5 4 h4 g4 5 ^e5) 28
3 Other Gambits after 3 ^f3 g5 and 3...<5^c6 50
4 Cimningham Defence (3 ^f3 ^eT) 76
5 Modem Defence (3 ^f3 d5) 88
6 Bishop and Mason Gambits (3 .^c4 and 3 ^c3) 98

Part Two. King’s Gambit Deciined

7 Nimzowitsch Counter-Gambit (2...d5 3 exdS c6) 115
8 Falkbeer Counter-Gambit (2...d5 3 exdS e4) 127
9 Classical Variation (2...±c5) 134

Part Three. Odds and Ends

10 Second and Third Move Alternatives for Black 147

Index of Games 159

 BIBLIOGRAPHY |

Books
Winning with the King's Gambit, Joe Gallagher ^atsford 1992)
Encyclopaedia of Chess Openings (ECO), 2nd edition (Chess Informator 1981)
King’s Gambit, Korchnoi and Zak (Batsford 1986)
Play the King’s Gambit, Estrin and Glaskov (Pergamon 1982)
The Romantic King’s Gambit in Games and Analysis, Santasiere and Smith
(Chess Digest 1992)
Das angenommene Konigsgambit, Bangiev (Schach-Profi-Verlag 1996)
Developments in the King’s Gambit 1980-88, Bangiev (Quadrant Marketing Ltd,
London 1989)
Modem Chess Openings Encyclopaedia, edited by Kalinichenko (Andreyevski
Flag, Moscow 1994)
The Gambit, M.Yudovich (Planeta Publishers, Moscow 1989)

Periodicals
Informator
New In Chess Yearbook
British Chess Magazine (BCM)
Chess Monthly

Magazine Articles
The King’s Bishop Gambit, Stephen Berry, Chess Monthly, November/
December 1981
Play the King’s Bishop Gambit!, Tim Wall, British Chess Magazine, May 1997

Video
An Aggressive Repertoire for White in the King’s Gambit, Andrew Martin,
Grandmaster Video 1995
 imOVUCTION

fiifKi ■Iri
Hasil^Aga

In the 19th century the art of defence does in other openings. As often as
was little understood. Hence, enter¬ not, his strategy consists of stifling
prising but unsound gambits often Black’s activity and then winning in
enjoyed great success. In those halcyon an endgame thanks to his superior
days for the King’s Gambit, boldness pawn struaure. Here is an example of
and attacking flair were more impor¬ this in action.
tant than rigorous analytical exacti¬
tude. The King’s Gambit proved the
perfect weapon for the romantic
player: White would push aside the
black e-pawn with 2 f4! and then over¬
run the centre, aiming to launch a
rapid attack and slay the black pieces
in their beds.
Nowadays, after a century of im¬
provements in technique and the ac¬
cumulation of theory by trial and er¬
ror, things are somewhat different.
Black players have learnt how to de¬ This position is taken from the
fend and any impetuous lunge by the game Illescas-Nunn, which is given in
white pieces will be beaten off with the notes to Game 45 in Chapter 7.
terrible losses to the attacker. White has the better pawn structure
Even in the King’s Gambit, there¬ (four against two on the queenside)
fore, White is no longer trying to at¬ and any en(%ame should be very good
tack at all costs. He has had to adapt for him. On the other hand. Black has
his approach and look for moves with dynamic middlegame chances, as all
a solid positional foundation, just as he his pieces are very active. White found
The King’s Gambit

a way to force an endgame here with ity and White’s better structure is cen¬
13 'Bfel! SeS 14 'lifh4! 'Iifxh4 (more or tral to the modern approach to the
less forced) 15 <5lxh4. There followed King’s Gambit.
15...^e3 16 .^xe3 Slxe3 17 Sael Sxel
18 Sxel and White’s queenside pawns
were much more valuable than Black’s
ineffectual clump on the kingside.
Furthermore, Black has not the slight¬
est covmterplay. It is no surprise that
White won after another 22 moves.
There was no brilliant sacrificial at¬
tack in this game, yet White succeeded
in defeating a top-class grandmaster.
Here is another example, taken from
Game 15 in Chapter 2.
This position was reached in Short-
Shirov, Madrid 1997, after White’s
ninth move (see Chapter 2, Game 8).
White has established the ideal pawn
centre, while Black has doubled f-
pawns. Therefore, statically speaking.
White is better. However, Shirov has
correctly jutted that his active pieces
are more important than White’s su¬
perior pawn struaure. Black has a lead
in development and can use this to
demolish the white centre. The game
Despite the fact that he is a pawn continued 9...'life7! 10 ^c3 Ad7 11
down. White’s chances would be no .^f3 0-0-0 12 a3?! ^xe4! and White’s
worse in an endgame. After all, he has proud centre was ruined, as 13 Axe4
control of the excellent f4-square and f5 regains the piece with advantage.
could aim to exploit the holes in the Shirov quickly followed up this posi¬
black kingside, which is looking dis¬ tional breakthrough with a decisive
jointed. However, as Tartakower re¬ attack. The time faaor was of crucial
marked ‘before the endgame the gods importance here: in the ‘arms race’ to
have placed the middlegame’. White is bring up the reserves White lagged too
behind in development and in the far behind.
game Black exploited this to launch an So what is Black’s best defence to
attack on the white king after 10 4ld2 the King’s Gambit? Three general ap¬
SeS 11 ^xe4 Slxe4+ 12 '^f2 c5! etc., proaches are possible:
when White was soon overwhelmed. a) take the pawn and hold on to it,
This conflict between Black’s activ¬ at least temporarily, with ...g7-g5.

10
 Introduction

b) play ...d7-d5 to counterattack. However, one should not forget the

c) decline the pawn in quiet fashion. Bishop’s Gambit 3 .^c4. Fischer fa¬
Of these options, the last one is the voured this move and at the time of
least promising. White shouldn’t be writing it has been successfully
allowed to carry out such a key strate¬ adopted by Short and Ivanchuk (see
gical advance as f2-f4 without encoun¬ Chapter 6). Furthermore, when I told
tering some form of resistance. Black David Bronstein I was writing a book
normally ends up in a slightly inferior, on the King’s Gambit, he replied ‘You
though solid, position. Nevertheless, want to play the King’s Gambit? Well,
undemonstrative responses remain Black can draw after 3 ^f3. Play 3
popular, mainly for practical reasons: .^c4 if you want to win!’ However, as
there is less theory to learn than in the a word of warning we should remem¬
main line. ber the words of a great World
Option b) is under a cloud at the Champion who grew up in the glori¬
moment. Although defences based on ous age of the King’s Gambit: ‘By
...d7-d5 allow Black free and rapid de¬ what right does White, in an abso¬
velopment of his pieces, often his infe¬ lutely even position, such as after
rior pawn structure comes to haunt move one, when both sides have ad¬
him later in the game. vanced 1 e4, sacrifice a pawn, whose
That leaves option a), 2...exf4. This recapture is quite uncertain, and open
is undoubtedly the most challenging up his kingside to attack? And then
move after which play becomes highly follow up this policy by leaving the
complex. As will be seen in Chapters 1 check of the black queen open? None
and 2, White has no clear theoretical whatever!’ Emanuel Lasker, Common
route to an advant^e after 2...exf4 3 Sense In Chess, 1896. A hundred years
^f3 d6 or 3...g5, while the variations on, the jury is still out!
in Chapter 3 have a poor standing for
White. Black should therefore bravely Neil McDonald
snatch the f-pawn. February 1998

11
 CHAPTER ONE \

Fischer Defence (3 d6)

1 e4 e5 2 f4 exf4 3 ^f3 d6 The idea behind 3...d6 is simple. In

‘This loss (gainst Spassky at Mar essence. Black wants a Kieseritzky
Del Plata 1960) spurred me to look for Gambit (Chapter 2) without allowing
a “refutation” of the King’s Gambit ... White to play ^e5. If after 4 d4 g5
the right move is 3...d6!’ - Bobby White plays 5 .^c4. Black can enter
Fischer, My Sixty Memorable Games. the Hanstein Gambit with 5....^g7 (or
It is ironic that Fischer, who hardly the Philidor after a subsequent 6 h4
ever played l...e5 as Black and only h6). The Hanstein seems favourable
adopted the King’s Gambit in a hand¬ for Black since he has a very solid
ful of games (always with 3 .^c4), kingside pawn structure. It is better
should have discovered one of Black’s for White to strike at the black pawn
most effective defences. Or perhaps structure immediately with 5 h4!, as
we should say rediscovered, as 3...d6 he also does in the Kieseritzky. Al¬
was advocated by Stamma way back though after 5...g4 6 ^gl. White’s
in 1745, but subsequently ignored. knight has been forced to undevelop
This neglect is puzzling. Why wasn’t itself. Black has had to disrupt his
the strength of 3...d6 appreciated in kingside structure with ...g5-g4. The
the heyday of the King’s Gambit by strange looking position after 6 ^gl is
Anderssen, Morphy and others? We the subject of Games 1-4, while 6 <5ig5
can either conclude that even in the is seen in Game 5.
field of ‘romantic’ chess modern play¬ Instead of 4 d4. White can try 4
ers are way ahead of the old masters, .^c4, when Black responds 4...h6, hop¬
or point to the creativity of a genius ing for 5 d4 g5 etc., when he reaches
able to find new ideas in familiar set¬ the favourable Hanstein. However,
tings. After all, who would look for an White can try to cross Black’s plans
improvement on move three of any with either 5 d3 (Game 6) or 5 h4
opening? (Game 7).

12
 Fischer Defence (3 fhf3 d6)

many new ideas that Gallagher pio¬

Game 1 neered and then publicised in his
Short-Akopian Winning with the King’s Gambit. After
Madrid 1997 7...ke7 8 'td2 h6 (8...f6 9 ±h6! ^xh6
10 'llxh6 was good for White in Gal¬
1 e4 e5 2 f4 exf4 3 ^f3 d6 4 d4 g5 lagher-Bode, Bad Worishofen 1991) 9
5h4! ±xe7 fxg2 (Black has to interpose this
The best move. White imdermines move as 9...<5lxe7 10 gxf3 is bad for
the black pawn structure before Black him) 10 .^xg2 <5lxe7 11 <5lc3 ^g6 12
has the chance of solidifying it with White had good compensation
...h7-h6 and The resulting posi¬ for the pawn in Gall^her-Ziatdinov,
tion may or may not be good for Lenk 1991. We have the typical dis¬
White, but one thing is clear; if he de¬ jointed black kingside to contrast with
lays even a move, e.g. with 5 .^c4, White’s solid centre.
then Black will definitely have good b) 6...^f6. Instead of defending the
chances after 5....^g7 6 h4 h6 etc. (see f4-pawn. Black counterattacks against
Chapter 3, Games 19 and 20). the e4-pawn. After 7 .^xf4 ^xe4 8
5...g4 6 4ig1 .^d3 d5 (Black tried to make do with¬
The Allgaier-related 6 ^g5?! is ex¬ out pawn moves in Hebden-Borm,
amined in Game 5. Orange 1987, but was in deep trouble
after 8...'te7 9 ^e2 kg7 10 0-0 0-0 11
.^xe4! 'Iifxe4 12 ‘5^bc3 #c6 13 'lifd2 d5
[now he has to move a pawn to pre¬
vent 14 ±h6] 14 ^g3 etc. Another
way to bolster the knight is 8...f5, but
White had a good endgame after 9
^e2 kg7 10 ±xe4 fxe4 11 ±g5 ±f6 12
^bc3 .^xgS 13 hxgS 'ItxgS 14 <5lxe4
#e3 15 ^f6-H ^d8 16 #012 #xd2+ 17
^xd2 ^c6 18 Safi ^e7 19 Sxh7 etc.
in Hebden-Psakhis, Moscow 1986) 9
.^xe4 dxe4 10 ^c3 .^g7 11 ^ge2 0-0
6.. .1.h6 12 #d2 f5 13 0-0-0 ^c6 14 h5 a6, Yak-
If Black’s last move was forced, here ovich-Zuhovitsky, Rostov 1988, and
he is spoilt for choice. Alternatives now Bangiev thinks that White is bet¬
include 6...'llf6 (Game 3, which may ter after 15 h6.
transpose to the present game) and 7 ^c3 c6
6.. .f5 (Game 4). Two other moves Here three other moves are possi¬
should also be mentioned: ble;
a) 6...f3. This was popular once, but a) 7...^f6 aims to start an immedi¬
perhaps Black has been frightened off ate attack on White’s centre after 8
by the move 7 .^gS! This is one of the ^ge2 d5!? Then the game Christoffel-

13
The King's Gambit

Morgado, Correspondence 1995, con¬ #d3 a6! 9 .^d2 ^c6, which looked
tinued 9 e5?! ^h5 10 g3 ^c6 11 ±g2 good for Black in the game, Gallagher
^e7 12 ^xf4 ^xf4 13 ^xf4 ^xf4 14 suggests 8 4ige2, when 8...#f6 9 g3
gxf4 c6 15 'lle2 h5 and Black had a hcg3 10 ^g3 Axel 11 Sxcl #f4 is
small advantage in view of his control not too different from the position
of the important f5-square. Gallagher reached in Games 1 and 2.
suggests that White’s play can be im¬
proved with the more dynamic 9
^xf4!? ^xf4 10 ^xf4 dxe4 11 ^c4!,
looking for an attack down the weak¬
ened f-file. After ll...^c6! (Black must
attack d4, not just to win a pawn but
also to exchange queens) 12 0-0 'lifxd4+
13 'lifxd4 ‘5ixd4 14 <5ifd5 4ixd5 15
^xd5 ^e6 16 ^f6+ <^e7 17 Sael Gal¬
lagher concludes that White has more
than enough for his pawns. Indeed, he
should regain them both over the next
couple of moves whilst retaining a 8 ^ge2 #f6 9 g3 fxg3 10 ^g3 Axel
positional advant^e. 11 axel #h6?
b) 7...^c6 is Black’s second option. After this White achieves easy de¬
Now 8 .^b5 a6 9 .^xc6+ bxc6 10 'lld3 velopment. The correct ll...#f4,
#16 11 .^d2 <5ie7 12 0-0-0 was tmclear which prevents White’s smooth build¬
in Bangiev-Pashaian, Correspondence up by attacking the knight on g3, is
1987. The critical move is 8 ‘5ige2, examined in the next game.
which leads to the sharp variation 8... 12 Ad3! #63-1- 13 ^ee2 ^e7 14
f3 9 4if4 f2+! 10 '^xf2 g3-i- 11 '^xg3 #d2!
‘5if6. Black has sacrificed his f- and g- This game demonstrates that the
pawns to expose the white king in King’s Gambit often offers White
similar fashion to the 5...d6 variation good enc^ame chances, even when he
of the Kieseritzky (see Chapter 2, is a pawn down.
Games 8-10). This position has been 14...#xd2-i- 15*xd2 d5?
analysed extensively by Gall^her, It is never a good idea to open the
whose main line runs 12 .^e2 Slg8-i- 13 centre when you are underdeveloped.
'^f2 ^g4+ 14 .^xg4 .^xg4 15 #d3 White now regains his pawn while
±g7 16 ±e3 #d7 17 ^cd5 0-0-0 18 b4 maintaining his positional advant^es.
Slde8 19 b5 ^d8 20 c4 4ie6, and now It was better to dig in with 15...Ae6,
21 c5 dxc5 22 dxc5 .^xal 23 Sxal e.g. 16 c4 4ia6 or 16...c5.
^xf4 24 .^xf4 gives White compensa¬ 16ace1 Ae6
tion for the exchange. If 16...dxe4 17 ^xe4 the threat of 18
c) 7....^e6 was tried in Gallagher- ^d6+ is very disruptive.
Hiibner, Biel 1991. Now instead of 8 17^14 0-0

14
 Fischer Defence (3 ^f3 d6)

Giving back the pawn, as 17...6xe4 Sf8 29 Sel g3 30 i.f5 ^b6 31 b3

18 Sxe4 leads to disaster on the e-file. <53e8 32 <53xb7 <53g7 33 i.h3 Sf4 34
18 exd5 <£ixd5 19 ^xe6 fxe6 20 ^xc5 Sxh4 35 i.g2 Sh2 36 Se2 ^f5
Sxe6 37 i.e4 ^d6 38 i.f3 Sh6 39 ^e6
White regains his pawn with excel¬ Sf6 40 i.g2 ^d7 41 c5 ^f7 42 d6
lent chances. He has more space in the <53fe5 43 i.d5 Sf5 44 c6 <53b6 45
centre, a lead in development and the Ag2 Sf2 46 Sxf2 gxf2 47 <S?e2 1-0
opportunity to attack the sickly black
g-pawn, which, although passed, is Game 2
well blockaded and difficult to sup¬ Fedorov-Pinter
port. PuU 1997
20.. .<ad7 21
It was even better to play 21 J.f5 I e4 e5 2 f4 exf4 3 <53f3 d6 4 d4 g5
according to Short, when after 5 h4 g4 6 <53g1 i.h6 7 <53c3 c6 8
21.. .^7f6 22 c4 4ib6 23 'i>d3 White is <53ge2 #16 9 g3 fxg3 10 <53xg3 i.xc1
in complete control. II Sxcl «f4!
21.. .<^h8 22 Sfl &ae8 23 Sxe8
Sxe8 24 c4 <53516 25 <53g3 c5

An attempt to disrupt the build up

of White’s position. The attack on the
A typical King’s Gambit situation knight means that White has no time
has arisen. The black kingside pawns for .^d3 as played in the game above.
are inert, while the white centre is 12^ce2«e3 13 c4?!
mobile and strong. Therefore Ak¬ White finds an ingenious way to
opian concedes a proteaed passed expel the queen. Nevertheless, the
pawn, hoping to entice the knight endgame with 13 ’B^d2 ’B^xd2-i- 14
from the excellent blockade square on 'i>xd2 seems a better approach.
g3 and so activate the g-pawn. The 13...<53e7 14 Sc3 «h6 15 i.g2
dternative was to wait passively while White could still have played for an
White increased his space advantage endgame with 15 ’B^d2. However,
with b2-b4 etc. \5..Mx62+ 16 ^xd2 c5! 17 l.g2 ^bc6
26 d5 ^g7 27 <53f5-(- ^h8 28 <53d6 looks better for Black. Why is this

15
The King’s Gambit

endgame worse for White than in dxc5 18 Sd3 ^bc6 19 Sd6 WahM} Of
Short-Akopian above? The point is course, the position remains very
that White has played c2-c4 here, complicated and there could be a
which means that Black’s counterblow knockout blow concealed among the
...c6-c5! cannot be met with c2-c3, thickets of variations.
maintaining control of the central 17Sf6
dark squares. The white centre is thus Now, in view of the threat h4-h5.
spht after the inevitable d4xc5 and the White wins the important d6-pawn,
e5-square becomes a strong outpost for after which he can always claim posi¬
a black knight. White is correa to tional compensation for the pawn
seek a middlegame attack in the game. deficit.
17.. .1'xh4 18Sxd6 c5
Too late!
19 ^f5 WgS 20 SdS
After 20 Sg31?, 20...h5 looks okay
for Black, but not 20...4ic6 21 nxg4l
nor 20...1.xf5? 21 exf5 «xf5 22 l.xb7
^d7 23 Sd51 lfe6 24 l.xa8 axa8 25
4if4ll 4kf4? 26 ’B^xg4+ and White will
be the exchange up in the endgame.
20.. .cxd4 21 Sg3 WfG 22 Sxg4
23 ^f4 i.xd5 24 ^xdS WeS 25 Sg5
^h8 26 Sh5 ^d7
15.. .0-0 There was a draw by repetition af¬
Here 15...c5 is the natural positional ter 26...Se8 27 ^fe7 %7 28 ^f5 Mtb.
move, imdermining White’s centre. 27«f3
But the crucial question is: can White A last winning try. White could
overwhelm his opponent before he have forced a draw with 27 Hxh7+
can develop his pieces? It seems that ^7 28 ’th5+ *g8 29 4ih6+ ^h7 30
the answer is yes after 16 Sd3! 4ibc6 ^f5+ ^g8 31 ^6+.
17 dxc5 dxc5 18 Sd6. For example, 27.. .5fe8 28 ^h6 l'g7 29 ^xf7+
18.. .1.e6 (18...1fe3 19 4ifl! wins the *g8 30 ^h6+ *h8 31 ^f7+ 'A-'A
queen, while 18...’B^g7 19 4ih5 ’B^xb2 White has to force the draw in view
20 4if6+ 'i’f8 21 0-0 gives White a big of the material situation.
attack) 19 ^f5! 1^16 20 Sxe6! fxe6 21
^d6+ ^d7 22 e5! %6 23 ^f4 WgS 24 Game 3
^xb7+ ^c8 25 ^xc5 ’td8 26 m Gallagher-G.Flear
with a very strong attack. Lenk 1992
16 0-0<ag6?
Here 16...c5! was the most challeng¬ 1 e4 e5 2 f4 exf4 3 ^f3 d6 4 d4 g5
ing move. As far as I can see Black 5 h4 g4 6 ^g1 WfO 7 ^c3 ^e7
then has good chances, e.g. 17 dxc5 After 7...c6 8 4ige2 J.h6 play will

16
 Fischer Defence (3 ^f3 d6)

transpose to the two games above. This idea received a practical test in
Gallagher points out that the attempt the game Russell-Beaton, Scotland
to refute 7...c6 with 8 e5 falters after 1994 (through a different move order
8...clxe5 9 4ie4 We? 10 dxeS ’i^xeS 11 beginning 8 ’i^dS!?). Unfortunately,
"Wei J.e7 12 J.d2 4if6! Meanwhile, White blundered immediately with 12
Bangiev recommends 7...c6 8 4ige2 4id5?, when he had nothing for his
4ih6, but this is either a brainstorm or pawn after 12...4kd5 13 exd5 ^e7 14
a misprint. 4ic3 0-0-0 etc. The key variation is the
calm 12 'i’bl 0-0-0 13 J.cl, when John
Shaw gives 13...f3 as unclear, while
13.. .5.e8 14 g3 f3 15 4if4 is Bangiev’s
choice. But doesn’t Black have an ex¬
cellent position after, say, 15...Wh8
and 16...f5 here?
9.. .<abc6 10 <5365!
The only way to exploit the queen’s
absence from d8 is to attack c7. After
10 g3 J.g7! 11 d5 fxg3! 12 ^g3
(White cannot allow \2..M{2+ and
13.. .g2) 12...^d4 13 l.g2 ^f3-(- 14
8^ge2i.h6 9«d2!? J.xf3 ’i^xf3 15 4ice2 .^e5 Black was
Note this idea only works after winning in Bangiev-Figer, Correspon¬
...^e7. If you put the knight back on dence 1987.
g8 and play...c7-c6 instead, then 9 10.. .<^d8 11 d5
Wd2.^? loses a piece after 9...f3. This looks horribly anti-positional,
Gall^er actually prefers 9 ’i^dS as it gives up the e5-square to the black
here. Play could go 9...a6 (to play knight. Bangiev reconimends 11 e5!,
...4ibc6 without allowing 4ib5) 10 which leads to a highly contentious
J.d2 4ibc6 11 0-0-0 J.d7 when a criti¬ position after ll...''B^f5 12 exd6 ^d5 13
cal position is reached: dxc7+ ^d7.

17
The King’s Gambit

The Russian Master claims that Black’s king faces an attack from all
White is better in the complications. White’s pieces) 19 Shi l.xd2-(- 20
However, according to Gallagher Wx62 Wgb 21 Sh6! Sd8 22 Sxg6 hxg6
‘Bangiev didn’t suggest a way to beat 23 ^c7 l.e6 24 <53xa8 Sxa8 25 l^xdb-i-
off the black attack. I can’t see any¬ ^e8 26 l.b5 Sc8 27 ’txe5 1-0 Gal-
thing resembling a White advant^e.’ l^her-Fontaine, Bern Open 1994.
Who is right? In a book published af¬
ter Gallagher’s comments, Bangiev
comes up with the goods: 14 4ig3!?
Somewhat surprisingly, this seems
good for White! For example:
a) 14...1fe6-(- 15 l.e2 Se8 (15...^e3
16 d5l l^xdS 17 ’txd5+ ^xd5 18
i.xg4+) 16 0-0 fxg3 17 lfxh6 Wxh6 18
.^xg4-(- Web 19 Sxf7-(- Se7 20 J.xe6-(-
^xeb 21 Sxe7-(- 4idxe7 22 J.g5 'i’d7
23 Sel when White has three pawns
for the piece and a dangerous initiative
since the black queenside is buried. This seems very convincing, but
b) 14...Se8-(-15 ^dl Web (15...^e3-(- 14.. .a6!? would have been a much
16 Wxe3!) 16 l.d3 ^e3+ 17 Wxe2>\ fxe3 tougher defence. Then Black would
18 l.f5 e2+ 19 ^el l.xcl 20 Sxcl a6 win after 15 ^xd6 l^xdb 16 0-0-0 l.d7
21 J.xe6-(- fxe6 22 ^c3 4ixd4 23 17 J.xf4 J.xf4-(- 18 ^xf4 WhA etc., so
^cxe2 4ixe2 24 ^xe2 ^xc7. Here the White has to try 15 4ibd4. With the
weak black pawns on e6 and g4 give knight chased from b5, 15...f3! is now
White a positional advantage (analysis safe, e.g. 16 J.xh6 (after 16 0-0-0 fxe2
by Bangiev). 17 Axe2 J.d7 White has little to show
Judging from this, 11 e5 seems to be for his piece ) lb...{2+ 17 'i’dl ’B^xh6
a much better try than 11 d5. and Black is much better.
11.. .<ae5 12 <53x14 12.. .a6 13^d4g3!
In a later game Gall^her improved White has regained his pawn but is
with 12 Wei c6 (forced) 13 dxc6 in serious trouble due to the pin on f4.
^7xc6 14 l.d2. Flear’s excellent move prevents White
from supporting the pinned knight
see following diagram_
with g2-g3.
Black now tried 14...f3 and was 14 ^de2 Sg8 15 #d4 i.g4 16 i.e3
soon overwhelmed: 15 0-0-0! fxe2 16 i.xe2 17 ^xe2 ^f3+\ 18 gxf3 «xf3
±xe2 ^e7 (if 16...a6 17 Shfl Web 18 19 i.xh6 Wxhl 20 i.g5 g2 21 ^2
^xd6 and White has an enduring at¬ SxgS
tack for his piece; maybe 16...J.d7 is Instead of giving back the exchange,
best) 17 Shfl Wgb 18 h5! Wxh5 (if the computer program Fritz prefers to
15.. Web 19 ^c7 Wxa2 20 i.xh6 and win another one with 21...h6. Now a

18
 Fischer Defence (3 ^f3 d6)

bishop retreat from g5 allows 22...glW two squares. This hardly accords with
with a winning attack, so 22 Axg2 (22 the precepts of classical chess, which
±{6 Wh2\ 23 ^e3 Sg3+ 24 <^d2 gl^t require rapid and harmonious devel¬
wins) 22...1fxal 23 M6 HUxal 24 e5 opment of the pieces.
Wa5 and the white attack will fail, 7-5^c3
with huge losses. Here 7 .^xf4 fxe4 8 ‘?3c3 ^f6 trans¬
22 hxg5 gxf1«+ 23 Sxfl «h4+ 24 poses to the main game.
^g3 ^d7 25 «f6 Sg8 26 Shi «xg5 An important tactical point is the
27 «xg5 Bxg5 28 Bxh7 ^e8 fact that 7 exf5? fails to 7..Me7+. For
White seems to be a little better example, 8 l.e2 i.xf5 9 ^c3 (if 9 i.xf4
here after 29 4ih5. We4!) 9...i.h6 10 ^d5 We4 11 ^xc7+
'i’d7 and Black wins (Raetsky). Or if 8
Game 4 ’B^e2 J.xf5 9 J.xf4 Axc2! and White
Hector-Leko has hardly any compensation for the
Copenhagen 1995 pawn. It is a pity that 8 ‘?3e2 doesn’t
seem to work after 8...f3, e.g. 9 .^g5
1 e4 e5 2 f4 exf4 3 ^f3 d6 4 d4 g5 fxe2 10 kxe2 ^ 11 0-0 i.g7 12
5 h4 g4 6 -^gl f5 J.b5-(- ‘^d8! or 9 gxf3 gxf3 10 flh3 fxe2
11 J.xe2 J.h6!? In neither case does
White have enough play for a piece.
7...-ate 8 i.xf4
The critical move. In Shevchenko-
Raetsky, Russia 1992, White played
the careless 8 We.2} and after 8....^h6 9
exf5-(- 'i’f7! Black suddenly had an
overwhelming lead in development.
White was swept away in impressive
style: 10 W{2 Se8+ 11 <^dl g3 12 Wf3
i.xf5 13 ±c4+ ^g7 14 ^ge2 i.g4 15
Wxh7 d5! 16 i.d3 (if 16 WxaS dxc4 17
An imaginative idea. White hasn’t ‘WxaJ ^c6 followed by 18...^xd4
yet got any pieces in play, so Black crashes through) 16...‘?3e4! (completing
feels that he has time to strike at his the strategy began with 6...f5; Black
opponent’s centre and dispose of the has absolute control of e4) 17 WxaS (if
strong e-pawn. It looks risky to re¬ 17 l.xe4 dxe4 18 WxuS 'Wxd4+ 19 i.d2
move the remaining pawn cover from f3! - Raetsky) I7...^i2+ 18 <^el
Black’s king, but hasn’t White done ^xhll9 Wxd5 lfxh4 20 i.c4 <^h8 21
the same thing with 2 f4? Further¬ m7 l.h5 22 mxc7 ^f2 23 <^fl Whl+
more, White’s play is hardly above 24 ^gl 4ig4 0-1, as 25 ^ce2 ^h2-(- 26
criticism. In the first six moves he has 'i’el ’B^xgl-i- is more than flesh can
developed and then undeveloped his stand. White played the whole game
knight, and moved his rook’s pawn without his queen’s rook or bishop.

19
The King’s Gambit

8.. .fxe4 9l'd2 hardly inspiring. White doesn’t play

White has also tried 9 Wei d5 10 the l^g’s Gambit to ^ree a draw
J.e5, when Bangiev recommends after 11 or 12 moves!
10.. .c6! 11 4idl ^bd7 12 ^e3 4ixe5 13 10.. .C6 11 ^ge2 i.e6 12 <5314
dxe5 ^d7 14 lfxg4 ^35+ 15 c3 ^xe5 If 12 %5 ^bd7 13 ^f4 We7l 14
16 ’i^h5+ 4if7 as clearly good for ^h5 ^xh5! 15 l.xh8 ^g3 16 Sgl
Black. ’i^xg5 17 hxg5 J.e7 gives Black excel¬
At the time of writing, theory has lent play for the exchange - Leko.
yet to decide on the strongest response 12.. .1.f7 13 <53d1 <53bd7 14 <53e3
to 7...f5. Nevertheless, I would sug¬ <53xe5 15dxe5«c7!
gest that 9 d5 ought to be considered. I This simple move refutes White’s
like the idea of preventing Black con¬ attack by pinning the e-pawn and pre¬
solidating his centre with 9...d5. In his paring ...0-0-0. Since the e-pawn is fa¬
annotations to the Heaor game, Leko tally weak. White will soon be two
gives 9 d5 a question mark, claiming pawns down without any real com¬
that Black is a little better after 9...1.g7 pensation.
10 h5 0-0 11 h6 MS 12 M l^eS. 16 #03 0-0-0 17 0-0-0 ^h5 18 <53e2
However, instead of pushing the h- i.h6 19 ^b1 i.xe3 20 #xe3 ^b8 21
pawn White can mobilise his pieces, #g5 Shg8 22 WfS i.g6 23 #g5
e.g. 9...1.g7 10 Wd2 0-0 11 ^ge2, Sde8 24 «xg4 «xe5 25 «g5 «xg5
planning moves like 0-0-0, ^d4 and 26 hxg5 Se5 27 g4 ^g7 28 ^f4
l.c4. <53e6 0-1

Garnet
Morozevich-Kasparov
Paris (rapidplay) 1995

1 e4 e5 2 f4 exf4 3 <53f3 d6 4 d4 g5
5 h4 g4 6 ^g5

9.. .d5 10i.e5?!

White’s position begins to fall apart
after this. According to Leko, White
should have played 10 ^b5 4ia6 11
^c3 c6 12 J.xa6 bxa6 13 ^ge2 with
unclear play. However, since Black
can force a draw by repetition with
11.. .41b8, this recommendation is White plays in enterprising style.

20
 Fischer Defence (3 d6)

hoping to bamboozle the World has already played ...d7-d6, so he

Champion with a rarely seen sacrifi¬ would be a tempo down if he were to
cial line. Since this was a rapidplay revert to ...d6-d5 after J.c4-(-.
game, such an approach makes some It is also worth comparing the sacri¬
sense. fice here with the line 1 e4 e5 2 f4 exf4
6...h6 3 4if3 g5 4 h4 g4 5 4ie5 d6 6 4ixf7?, as
An interesting moment. According played in Schlechter-Maroczy, Vienna
to Fischer it is better to play 6...f6!, 1903. (This is real coffee-house chess. I
when 7 ^h3 gxh3 8 Wh5+ ^d7 9 have a book on Schlechter that is full
.^xf4 ’B^eS! 10 ^8^13 'i’dS leaves White of fine positional games. Yet in those
with little for the piece. Another pos¬ days nobody was immune from the
sibility given by ECO is 7 J.xf4 fxg5 8 outlandish sacrifices which seem ri¬
l.xg5 l.e7 9 M l.e6! 10 ^c3 ^d7 diculous to modem eyes.) After
and again Black should be able to de¬ 6.. .'i>xf7 7 i.c4-(- 'i>e8 Black was
fend successfully. This opinion is sup¬ clearly better. In the Kasparov game
ported by Gallagher. Why did Kas¬ we have reached a similar position
parov avoid 6...f6 then? Perhaps he with the moves d2-d4 and ...h7-h6
was afraid of an improvement or per¬ thrown in. This should help White.
haps he had simply forgotten the Or does it? The move ...h7-h6 pre¬
theoretical refutation. vents .^g5 in some lines and, as we
7 <axf7 <S?xf7 8 i.xf4 i.g7 9 i.c4+ shall see, h7 proves a good square for
<^e8 the black rook...
10 0-0??
Very stereotyped. The white king
will prove to be a target on the king-
side. It was better to play 10 ^c3, in¬
tending 11 ’B^d2, 12 0-0-0 etc. (if ...k)c6
then J.e3) with an enduring initiative
which would have offered fair chances
in a rapid game. If this plan fails then
the whole variation is simply bad for
White.
10.. .^c6 11 i.e3
White might as well play 11 c3, as
White now has a favourable version the coming incursion on the f-file leads
of the Allgaier Gambit, since normally nowhere.
after 1 e4 e5 2 f4 exf4 3 4if3 g5 4 h4 g4 11.. .«xh4!
5 ^g5 h6 6 4ixf7 ^x{7 7 Ac4+ Black A good defensive move, vacating d8
responds 7...d5! (or if 7 d4, then 7...f3! for the king, and a strong attacking
8 Ac4-(- d5). The point is that Black move, threatening 12...g3.
usually gives up the d-pawn to speed 12Sf7 Sh7!
up his development. In the game Black Another dual-purpose move. Black

21
The King’s Gambit

defends the bishop and threatens sound bizarre but all is soon re¬
13...Axd4! 14 flxhZ? Axe3+ and mate vealed!). After the alternative 5 d4,
on f2. play could transpose to a Hanstein
13 e5 <aa5 with 5...g5 etc. (see Chapter 3, Game
This beats off the attack with fright¬ 20). Since the Hanstein looks suspect
ful losses. It is no wonder that the at¬ for White, this is another reason to
tack fails: not only has White sacri¬ consider 5 d3. However, the analysis
ficed a piece, but the queenside rook below also gives 5 d3 a thumbs down,
and knight may as well be any place so the conclusion seems to be that 4
but on the board. J.c4 is inaccurate: 4 d4 is the only de¬
14 i.d3 <S?xf7 15 Wfl-t- <S?e7 16 cent try.
i.xh7 Ae6 17 ^d2 Sf8 18 exd6+ 5.. .g5 6 g3 g4
cxd6 19 «e2 <S?d8 20 c3 ^7 21 Four other moves are possible:
Sel .^c4! a) 6...fxg3 7 hxg3 .^gZ looks dan¬
Of course capturing twice on c4 gerous for Black after the sacrifice 8
now leaves el en prise. The game ^xg5 hxg5 9 Sxh8 l.xh8 10 Wh5 ^{6
move allows a mercifully quick finish. 11 ^c3. However, White has still has
22 i.f2 Sxf2 23 «xf2 g3 0-1 to prove the win after, say, ll...'i’f81?
Mate on h2 or loss of the queen fol¬ 12 ^d5 We5l? or 12 J.xg5 Wg61
lows. b) 6...J.h3 was played in Gallagher-
Lane, Hastings 1990, when 7 4id4?!
Game 6 d5! 8 exd5 J.g7 led to obscure play.
Gallagher-Kuzmin Gallagher suggests that 7 ’B^dZ was bet¬
Biel 1995 ter, preventing ....^g2 and intending to
capture on f4 (the immediate 7 gxf4 is
1 e4 e5 2 f4 exf4 3 <^f3 d6 4 i.c4 less good, as 7...g4 8 4id4 ’B^h4-(- looks
h6 5 d3 annoying; whereas after 7 Wd2 Black’s
check on h4 can be answered by Wf2).
c) 6...^c6 7 gxf4 g4 (if 7...J.g4 Gal¬
lagher suggests 8 c3, hoping for
8.. .gxf4 9 J.xf4 4ie5? 10 .^xe5 and 11
J.xf7-(- winning) 8 ^gl ’B^h4-f 9 “^fl
f51 (much more dynamic than 9...4if6
10 ^g2 ^h5 11 ^c3 g3 12 Wei Sg8 13
h3 with advantage to White, as given
by Gallagher; note that 12...^xf4-(- 13
J.xf4 Wxf4 14 ^d5 is bad for Black, as
is 13...4ixf4-(- for the same reason)

see following diagram_

This is Gallagher’s pet idea. White
strengthens his centre and keeps d4 10 4ic3 ^f6 11 'i’g2 fxe4 12 dxe4
free for his king’s knight (this may .^dZ 13 h3? (following the plan out-

22
 Fischer Defence (3 d6)

lined in the last bracket, but here it is bishop on g7 rather than the knight
inappropriate; 13 J.e3, intending 14 on c6. Perhaps Black should try 9...f5,
Wei, looks better, when White may as 9...^f6 10 ^g2 ^h5 11 ^c3 g3 12
have the advantage) 13...^h5! 14 hxg4 Wei Sg8 13 h3 is good for White (if
Wg3+ 15 'i’fl 4ixf4 16 Wf3 (the queen 13...^xf4+ 14 l.xf4 Wxf4 15 ^d5
exchange saves White from immediate etc.).
collapse, but he has two weak pawns 7 <ad4 f3 8 c3
on e4 and g4 and a hole on e5, whereas Gallagher suggests the alternative
Black only has one weakness on h6; plan of 8 l.e3, ^c3, Wd2 and 0-0-0 in
nevertheless, he uses his slight lead in his book.
development to avoid the worst)
16.. .Wxf3+ 17 ^xf3 ^g6 18 g5
(liquidating one of his weak pawns)
18.. ..^g4 19 J.e2 0-0-0!? (on this move
or the last Black could have played
...h6xg5, but Beliavsky chooses a dy¬
namic pawn sacrifice) 20 gxh6 J.e7
with unclear play which eventually
led to a draw in Belotti-Beliavsky,
Reggio Emilia 1995/96.

8...^c6!
This is Kuzmin’s improvement.
Rather than prevent the white knight
going to d4 with 6....^g7 or 6...^c6,
Black attacks it when it reaches this
square. Black has tried two other
moves:
a) 8....^g7?! (actually the move or¬
der was 7...J.g7 8 c3 f3) 9 Wb3 Wd7
(forced because if 9...We7 10 ^f5!
d) 6...J.g7 7 c3? 4ic6! (ruling out 8 J.xf5 11 Wxb7 wins) 10 .^f4 ^c6 (too
«3d4) 8 ^a3 l.e6 9 Wb3 Wd7 10 gxf4 late) 11 ^f5 .^e5 12 ^d2 and White
Axc4 11 Wxc4 (if 11 ^xc4 d5l) 0-0-0 had good play for the pawn in Gal-
12 .^d2 (McDonald-Morris, Douai lagher-G.Flear, Paris 1990.
1992) and now Black should have b) 8...4id7!? is an improvement on
played 12...Wh3! 13 Sfl d51 with a big Flear’s 8....^g7, played by ... his wife!
advantage as 14 exd5? Se8-(- 15 g4 The knight heads for e5, which is a
wins for Black. Critical was 7 gxf4l g4 more efficient way of defending f7
8 ^gl Wh4-(- 9 'i’fl and we have op¬ from attack by Wb3 than 9...Wd7 in
tion c) above but with the black the previous variation. The game

23
The King’s Gambit

McDonald-C.Flear, Hastings 1995, 15...Sa6!

continued 9 4ia3 4ie5 10 J.f4 ^xc4 11 The last difficult move. On the
^xc4 4ie7! (and now the other knight other hand, 15...cxd5? 16 4ib5 would
heads for e5!) 12 Wz4+ ±d7 13 *53 have given White a dangerous attack.
^g6 14 l.e3 ^e5 15 0-0-0 Sb8? (too 16 «b3 cxd5 17 -^bS l'b4+
passive; 15...J.g7 is fine for Black) 16 The exchange of queens kills off
^xe5 dxe5 17 4if5! (now White has White’s initiative.
good chances) 17... J.xf5?? 18 'lft)5-(-! (a 18 «xb4 axb4 19 ^c7-h *d8 20
deadly zwischenzug) 18...’B^d7 19 ^xa6 bxa6
Wxe5+ l.e6 20 Wxh8 h5 21 i.c5 and The dust has cleared and Black has a
Black resigned. Despite the unfortu¬ decisive material advantage.
nate outcome, Black’s opening idea 21 e5 ^e7 22 Scl i.e6 23 h3 gxh3
seems good. 24 ^2 ^d7 25 ^xf3 ^c6 26 g4
9 ^a3?! <axd4+ 27 *e3 ^c6 28 d4 f6 29
Here 9 ^xc6 bxc6 would be posi¬ exf6 i.xf6 30 Sxh3 i.xg4 31 Sxh6
tional capitulation, so White should i.xd4+ 32 ^d3 Sxh6 33 i.xh6 i.xb2
try 9 ’B^a4, when Kuzmin analyses 34 Sfl i.e6 35 i.d2 a5 36 Shi
9.. .1.d7 10 *53 ^e5 11 ’txb7 ^xc4 i.f5+ 37 *e3 *d6 38 Sh6+ *c5 0-1
12 dxc4 .^g7 as slightly better for
Black. Game 7
9.. .^xd4 10cxd4 Ag7 C.Chandler-Howard
White’s centre is dislocated and will Correspondence 1977
inevitably crumble. Therefore, Gal¬
lagher goes for a do or die attack. 1 e4 e5 2 f4 exf4 3 ^f3 d6 4 i.c4
11 WbSWe? 12i.f4c6 h6
Not 12...1.xd4? 13 l.xf7+ 1^x17 14 An interesting alternative idea here
Wa4-(- J.d7 15 ’B^xd4 etc. is 4...J.e7!?, as played in McDonald-
13«b4 Skembris, Cannes 1993. After 5 0-0
Playing for traps as 13 d5 4if6 14 ^f6 6 d3 d5 7 exd5 ^xd5 8 l.xd5
dxc6 bxc6 is bad. ’B^xd5 9 J.xf4 0-0 White had a minus¬
13.. .a5! cule advantage. In effea. Black has
Kuzmin avoids the draw with played a Cunningham Defence but
13.. .d5? 14 l.d6 lfg5 (14...1fe6?! is avoided the normal problem after 1 e4
risky: 15 J.xd5 cxd5 16 ^b5 etc. ) 15 e5 2 f4 exf4 3 ^f3 J.e7 4 J.c4 4if6 of 5
l.f4 We7 etc. e5!, chasing his knight from the cen¬
MWbG tre. The drawback is that he is a
If 14 ’B^xd6 ’B^xd6 15 J.xd6 b5 16 tempo down on the line 5 d3 d5 6
J.b3 a4 17 J.c2 J.xd4 and wins exd5 4kd5 7 J.xd5 ’i^xd5 8 J.xf4.
(Kuzmin). However, 5 d3 is hardly an ultra-sharp
14.. .dS 15i.xd5 move, so it seems that Black can afford
The only chance, as 15 .^b3 Sa6 16 this liberty.
Wc7 Wxc7 17 J.xc7 J.xd4 is hopeless. 5 h4

24
 Fischer Defence (3 fhf3 d6)

Attacking the ghost of the pawn on The sacrifice 8 ^e5 doesn’t look
particularly brilliant after 8...dxe5 9
g5.
5...ate WxhS g6 and 10...^xd4. I also didn’t
Instead Black could go hunting for like the look of 8 ^e2 Wf6 or 8 ^d5
more pawns with 5...^e7 6 ?^c3 (a ^g3 9 Sgl g5 etc. Therefore, I tried
more solid approach is 6 d4 ^g4 7 the unusual looking move 8 d5!? ^e7
i.xf4 i.xh4+ 8 g3) 6...i.g4 7 d4 9 ^d4 ^g3 10 Sh2, when I was hap¬
^xh4+ 8 '^fl g5. However, according pily contemplating 11 .^xf4 next
to an article in Chess Monthly, January move, or if 10...^g6 then 11 h5 ^e5
1976, 9 l'd3 then gives White suffi¬ 12 .^b5-i- followed by .^xf4. However,
cient play, e.g. 9...^xf3 (more or less Flear found a brilliant move which
forcecC as 9...i.g3 10 ?3xg5 WxgS 11 shows up all the weaknesses created
l'xg3 fxg3 12 ^xg5 is best avoided) 10 by 8 d5: 10...g5!! 11 hxg5 ?3g6. Black
#xf3. White has chances in view of has returned the extra pawn to keep
his lead in development, his two bish¬ hold of f4. 10...g5 has also cleared the
ops and the awkward position of the diagonal al-h8 for the dark-squared

bishop on h4. bishop, which White has weakened

6 ‘Sic3 ^g4 with d4-d5. The e5-square is now

Another sharp possibility is 6...^e7 firmly in Black’s hands and is a central
7 d3 ^h5 8 ^e5 dxe5 9 #xh5 0-0 10 outpost for a black knight or bishop.
g3!?, planning to answer 10...fxg3 with The game continued 12 .^b5-i- .^d7 13
11 Axh6. However, the best move is ±xd7+ Wxd7 14 gxh6 i.xh6 15 ^f3
probably 6...^c6!, as played in and now the simple 15...0-0-0, plan¬
McDonald-G.Flear, Hastings 1992/93. ning 16...Sde8 etc., attacking e4, is
good for Black. The white king is a
see following diagram_ long way from the safety of the queen-

After 7 d4 ‘^ihS Black was ready to side. In the game Black tried the pre¬

complete his development with mature 15...f5, when 16 exf5 Wxf5 17

....^e7, ...^g4 and ...0-0, so White #d3! was unclear.

should do something fast. 7 d4 8 ‘5ie5!

25
The King's Gambit

Breaking the pin in some style. Of 14 c6 15 fldi 16 «g4

course 8..Axdl 9 i.xf7+ ^e7 10 ^d5 With his knights scattered and his
is mate. kingside undeveloped, Black is lost.
8.. .dxe5 9 Wxg4^f6 16...«xd1-i- 17*xd1 ^xal 18 ^d5!
The critical move is 9...^g3 10
i.xf4 ^xhl 11 i.xe5 (11 dxeS!? seems
better - in the game the pawns look
pretty on d4 and e4, but the e5-pawn
becomes a battering ram and the d-file
is opened; in fact it is difficult to see a
good answer to the plan of e5-e6 in
conjunction with Sdl) ll...#d7 12
#f3 ^c6 13 0-0-0 and Black eventually
won in Chandler-Haldane, Corre¬
spondence 1977.
10«f5
If 10...#xd4 11 Wc8+ (11 ^d5 ^d8) This elegant winning move is better
11.. .'^e7 12 ^d5-i-! ^xd5 13 .^xd5 and than 18 #xh5 0-0-0-i-. The knight on
Black seems lost in view of the attack h5 won’t run away, so White prevents
on b7. For example, 13...c6 (13...<^16 the black king from escaping to the
14 Wxc7 ±e7 15 #xb7) 14 Wxh7+ queenside.
?3d7 15 #xa8 cxd5 16 Wxd5 etc. 18...fid8 19 Wxh5.^c5
11 dxe5 ^4 12 «xf4 ^c2+ 13 After 19...g6, 20 #f3 cxb5 21 b3, in¬
^e2 Wd4 tending 22 .^b2 winning the knight, is
If 13...<§3xal 14 exf6 #xf6 15 ^d5! simplest.
will win material - Chandler. 20 «g4 *f8 21 «f5 Jie7 22 e6 1-0

26
 Fischer Defence (3 ^f3 d6)

Summary
Although it is difficult to agree with Fischer that 3...d6 refutes the King’s Gam¬
bit, it is certainly one of the best defences. At the time of writing, White can
only hope for an ‘vmclear’ verdict after best play in the main line, with 6...f5!?
(Game 4) looking particularly challenging. The divergences from the main line
with 4 .^c4 (Games 6 and 7) don’t seem very promising for White either.

1 e4 e5 2 f4 exf4 3 d6

4d4
4 i.c4 h6 (D)
5 di-Game 6
5 h4 - Game 7
4.. .g5 5 h4 g4 6 4ig1 (D)
6 ^g5 - Game 5
6.. ..^h6
6...#f6 - Games
6.J5-Game4
8 ^ge2 «f6 9 g3 fxg3 10 4ixg3 i.xc1 11 fixcl (D) «f4
n...Whe - Game 1
12 fhce2 - Game 2
 CHAPTER TWO \ mtmmmt]

Kieseritzky Gambit [£m£m M£t

(3 af3 g5 4 h4 g4 5 ^eS)

After on f3 is attacked and must move to

1 e4 e5 2 f4 exf4 3 g5 safety. This disruption in White’s po¬
the continuation sition should give Black enough active
4 h4! g4 5 ‘5ie5 play to compensate for his positional
the Kieseritzky Gambit, is White’s weaknesses.
strongest continuation, and is the sub¬ Black has several replies to 5 ?3e5,
ject of the present chapter. It is impor¬ the most popular of which at present
tant for White to undermine the pawn are 5...d6 (Games 8-11) and 5...^f6
on g5 before it can be reinforced with (Games 12-17). The first of these re¬
...h7-h6 and ....^gZ. If Black were turns the gambit pawn immediately in
given time - even a single move - to an attempt to seize the initiative, while
support his g5-pawn then he would the second forces the exchange of the
have a solid, well entrenched chain of black f-pawn for the white e-pawn,
pawns on the kingside. He could then unless White adopts the aggressive 6
ignore any later h2-h4 thrust since g5 .^c4 (see Games 12-14). Other Black
would be securely defended. In con¬ options at move five are considered in
trast, after the immediate 4 h4! Black the notes to Game 8.
has no time to set up a wall of pawns,
as 4...h6 loses to 5 hxg5. Therefore, he Game 8
has to advance his g-pawn again, Short-Shirov
which destroys any hopes of a com¬ Madrid 1997
pact pawn formation; his kingside is
permanently wrecked. On the other 1 e4 e5 2 f4 exf4 3 ‘5if3 g5 4 h4! g4
hand, things are also not so simple for 5 ^e5 d6
White. The move h2-h4 loosens the With this move Black sacrifices the
white kingside and the advance 4...g4 g4-pawn in order to achieve a smooth
is awkward to meet since the knight and active development of his pieces.

28
 Kieseritsky Gambit (3 ^f3 g5 4 h4 g4 5 ^e5j

.^c4, which will almost certainly

transpose to Game 12 below, where
the opening moves were 1 e4 e5 2 f4
exf4 3 ^f3 g5 4 h4 g4 5 ^e5 ^f6 6
.^c4 d5 7 exd5 .^g7 8 d4 ^h5. Notice
that in this sequence White could not
play ?3c3 instead of J.c4, e.g. at move
six, 6 ^c3? d6! forces 7 ^d3, which
looks silly with the d-pawn still on d2
rather than d4. Therefore, the move
order of Game 12 makes more sense
than 5....^g7 straightaway, as it rules
The main alternative, 5...^f6, is ex¬ out White’s strong 7 ?3c3 idea. For
amined in Games 12-17. analysis of the position after 7 .^c4,
Other possibilities for Black are: the reader is referred to Game 12.
a) 5...J.g7. The most important of b) 5...d5. A natural move, but inap¬
the less popular moves. Indeed, Zak propriate here. Black does nothing to
and Korchnoi go as far as giving it an challenge the knight on e5 or defend
exclamation mark. After 6 d4 (Keres the important f4-pawn. White can get
analyses 6 ^xg4 d5 7 d4! dxe4 8 .^xf4 a clear plus with some vigorous
Wxd4 9 #xd4 .^xd4 10 c3 .^xg4 11 moves: 6 d4! <53f6 7 .^xf4 ^xe4 8 <53d2!
cxd4 4ic6 12 .^bS 0-0-0 13 .^xc6 bxc6 ^xd2 9 'txd2 J.g7 (if 9...i.d6 White
14 0-0 f6 15 ?3c3 Sxd4 16 Sael, when castles queenside then attacks with
White regains one of his pawns and .^d3, etc.) 10 .^h6 and White had a
has good play, but he cannot hope to dangerous attack in Teschner-Dahl,
win after say 16...^e7 17 <53xe4 ^d5) Berlin 1946.
6.. .‘53f6 (6...d6 7 ^xg4 .^xg4 8 Wxg4 c) 5...h5. Every game I have seen af¬
.A.xd4 9 ?3c3! is good for White - Gal¬ ter this move has ended in disaster for
lagher) 7 ^c3! (this is the reason that Black, which is only to be expected.
5.. ..^g7 is out of favour) 7...d6 8 ?3d3 As Bronstein remarks, is there any
0-0 9 ^xf4! ^xe4 (or else Black has a other variation in which the first piece
rotten structure for nothing) 10 ^xe4 Black develops is his king’s rook?
Se8 11 <^12 Sxe4 12 c3 Wf6 (again this Bronstein himself took apart this
is do or die, as White intends the sim¬ variation in a famous game: 6 .^c4
ple 13 .^d3 with advantage) 13 g3 .^h6 Sh7 7 d4 M6 (7...d6 8 ^xf7!) 8 ?3c3
-14 .^d3 .^xf4 15 .^xf4 Sxf4-i- 16 gxf4 ^c6 9 ^xf7! Sxf7 10 i.xf7-i- ^xf7 11
Wxf4-f 17 '^e2! (some precise moves, .^xf4! .^xf4 12 0-0 #xh4 13 Sxf4-i-
discovered by Rubinstein, will beat off ^g7 14 #d2! d6 15 Safi ^d8 16 ^d5
the attack) 17... g3 18 #d2! .^g4-i- 19 .^d7 17 e5! dxe5 18 dxe5 .^c6 19 e6! (as
*el g2 20 #xg2 ^c6 21 Se8 22 often happens, a pawn advance is the
fifl! White wins. final straw for a beleaguered defence)
Instead of 7 ‘53c3! White can play 7 19..i.xd5 20 Sf7-i- ^xf7 21 Sxf7-i- ^h8

29
The King’s Gambit

22 Wc3+ 23 Sxf6 #xf6 24 #xf6+ elsewhere, he would have a winning

^h7 25 #f5+ ^h6 26 #xd5 ^g6 27 position. However, it is not easy to
#cl7 1-0 Bronstein-Dubinin, Lenin¬ begin a siege of f4 since White has his

grad 1947. own weaknesses to defend and, as we

d) 5...^c6 6 d4 mi? (6...<§3xe5 7 shall soon see. Black’s pieces will be
dxe5 d6 8 .^xf4 is known to be good developed very rapidly to aggressive
for White) was tried in the blindfold squares. Therefore, as stated above, 7
game Nunn-Piket, Monaco 1995. ^xf6-i- seems to be correct.
Now instead of 7 ^c3?! .^b4 8 ?3d3? 7.. .^c6
#xd4 when, after 63 moves, the game Black has to play energetically; oth¬
was won by ... White, 7 <53xc6 looks erwise the wea^ess of the doubled f-
good, e.g. 7...dxc6 8 e5 #f5 9 i.d3 pawns could lead to a lost position.
#d7 10 c3 c5 11 i.xf4 cxd4 12 cxd4 c5 8 d4 .^h6 9 ^e2
13 d5! or 7..Mxc6 8 i.d3 d5 9 0-0 The drawbacks of having the knight
dxe4(?) 10 i.xe4! on f2 rather than f3 are becoming ap¬
6 ‘5lxg4 parent. With the knight on f3, a good
Black continues to harass the white and natural developing move would
knight. The other possibility 6....^e7 be 9 .^d3, but here that simply loses
is the subject of Game 11. the d4-pawn. Therefore, White plans
7af2 to put the bishop on f3, where it forti¬
The alternative 7 ?3xf6-i-, which fies e4 and also defends g2 and the
seems to be a better continuation, is kingside in general j^ainst ...Sg8.
the subject of Games 9 and 10. Short, Once e4 is well defended and the king-
however, has no wish to see his oppo¬ side secure. White can turn his atten¬
nent’s queen activated after 7 ^xf6-i- tion to the f4-pawn.
#xf6 and so retreats his knight. At the 9.. .«e7 10 .^d7 11 .^f3 0-0-0
same time he defends e4. However,
White’s plan seems fatally flawed. The
knight has made three moves to end
up on a square that will prove both
aggressively and defensively to be
worse than f3. By retreating. White
also gives his opponent the free devel¬
oping move ...4if6, which means that
Black now has a lead in development.
This is a dangerous state of affairs for
White, since his kingside looks fragile
- the pawn on h4 cannot be supported
by the g-pawn and Black can attack 12 a3
down the g-file. Of course, if White This allows Black to dissolve the
were able to assume the initiative and white centre. Correct was 12 0-0, al¬
capture the pawn on f4 without loss though Black would have had attack-

30
 Kieseritsky Gambit (3 ^f3 g5 4 h4 g4 5 ^e5)

ing chances after 12...Shg8 etc. The «xf6 8 ac3

weakness of the h4-pawn would
greatly abet the attack.
Here we see again the unfortunate
situation of the knight on f2. Ideally
White would like to castle queenside,
but how can he achieve his? Both 12
.^d2 and 12 We2 lose to 12...?^xd4,
while if 12 #d3 then 12...?^b4 harasses
the queen. With the knight on f3
rather than f2, d4 would be safe and
White could continue with #e2, .^d2
and then 0-0-0. It is therefore easy to
conclude that 7 <53f2 has proved un¬ 8.. .C6
sound. Black secures control of d5 to pre¬
12.. .‘Sixe4! IS'SidB vent his queen being driven away
If 13 ?3fxe4 then 13...f5 regains the from its excellent post by 9 ^d5. The
piece with a clear advantage in devel¬ other method was 8...Ae6, which is
opment and king safety. considered in Game 10.
13.. .«e8 14 0-0 f5 9 Wf3
White’s once proud e4-pawn has Instead 9 d4 could be answered by
been replace by a powerful black 9.. ..^g7, when d4 is hard to defend.
knight. White therefore tries a more re¬
15 c3 fig8 16 fiel «f7 17 i.xf4 strained approach, intending #f3, d2-
.^xf4 18 4ixf4 <2ie7! d3 and <53e2 to win the f4-pawn. Black
An excellent move, preparing has to respond energetically by utilis¬
...Ac6 to put pressure on the vulner¬ ing the g-file.
able g2-square. White tries to disrupt 9.. .6.8
the gradual build-up of Black’s attack Not 9....^h6? because of 10 g4! with
by capturing on e4, but this leads to a clear advantage to White. This trick
tactical disaster. to exploit the pin on the f-pawn to
19 ‘^xe4 fxe4 20 fixe4 d5! 21 Wb3 straighten out White’s pawn structure
fidf8 22 fixe7 «xe7 23 ^xd5 «xh4 is well worth knowing. Sometimes it
24 ^e3 c6 25 Sfl Wg5 26 c4 Se8 occurs in a different form, when
27 ^d1 «h4 28 d5 «d4-i- 0-1 White has played d2-d4, threatening
.^xf4, and Black has defended the f4-
Game 9 pawn with ....^h6. Then, if the bishop
Nunn-Timman on h6 is undefended, a diagonal pin
Amsterdam 1995 can also be exploited with g2-g4!
10«f2
1 e4 e5 2 f4 exf4 3 <^f3 g5 4 h4 g4 A finesse, but there seems nothing
^e5 d6 6 ‘Sixg4 <2if6 7 wrong with he immediate 10 d3.

31
The King’s Gambit

However, if 10 then 10...?^a6!? Before White can castle he has to

with the idea of 11...^b4 looks awk¬ work out how to develop his bishop
ward for White. on fl, since after 14 Axf4 both 15
10....^g4 11 d3 .^h6 12 ^2 i.xf4 i.xe2 16 ^e2 #xb2 and 15
It would be bad to play 12...f3 13 #xf4 #xf4 16 i.xf4 i.xe2 17 <^xe2
i.xh6. Sxg2-i- are unsatisfactory. He decides
13 4^x14 0-0-0 to fianchetto, but further weaknesses
are created on g3 and f3.
14.. .We5!
This clears the way for a pawn at¬
tack on White’s centre. The position is
now very sharp and unclear.
15i.g2
The pawn snatch 15 #xa7 is dan¬
gerous, e.g. 15....^f3 16 Sgl f5 -
Korchnoi.
15.. .f5 16 0-0fxe4 17 i.d2?
After 17 .^xe4 Sdf8 Black would
have good play for the pawn, e.g. if
White has won the weak f4-pawn the rash 18 #xa7? then 18...i.h3! 19
and if he succeeds in consolidating he Sf3 .^xf4 wins for Black. However,
will be winning. However, his devel¬ after a sensible reply such as 18 .^g2
opment has suffered badly. In particu¬ Black would find it very difficult to
lar his king’s position is alarming. break through on the kingside, espe¬
Where can his king seek refuge.^ He cially as he has no more pawn thrusts
cannot very well castle queenside be¬ at his disposal. Chances would remain
cause moving the bishop on cl allows balanced.
...Wxb2. And besides, first of all the 17.. ..^13!
black bishop would have to be driven Black seizes the chance to exchange
way from the g4-square where it con¬ off the light-squared bishops. This fa¬
trols dl, which would not prove easy. vours him in two ways. First, the ex¬
It is also dangerous to stay in the cen¬ change clears the way for a rook as¬
tre, since Black can prepare the line sault against g3. And second, although
opening ...f7-f5 or ...d6-d5 pawn ad¬ White’s pawns on g3 and h4 and the
vances. This leaves the kingside, which knight on f4 are well entrenched on
is not very appealing since Black will the dark squares, the light squares such
have a readymade attack along the g- as f3, g4 and h3 have been compro¬
file. Nevertheless, castling kingside is mised. With the disappearance of
clearly White’s best option. Black can White’s light-squared bishop these
attack but at least there are many squares become severely weak.
white defenders at hand. 18 Sael Sdf8 19 dxe4
14 g3 If 19 .^xf3 then 19....^xf4 20 .^xf4

32
 Kieseritsky Gambit (3 ^f3 g5 4 h4 g4 5 ^e5)

Wxf4 and the g-pawn drops. Black develops and protects the d5-
19.. ..^xg2 20 <4>xg2 Sg4 21 <4>h3 square. For 8...c6 see the previous
In order to defend g3 next move game.
with a rook. 9Wf3
21 ...SfgS 22 Sgl ^f6 An important moment. The main
The knight joins in the attack and alternative is 9 #e2, as recommended
threatens e4. Black’s onslaught now by Gallagher in his book. This threat¬
increases in intensity until the fragile ens 10 Wb5-i- and rules out 9...?3c6
white kingside collapses. because of 10 ?3d5. Then a critical po¬
23 #f3 #e7 24 fig2 .^xf4 25 .^xf4 sition is reached after 9...?3d7 10 b3
We6 26 *h2 27 See2 (this is virtually the only way to de¬
No better is 27 .^g5 h6. velop the bishop) 10...Sg8 11 Ab2.
27.. .5f8! 28 Sef2 ‘Sixf4 29 gxf4
Bxh4+ 30 4?g1 «xa2!
Unexpectedly the final break¬
through occurs on the queenside.
Now White’s only chance was 31 Sg4,
but in any case the game was not to be
saved.
31 «g3 WbU- 32 Sfl fihU- 33
*xh1 WxfH- 34 Sgl «xf4 35 «h3-i-
*b8 36 «xh7 a6 37 5g8 WcU- 0-1

Game 10
Gallagher-Btyson Does White have the advantage or is
Hastings 1994 Black’s counterplay sufficient? There
are two variations to analyse:
1 e4 e5 2 f4 exf4 3 ‘Sif3 g5 4 h4 g4 a) 11...0-0-0 12 0-0-0 i.g4 13
5 iSieS d6 6 ‘Sixg4 4^16 7 4ixf6-i- i.xdl 14 #xa7 i.g4 15 i.a6 ^c5 16
Vxf6 8 ^c3 ^e6 J.b5 ?3d7? 17 ?3d5 18 J.c6! and
White wins. This pretty variation is
given by Gallagher. However, he
mentions, but doesn’t analyse, 16...c6!
This looks no better than unclear for
White, e.g. 17 ^d5 Wxb2-i-! 18 '^xb2
.^g7-i- 19 '^bl cxd5 20 exd5 with an
unusual material balance. Neverthe¬
less, White can get the advantage after
11...0-0-0. Simply 12 #12!, attacking a7
and planning 0-0-0 next move, gives
him a good game.
b) ll...i.g4! 12 W{2 d5 13 i.e2 J.c5

33
The King’s Gambit

14 #fl ^xe2 (14...(lxe4 15 ^xg4 Sxg4 Here 11...0-0-0 is bad after 12

16 0-0-0 ^d4 17 Sel looks a little bet¬ <53xa7-i- '^b8(?) 13 ^c6-i- bxc6 14 .^xa6.
ter for White) 15 WxeZ 0-0-0 16 0-0-0 However, a critical alternative is
dxe4 17 ^xe4 and after 17...#g6? 18 11.. .5.8 12 e5 #g6!.^ (not 12...dxe5 13
#c4! White obtained the better #xb7). Now White can try 13 h5,
chances in De La Villa-Femandez, when 13...Wxc2? 14 ^a3!! wins
Barcelona 1990. He is attacking the Black’s queen. However, 13...Wg4 14
bishop and also has the positional exd6 c6 15 ^c7-i- ^xc7 16 dxc7 ^d7 is
threat of 19 ^g5. Instead, 17...#c6! unclear. Probably his best chance is 13
looks much better, since it defends the <53xd6-i-, when 13... cxd6 14 .^xa6 dxe5
bishop and has an x-ray attack on g2 15 .^xb7 Sd8 gives complications
through the knight on e4, which is which seem to favour White.
precariously placed. I can’t see any 12 g4
advantage for White here. This seeks to achieve a bind on the
9.. .1.h6 position. If Black does nothing fast
The reason why 9 #f3 has long then White will develop his pieces and
been discredited is 9...Sg8! 10 #f2 pick off the f4-pawn. Therefore an
^c6 11 .^b5 (or else 11...^d4 follows aggressive response is required from
strongly) 11...0-0-0! 12 .^xc6 bxc6 13 Black.
d3 (13 #xa7 Sxg2 gives Black the 12.. .«e7 13 .^xf4 .^xf4 14 «xf4 f5!
stronger attack) 13...J.h6 in De La Black’s covmterplay comes just in
Villa-Fernandez, Salamanca 1990. That time. White now finds that he cannot
game continued 14 Sfl Sg4 15 g3 #g7 hold on to his e-pawn in view of the
16 i.xf4 i.xf4 17 gxf4 Sg2, when vulnerable position of his king.
Black’s initiative offered him at least a 15 gxf5 .^xf5 16 ^c3 SaeS 17 0-0-0
draw. Gallagher mentions this game in If 17 J.d3 .^xe4!, while 17 .^xa6
his book, yet here he plays 9 #f3 bxa6 doesn’t help.
anyway. It would be intriguing to 17.. ..^xe4 18 Sgl-H ^h8 19 Wxe4
know what improvement he had in Wxe4 20 ^xe4 5xe4 21 .^g2 fie2
mind. 22 Sdel
10 4^b5 Black has enough activity to draw.
This moves looks a little odd since 22.. .fief2 23 i.xb7 ^b4 24 fig2 c6
White embarks on a tactical adventure 25 fixf2 fixf2 26 fie8-<- <ig7 27 a3
with his queenside undeveloped. ^xc2 28 .^xc6 4ixd4 29 .^d5 ^2+
However, Gdlagher has prepared a 30 'i’bl 31 .^c6 'i>f6 32 fia8
forcing variation that seeks to exploit 4id3 33 .^e4
some concrete features of the position.
Note that if Black had played 9...Sg8, Game 11
10 ^b5 would fail to 10...?3a6 11 d4 Winants-Z.AImasi
c6, since the d6-pawn is defended by Wiijk aan Zee 1995
the bishop.
10.. .41a6 11 d4 0-0 1 e4 e5 2 f4 exf4 3 4if3 g5 4 h4 g4

34
 Kieseritsky Gambit (3 i^f3 g5 4 h4 g4 5 5^e5j

5 ^eS d6 6 5^xg4 ±e7 Wxg3 fxg3 13 i.xg5 gxi2+ 14 <^xf2 is

A logical move which seeks to gain clearly good for White) 12 ‘^fl! '53c6
time by attacking the h-pawn. (too late!) 13 ^e2 i.g4 14 ^xg4 Wxg4
15 c3 0-0-0 16 Wxg4-(- ^xg4 17 i.xf4
±xf4 18 <53x14. White has regained his
pawn and now enjoys the advantages
of a better centre, a bishop against a
knight, and lots of weak black pawns
to attack. Needless to say, Gallagher’s
technique was relentless.
7.. .±xh4+ 8 53f2 «g5 9 «d2
Bangiev claims that 9 Iiff3!? is inter¬
esting. Then there is a more or less
forced sequence 9....^g3 10 <53c3 <53c6
(10...i.g4?! 11 Wxg3!) 11 <53e2. White
7d3!? has the edge after both ll....^xf2+ 12
This new idea was suggested by Gal¬ Wxf2 and ll...^e5 12 Wxg3!, so Black
lagher and received its first interna¬ has to speculate with ll...<53b4!? 12
tional test in this game. Previously ‘^dl Wc5 (probably better is
White had played 7 d4, but Black had 12.. .<53xc2!? 13 <^xc2 Wc5+ 14 <^dl
good counterplay after 7....^xh4+ 8 ±xf2 15 ±xf4 with unclear play) 13 c3
^{2 Wg5! 9 Wf3 (the natural move, Wxf2 14 Wxf2 i.xf2 15 cxb4 i.e3 16
threatening 10 ±xf4 or 10 lfxf4) ^xf4 i.xcl 17 Sxcl c6 18 ^h5 and
9...'53c6! 10 Wxf4! I.xf2+ 11 ‘^xf2 White has the initiative (analysis by
Wxf4-(- 12 i.xf4 ^xd4 13 '53c3! i.e6 Bangiev).
(not 13...^xc2? 14 ^d5 9.. .±g3 10<53c3<53f6
[14...'53xal 15 ^xc7+ <^d8 16 '53xa8 Developing and preventing 11 <53d5.
^c2 17 ±xd6] 15 Scl and White is 11 <53e2
better - Bangiev) 14 '53b5 '53xb5 15 This threatens 12 <53xf4. Almasi sees
±xb5+ ^d7. White has compensation that ll...<53g4, pinning f2, loses a piece
for the pawn - the two bishops and to 12 <53xg3 and therefore prepares this
lead in development - but this is not a move with
serious winning attempt. The game 11.. .«e5!?
Shumilin-Voikov, Correspondence This improves on Gallagher’s analy¬
1989, went 16 ±c4 and here a draw sis, which runs ll....^xf2+ 12 ‘^xf2
was agreed. Curiously, Gallagher- <53g4+ 13 ‘^gl <53e3 14 <53xf4 <53xfl 15
Neffe, Hamburg 1995, went instead 7 <^xfl and Black has a very inferior
Iiff3 but then transposed to the above position as his weaknesses remain and
variation after 7....^xh4+ 8 '53f2 Iifg5 9 his dynamism has vanished. Almasi’s
d4. Now 9...'53c6! is the Shumilin idea is to answer 12 <53xf4 with
game. Instead, Neffe played 9...1.g3? 12.. .<53g4!, which certainly looks very
10 ^c3 ^f6 11 i.d3 Sg8 (ll...i.g4 12 awkward for White.

35
The King's Gambit

12 5^xg3 fxg3 13 5^h3 5^c6 14 «c3 tage. White had to eliminate the g-
Clearing the way for the c 1-bishop. pawn with 24 .^xg3!, when Bangiev
14...Hg8 15 ±f4 «xc3+ 16 bxc3 gives the variation 24...Sxg3 25 Sxh2
5^h5 ^g4 26 Shhl f5 (26...Sxg2-i-? 27 <^f3)
27 '53f4 fxe4 28 Sxh7 as unclear.
24.. .hS 25 d4h4!
The black kingside now looks com¬
pact. Of course 26 dxe5 dxe5+ would
be very bad for White. There now
follows a gritty positional battle in
which Black eventually proves the
value of the kingside pawns.
26 libel f6 27 lle2 €ieg4 28 llbl a6
29 c4 ade8 30 ^d3 SgfO 31 Sbel
^d7 32 c3 ae7 33 a4 b6 34 a5 bxa5
35 Sal ab8 36 axa5 Sbl 37 c5 Sfl
The endgame is difficult to assess. 38 cxd6 cxd6 39 axa6 5if2+ 40
Black still has the gambit pawn but ihxfl gxf2 41 Hxf2?
White has a strong dark-squared White could still have defended
bishop. The key question is whether with 41 Sxd6-i- <^e8 42 Sd2! -
Black can convert his kingside pawns Bangiev.
from a defensive liability into a dy¬ 41.. .Hxf2 42 Jixhl h3!
namic, game-winning unit. Since at The triumph of the black kingside
the moment the pawns are dislocated pawns is complete. The f-pawn costs
and unable to support each other, this White the exchange and soon the h-
seems unlikely. However, in the com¬ pawn will cost him a piece.
ing struggle Almasi plays with great 43 Hxd6+ *e8 44 e5 Hxg2 45 ±f4
determination and exploits some er¬ h2 46 ±xh2 Hxh2 47 Hxf6 Hh3+ 48
rors by his opponent. Probably Black ^e4 axc3 49 aa6 ah7 50 aa8+ ^d7
is slightly better in this position, since 51 ^d5 ah5 52 Rg8 U.a3 53 ag7+
it is easier to imagine Black winning ^e8 54 ag4 aa5+ 55 ^e4 ^f7 56
than White. This casts doubt on the Hf4+ *e7 57 Hg4 Hhl 58 Hg7+ ^8
idea of 9 Wdl - Bangiev’s 9 Iiff3 looks 59 ad7 ^e8 60 SdO ^e7 61 ^d3
like a better try. ah4 62 ^c3 aa3+ 63 ^c4 SaO 64
17 ±e3 ±g4 18 ±e2 5ie5 19 *d2 ^c5 ad8 65 axd8 ^xd8 66 d5 ^d7
i.xe2 20 ^xe2 €if6 21 fiafl €ifg4 0-1
22 ±f4
Here 22 .^gl!.^ seems like a better Game 12
try, e.g. 22...<^e7 23 h6 24 Sh3 Winants-Van der Sterren
with unclear play - Bangiev. Wijk aan Zee 1995
22...5ih2 23 Sbl 0-0-0 24 *d2?
Now Black gains a serious advan¬ 1 e4 e5 2 f4 exf4 3 5if3 g5 4 h4 g4

36
 Kieseritsky Gambit (3 ^f3 g5 4 h4 g4 5 g^eSj

5 ‘aeB ‘afe 6 ±c4 This is examined in the notes to move

The most aggressive response to five in Game 8.
5...‘S^f6. The more solid 6 d4 is consid¬ 8 d4 5ih5
ered in Games 15-17 below. The alternative is 8...0-0, but after 9

6 ...dS 0-0 both 9...^h5 10 '53xg4 «xh4 11

This is virtually forced in view of ^h2 and 9...^xd5 10 i.xd5 Wxd5 11
the attack on f7. It has long been ^c3 Wd8 12 i.xf4 Wxh4 13 ^d5 are
known that 6...We7 is bad, e.g. 7 d4 d6 good for White - Gallagher.
8 l.xf7+ ‘^dS 9 ±xf4 dxe5 10 dxe5+ 9 0-0l'xh4 10 Wei!
i.d7 11 i.b3 12 ^d2 ^xe4 13 White has to force the exchange of
c3 ‘5^xc3 14 .i.g5+ with a clear advan¬ queens, as Black’s coming attack with
tage to White (Korchnoi and others). 11.. .'53g3 etc. looks dangerous.
7 exd5 10.. .Wxe1 11 axel 0-0 12 5ic3 5id7
Black’s strategy is to undermine the
knight on e5.
13 5ib5
This is very logical as it attacks the
most vulnerable point in Black’s posi¬
tion.
13.. .c5!?

7...Jk.g7
The alternative 7....^d6 is the sub¬
ject of the next game. Which of these
bishop moves is the stronger? The fi-
anchetto is of great value, since the
bishop will exert strong pressure
against the d4-pawn in the future. It
also strengthens the black kingside, The consistent move, attacking the
which means that the king will be se¬ defender of e5. However, theory sug¬
cure there. The drawback is that, gests that there is a problem with this
compared to 7....±d6, Black leaves the move, viz. the variation 14 ^c7 Sb8
c7- and f4-pawns undefended. As we 15 d6! cxd4 16 <53x17! and White wins,
shall see. White can try to exploit this e.g. 16...Sxf7 17 Se7 <53e5 18 Se8+
with a later '53b5. ±f8 19 Sxe5 etc. Since White avoided
It should be mentioned that this po¬ this variation in the game and as far as
sition can also be reached via an alter¬ I know the players have never re¬
native move order beginning 5....^g7. vealed their thoughts in annotations.

37
The King’s Gambit

we have to try to discover for our¬

selves what improvement Van der
Sterren might have prepared. Perhaps
it was 14 ^c7 Sb8 15 d6 ^b6!?

His analysis runs 18 e7? (18 exf7+

<^h8 19 i.e6 i.xe6 20 ^xe6 Sxf7 is
good for Black) 18....^d4+ 19 ‘ii’h2 (or
19 <^hl Sxc4 20 cxm+ <&cf8 21
Now the critical line is 16 dxc5 Se8-(- <^g7 22 axc8 ^g3+ 23 *h2
'2ixc4 17 '2ixc4 .^d4+ 18 ‘^h2 (after 18 ^fl-i- 24 <^hl Sc5 25 g3 Sh5-i- 26 *g2
‘^hl.5 ±xc5 19 Se5 Black has at least a 27 <&cfl Shi mate) 19...g3+ 20
draw by perpetual with 19...'2ig3+). <^hl Sxc4 21 exf8W+ <^xf8 22 Se8+
Black would lose a piece after <^g7 23 Bxc8 f3 24 c3 f2 25 i.g5 i.f6
18.. ..^xc5? 19 Se5, and the s^ressive 26 '53e8+ ‘^g6 27 <53x16 ‘^xg5 and
18.. .g3-i-19 <^hl f3 20 gxf3 i.h3 faUs to Black wins. Here we see the strength
21 ±e3. More interesting is 18...±f2, of Black’s attack against the white
but White has good compensation for king if White loses control. Despite
the exchange after 19 Se5 f5 20 '^^dS f3 White’s big material advantage, he will
21 gxf3 .i.g3+ 22 ‘^g2 ±xe5 23 ^xe5. lose the game because his king has be¬
Therefore, Black has to try 18...b5!? 19 come entombed on the h-file.
cxb6 (forced, as if the knight moves Gallagher suggests 18 .^b3! but
Black can capture on c5: 19 <5335 .^xc5 doesn’t provide any analysis. I suspect
20 Se5 i.xd6) 19...axb6 20 Sdl (20 d7 that White is also in trouble here, e.g.
.^bT) 20....^f2 21 d7 .^b7 and Black 18.. .fxe6 19 <S3xe6 (if 19 .^xe6+ i.xe6
has good attacking chances against 20 <53xe6 Se8 White is a pawn down
White’s king after ...f4-f3 etc. and pinned) 19...Bxb3! 20 axb3 (20
So it seems that 13...c5 stands up to ^xf8 Sg3! and if the knight moves to
analytical scrutiny. However, Black safety 21...f3 obliterates the kingside)
also has an alternative move, 13...c6. 20.. .5e8 21 Sa5!? i.d4+ 22 ^xd4 (22
According to Gallagher the critical <^h2 ^g7!) 22...Sxel+ 23 *f2 Sxcl 24
variation is now 14 dxc6 '^ixeS 15 dxe5 ]Sxh5 Sdl 25 ^xc6 Sd2+ and Black
bxc6 16 ^c7 Sb8 17 e6 Sb4! has every chance to win the endgame
after 26 ^el Bxc2 or 26 ‘ifegl .^b7.
see following diagram _
Finally we should consider 13...c6

38
 Kieseritsky Gambit (3 t^^f3 g5 4 h4 g4 5

14 ^c7 when, since 14...Sb8 15 d6 is .^xc4+ Sxc4 22 <53x04 Sxel+ 23 <^12

bad, Black has to offer the exchange fxg5 23 ^e5-i- <^18 24 ^xd7-i- <^e8 25
with 14...cxd5! In R.Byrne-Keres, ‘^xel ‘^xd7 and in a matter of moves
USA-USSR 1955, White took the bait White lost both of his d-pawns.
and after 15 ^xa8? dxc4 16 .^d2 ^xe5 In our main game White tries a
17 dxe5 ^f5 18 ^c7 Keres claims a more solid move.
large advantage for Black with 17 a4 SadS
18.. .5.8 19 .^c3 Axc2. Instead of ac¬ This contains a latent threat to the
cepting the exchange, Glaskov rec¬ d-pawn (18...±xb5) which persuades
ommends 15 Axd5! Sb8 16 c3 '53xe5 White to move his knight. But not to
17 dxe5 Sd8 18 e6 fxe6 19 ±xe6+ a7, since 18 <53xa7? loses a piece after
l.xe6 20 '53xe6 Se8 21 l.xf4 with a 18.. .5.8.
fairly equal position. 18 <53c7 ±f6!
14 c3 Black finds an excellent way to acti¬
After this defensive move all the vate his bishop.
complicated variations above are left I9±d2
behind. However, White cannot This threatens to win the exchange
count on gaining any advantage as with 20 .^b4.
Black can rapidly mobilise his pieces. 19.. .±h4 20 Seel
14.. .cxd4 15 cxd4 5^b6 16 ±b3 Here 20 Sedl was interesting, when
±d7! 20.. .1.e7 or 20...<53c8!? 21 i.b4 ±e7
The attack on the white knight is were possible continuations.
awkward. Of course. White has no 20.. .^g3 21 Hel 53f5 22 He4 53g3
wish to exchange off his strong knight 23 Seel 53f5
on e5 for the bishop. In Welling- A curious finish. White’s rook has
Zagema, Holland 1995, White tried to defend d4 and cannot capture on f4
the spectacular 17 '53c7? Sac8 18 '53e6!? because of a fork on e2. Meanwhile,
However, Black calmly replied Black is threatened with .^b4 or .^xf4,
18.. .5.e8!, not allowing the bishop on so he also has to repeat. A case of both
b3 to be unleashed after 18...fxe6 19 sides standing badly!
dxe6. There followed 19 ^g5 f6 20
d6+, which looks pretty strong as it is Game 13
mate after both 20...‘^f8 21 '53xh7 and Grasso-Pampa
20.. .<ii>h8 21 <23ef7+ <^g8 22 '53h6+ <^h8 Correspondence 1995
23 ^gf7. But Black confounded his
opponent’s plans once again with 1 e4 e5 2 f4 exf4 3 <53f3 g5 4 h4 g4
20.. .^c4!! And this is only to be ex¬ 5 <53e5 <53f6 6 ±c4 d5 7 exd5 ±d6
pected. Every black piece is in play, The alternative bishop develop¬
whilst the white rook on al and the ment.
bishop on cl are still slumbering. Why 8d4
should White be able to win by a di¬ Now Black has a choice between
rect attack? The game continued 21 the game continuation 8...0-0! and

39
The King’s Gambit

8...'5^h5 (see the next game). #g4+ leads to mate. Therefore, it

seems that the question mark after
12.. .f3 is justified.
Nevertheless, the problem of
12.. ..^f5! still remains.
In our main game White decided to
avoid all this by reintroducing a long
discredited move.
9 .&.xf4l? 5ih5 10 g3
If 10 0-0? «xh4 11 i.h6 Se8! with
decisive threats including 12...Sxe5 13
dxe5 .^c5+ - Gallagher.
10.. .f6 11 4ixg4!?
8...0-0! This is the first new move of the
The theoretical assessment of this game, diverging from Pillsbury-
line favours Black, based on the game Chigorin, Vienna 1903, which went
De La Villa-Am.Rodriguez, Bayamo 11 ^d3 ^xg3 12 i.xg3 i.xg3-i-13 <^fl
1991, which continued 9 0-0 '^ihS 10 WeS with a clear advantage to Black.
^xg4 Iifxh4 11 '2ih2 12 Sel .^f5! 11.. .«e8+
and Black obtained a good game after ll...'53xg3 is inaccurate, as 12 .^xg3
a subsequent ...^d7 and doubling of .^xg3+ 13 ‘^d2 .^f4+ would leave
rooks on the e-file. Here it is also Black a tempo down on the game.
worth mentioning the sharp 12...f3?, 12 *d2 5ixf4 13 gxf4 ±xf4+ 14
which according to theory fails after ^c3
13 ^xf3 WhU 14 <^f2 ^e4-(- 15 <^e3
lfh6+ 16 ‘^d3 (16 ‘^xe4 allows mate
in two) 16...Wg6 17 Sxe4 .^f5 18
'53bd2 Se8 19 .^xe4 20 '53xe4
Iifxe4 21 Whl! and White’s attacking
chances and safer king are supposedly
worth more than the exchange. This
line received a test in the game Olesen-
Kristensen, Copenhagen 1995, which
continued 21...f6 (to defend against 22
±d3, hitting h7 through the queen) 22
i.d3 Wc7 23 Wh5 ^d7 24 i.h6 c5 25
dxc6 bxc6 26 ‘^d2 (a precaution Here we see that capturing the g-
gainst 26....^b4+) 26...Sad8 27 Shi c5 pawn with 11 ^g4 has two distinct
and Black achieved counterplay. advant^es over Pillsbury’s 11 '53d3.
However, 27 Shi was a little stereo¬ First, White can attack Black along the
typed. Instead 27 Sel! appears to win newly opened g-file and second, White
at once as 27...Wf7 28 Ac4! Iifxc4 29 no longer has to fear an endgame. In

40
 Kieseritsky Gambit (3 ihf3 g5 4 h4 g4 5 i^e5)

fact, he would have winning chances a fatal blow before White succeeds in
due to his extra pawn. However, the freeing his queenside?
endgame is a long way off. Mean¬ 19.. .«d7
while, White’s king is in a very odd It turns out that Black also has
position. The question is whether problems with his king, since there is
White can mobilise his queenside no good way to dodge the coming
pieces while at the same time fending discovered check, e.g. 19...‘^h8? 20
off an attack, which will be abetted by m6+ <^g8 21 d(>+ or 19...<^g7 20
Black’s queenside pawns and bishop Wxc7+. Also 20 lfh6!?, intending 21
pair. e6+, looks unpleasant. So Black forces
14.. .b5 White’s hand by preparing 20...‘^g7 or
If 14...h5 15 #f3 is similar. 20.. .1.b7.
ISWfS! 20 d6+?!
Sometimes attack is the best form of After 20 lfh6! Black has nothing
defence. Now Black cannot avoid the better than 20...1fd6 to block the dis¬
exchange of his important dark- covered check. White then has the
squared bishop, as 15....^d6 16 '53xf6+ luxury of a choice between forcing
is disastrous. However, he finds an perpetual with 21 Wg6+ or playing to
excellent riposte. win with 21 “^cl, preparing 22 '53d2 or
15.. .h5! 16«xf4hxg4 22 If this analysis holds up, then
Black has sold his prize bishop at a 9 .^xf4 revives White’s chances in the
high price, as now the g-file is closed 8.. .0.0 variation.
and his king is much safer than his 20.. .*g7 21 dxc7 5la6 22 h5 5lxc7
opponent’s. The g-pawn could also 23 h6+ <^h8 24 <^c1 25 ±xe6
become valuable in the endgame. Wxee 26 €ld2 ae8?!
However, Black is still hoping to win This threatens a back-rank mate,
by a middlegame attack on the ex¬ but according to Grasso 26....^b7 was
posed white king. stronger. Then 27 Sgl Sg8 looks un¬
17±b3 aS clear, but not 27...Sac8 28 Iifxg4! Sg8
Threatening 18...a4. 29 Wxe6 Sxgl-i- 30 ^fl Sxfl-i- 31 <^d2
18 a4 Sf2+ 32 <^el Scxc2 33 Scl! and White
The check 18 d6+ would merely wins.
open up the c6-square for the black 27 €lb3 aa6 28 ^b1
queen. Of course, if 28 '53c5 Wd5 attacks
18.. .b4+ hi.
If 18...bxa4 19 Sxa4! brings the 28.. .1.d5 29 <^a2
white rook into the game. White has a clear advantage due to
19 ^d2 his ascendancy over the dark squares
Now the knight on bl and the rook and his safer king.
on al are temporarily stalemated. 29.. .±d7 30 Sael ±xa4 31 axe8+
There now follows an arms race: can ±xe8 32 Wxg4 mi 33 <^b1 ±d7 34
Black develop his queenside and strike Wf4 SaS 35 Sgl SgS 36 axg8+ ^xg8

41
The King's Gambit

37 5^xa5 *h7 38 «d6 f5 39 5^b3 f4 40 (15...i.f5) 16 i.d2 ^xe2+ 17 Sxe2

^d2 f3 41 5^xf3 «xf3 42 «xd7+ *xh6 Sxe2 18 '53xe2 '^ixdS wins a pawn.
43*66+1-0 White could find nothing better than
15 '53e5, when 15....^xe5 16 dxe5 Sxe5
Game 14 17 i.d2 i.d7 18 i.f3 Sae8 19 Sxe5
K.Kristensen-Sorensen Sxe5 20 Sel 5xel+ 21 .^xel ^f6 22
Copenhagen 1995 .^d2 ^c4 23 .^xf4 ^xb2 ^c4
24 .^xc7
gave Black some winning chances in
1 64 65 2 f4 6xf4 3 gS 4 h4 g4 the endgame, as White’s queenside
5 4^65 6 ^c4 dS 7 exdS ±d6 8 pawns are all weak.
d4 5^h5 12....^f5 is less good, when Gal¬
Black defends f4 and attacks h4 lagher claims an advantage for White
without further ado. after both 13 .^d3 .^xe5 14 Sxe5!
i.xd3 15 Sxh5 i.xc2 16 Sg5+ i.g6 17
Sxg4 '53a6 18 ±xf4 and 13 ^e4 ±xe4
14 Sxe4 f6 15 '53xg4 f5 16 '53h6+ ‘^g7
17 Se6 Sf6 18 Sxf6 ^xf6 19 i.d3!
White’s knight is trapped on h6, but
after 19...'53g3 or 19...'53g7 to guard f5.
White can use the c-pawn to deflect
the bishop on d6 from the defence of
f4. White’s knight should eventually
be freed after a subsequent .^xf4. Gal¬
lagher gives the possible continuation
19...'53g7 20 c4 c5 21 b4!? cxd4 22 i.b2
9 0-0 *xh410*6l! .i.e5 23 g4!? with unclear play, but I
Compare this with Game 12. White prefer White.
welcomes the exchange of queens to
secure his king from a mating attack.
He trusts in his sounder pawn struc¬
ture, slight lead in development and
strong knight on e5 to compensate for
the missing pawn.
10.. .*xe1 11 Hxel 0-0 12 ^c3
568?!
A critical moment. It was better to
challenge the knight on e5 immedi¬
ately with 12...'2id7! Then 13 '53xg4
^b6 14 Ae2 Se8! looked at least equal
for Black in Riemersma-Van der Ster- 13±d2!
ren, Holland 1993. The threat is The lack of pressure on e5 gives
15.. ..^xg4, and if 15 ^f2 then 15...'53g3 White time to devise a plan to destroy

42
 Kieseritsky Gambit (3 ^f3 g5 4 h4 g4 5 4^e5j

Black’s hold on d6.

13.. .±f5
Perhaps Black should have tried
14 '2ixg4 Sxel+ 15 Sxel
'2lb6, but the game has ceased to be of
theoretical interest.
14 5ib5! 5id7 15 5ixd7 i.xd7 16
5ixd6 cxd6 17±b4!
Now d6 is indefensible. White
therefore acquires a strong passed
pawn which, supported by the two
bishops, gives him a decisive advan¬
tage. Such reasoning overlooks the rela¬
17.. .±f5 18 .&.xd6 ±xc2 19 JLe7 tive dynamic strength of the f4- and g4-
i.a4 20 d6 5ig7 21 He4 5ie6 22 pawns. The g4-pawn is usually a posi¬
.^xe6 fxe6 23 fixf4 hS 24 fiel SacS tional nonentity in the Kieseritzky,
25 He5 Hc1+ 26 *h2 *g7 27 Hxh5 reducing the scope of the bishop on c8
1-0 and depriving the knight of the g4-
There is no good defence against the square. It also obstructs any counter¬
threat of 28 Sxg4+ 29 Sh7 mate. play based on ...Sg8 (we have already
seen the strength of the ...Sg8 attacks
Game 15 in the Shirov and Timman games ear¬
Matsuura-Van Riemsdijk lier in the chapter). In effect, the g-
Brazil 1995 pawn only had one purpose in life and
that was to defend the f4-spearhead; as
1 e4 e5 2 f4 exf4 3 5if3 g5 4 h4 g4 soon as it was driven to g4 by White’s
5 5ie5 5if6 6 d4 4 h4!, it lost most of its value. This
This is the main alternative to 6 explains why in the other main varia¬
l.c4. tion Black is happy to sacrifice the g-
6...d6 7 5id3 4ixe4 pawn immediately with 5...d6! in or¬
7...f3 is the subject of Game 17. der to gain active play.
8±xf4 The f4-pawn, on the other hand, is
White’s play may seem confusing to often a real nuisance to White. It re¬
a player unfamiliar with the positional stricts the bishop on cl to just one safe
complexities of the King’s Gambit. He square, the unimpressive d2, controls
has exchanged his proud e-pawn for the central e3-square and shuts White
the black f4-pawn, when at move six out of f4. It also blocks the f-file and so
he could have exchanged it for the prevents an attack on f7 with .^c4 and
black g-pawn (6 '53xg4 '53xe4). Surely it Sfl etc. So, although classically weak,
makes more sense to capture the g- from a dynamic view point the f4-
pawn, leaving Black with a doubled pawn has great value: it is the linchpin
and isolated f4-pawn? of Black’s position and holds his dark

43
The King's Gambit

squares together. That is why White This had been thought dubious, but
often plays g2-g3, allowing the f-pawn in view of Black’s improvement at
to advance to f3 and become a pro¬ move 12, it may in fact be the best
tected passed pawn. In return, the move. It avoids the unpleasantness of
bishop on cl gains access to some 9...We7 10 Ae2 (which transposes to
strong dark squares such as g5 or even the next game).
h6. 10 <53d2 Re8 11 <53xe4 Hxe4+ 12 *f2
The King’s Gambit often revolves c5!
around the f4-pawn. Can White dis¬
lodge it, or, better still, destroy it? If
he can do so at no great loss elsewhere,
then he usually has a pleasant game; if
the pawn remains firm then it can of¬
ten choke the life out of White’s posi¬
tion.
In the present variation. White
solves the problem of the f4-pawn by
destroying it immediately. The
queen’s bishop feels the benefit and is
excellently posted on f4. On the other
hand, the loss of the e-pawn is an Shades of Fischer! This looks much
enormous positional concession. better than 12...1ff6 13 g3 ±h6 14
The other move 8 is examined Wd2!, when White is ready to play 15
in the notes to Game 16 below. .^g2 with an advantage. It is always
8.. .±g7! good to get pawns involved in an at¬
As usual, this bishop proves very tack!
strong when it can be fianchettoed. 13 dxcS dxcS
9 c3 Black has negated White’s space ad¬
White’s main aims are to drive the vantage in the centre, activated his
knight away from e4 and defend d4. In queen without even moving having to
the famous game Spassky-Fischer, Mar move it, and opened up the white king
Del Plata 1960, he tried to combine to threats along the diagonal a7-gl
both ideas with 9 '53c3?! However, (after ...c5-c4 etc.).
after 9...<53x03 10 bxc3 c5! (the the¬ 14 g3 WbO 15±g2
matic move, striking at White’s cen¬ This leads to defeat after some fine
tre) 11 Ae2 cxd4 12 0-0 <53c6 13 .^xg4 play by Black. White had to try 16
0-0 14 i.xc8 Sxc8 15 %4 Black could <^g2 (16...C4?! 17 ^f2 Wxb2?? 18
have played 15...‘ifeh8 with a good po¬ Wd8+), although after 16...<53c6 it is
sition (Fischer). clear everything has gone wrong for
Therefore, White safeguards his White.
centre. 15...C4+ 16 *f1 He8 17 53b4 53a6!
9.. .0-0 18 53xa6.^f5!
 Kieseritsky Gambit (3 ^f3 g5 4 h4 g4 5

Now there is no good answer to the 17 .^xd3 ^b4-(- (the point) 18 '^b3
threat of 19...^d3+. C^xd3 and White resigned.
19 <^b4 Sad8 20 <^d5 b) The game Holmes-Hebden, Ply¬
If 20 #a4 simply 20...a5 eliminates mouth 1989, continued 13 Sel? #e6!
the knight, followed by....^d3+. and White was already in trouble as 14
20.. .5xd5! 21 i.xd5 i.d3+ 0-1 ■^bl loses to 14...^xd2-i- 15 ±xd2
If 22 '^g2 Wxh2+ 23 '^gl ±xc3 and «xe2 16 ±xe2 Sxe2! And if 14 ^xe4
there is no answer to 24...^d4+ or then 14...#xa2 anyway. Therefore,
24.. .«xal. White had no good way to defend a2.
He tried 14 a3 but, was quickly over¬
Game 16 whelmed after 14...«a2 15 «dl h5 16
Henris-Goossens g3 ±g7 17 Sh2 ^xc3! 18 bxc3 #xa3+
Charleroi 1994 19 '^c2 ±xd4 20 #al ^b4+ and
White resigned.
1 e4 e5 2 f4 exf4 3 4^13 g5 4 h4 g4 c) Another try is 13 d5, which Gal¬
5 -SieB 6 d4 d6 7 ‘Sid3 <5^X64 8 lagher refutes as follows: 13...^xc3! 14
i.xf4 «xe7 ^xa2+ 15 -^bl ^xe7 16 '^xa2
In view of Black’s convincing play ^xd5. White loses a piece and remains
in the game above, White should con¬ three pawns behind.
sider 8 #e2, which forces the reply Gallagher therefore suggests 13
8...#e7. However, White is also strug¬ ^xe4 as best, when after 13...#xe4 14
gling in this variation. The critical po¬ Wxe4 ±xe4 15 ^f2 f5 White’s posi¬
sition is reached after 9 .^xf4 ^c6 10 tion will be very hard to break down.
c3 ±f5 11 ^d2 0-0-0 12 0-0-0 Se8 Nevertheless, this isn’t what White
(12....^g7 deserves attention). wants when he plays the King’s Gam¬
bit. We can only conclude that the
variation 6 d4 is under a cloud for
White.
8.. Me77\
Since the previous game proves that
8.. ..^g7 is playable, this move, which
aims for ...0-0-0, seems inappropriate.
9 i.e2!?
see following diagram_

The best try for White, avoiding a

transposition after 9 We2 to the un¬
This position is very dangerous for pleasant variation examined at move
White. Here are three ways to lose: eight above.
a) Hajek-Bures, Correspondence 9.. .1.g7
1962, went 13 g3? ^xc3! 14 «xe7 Gallagher analyses 9...h5, 9...^c6
^xa2+ 15 '^bl Sxe7 16 '^xa2 ±xd3 and 9....^f5 as deserving attention.

45
The King’s Gambit

However, the game move is very 21 <^d1

natural. Black has four pawns for the piece
but his king has no safe place and he is
badly behind in development. What
follows is desperation.
21...h5!? 22 ^xc6 *f8 23 g3
24 Wdl h4 25 Scl h3 26 Sxc7 'tf4
27 We7+ *g7 28 Sxg3-H 1-0
It is mate in two after 28...#xg3 29
«xf7+'^h6 30«h5.

Game 17
Spassky-Xie Jun
Monaco 1994
10 4ic3! i.xd4?
This loses. It seems that Black has 1 e4 e5 2 f4 exf4 3 ‘Slf3 g5 4 h4 g4
nothing better than 10...^xc3 11 bxc3. 5 -SleB 6 d4 d6 7 ‘Sld3 f3
Then ll...c5?! (Il...^c6! is safer) fol¬
lows Fischer above (see Game 15,
move 9). However, Black has wasted a
tempo in playing 9...#e7. Not surpris¬
ingly, this changes the theoretical ver¬
dict: 12 0-0 cxd4 13 .^xg4 0-0 14 .^xc8
Sxc8 15 #g4 and White has a danger¬
ous initiative.
11 4ld5! I'dS
Henris gives ll...#d7 12 c3 Ag7 13
h5! h6 14 ^f2! ^xf2 15 '^xf2 as ‘good
for White’, but this was certainly a
better try for Black. Xie Jun is well prepared in the
12 c3 i.e6 13 #34-^! openings and comes up with a new
This unexpected move is much idea in this familiar setting. But I don’t
stronger than 13 cxd4. like it! Instead of capturing a pawn - a
13.. .-Sice 14 cxd4 i.xd5 15 4154! healthy centre pawn - Black gives up a
The point. If now 15....^e6 16 pawn and makes any future ...^xe4
^xc6! #d7 (16...bxc6 17 #xc6-(- wins liquidation problematical. White
the knight on e4) 17 d5 ^c5 18 #d4 maintains a strong centre: indeed, it is
wins material. made stronger by 7...f3. Certainly, the
15.. .-Slf6 16i.g5 kingside becomes inhospitable for his
The pin on f6 will prove fatal. king, but there is always the queen-
16.. .1.xg2 17 Sh2 h6!? 18 i.xf6 side, either through 0-0-0 or ‘^d2 and
Wxie 19 Sxg2 'txh4+ 20 *d2 l^gB-l- '^c2 (after preparation of course).

46
 Kieseritsky Gambit (3 thf3 g5 4 h4 g4 5 ^eS)

However, it seems that the verdict The immediate 14...gxf3 gives

on the position depends on a piece White the edge after 15 .^xf3 dxe4 16
sacrifice in the analysis below. *xe4+ ^e7 17 ^d2 M5 18 *e2.
8 gxf3 ^c6 9 c3 i.e7 10 i.g2 Sg8 15?3d2 dxe4
11 i.g5! h6 12i.xf6
After 12 .^xh6 ^h5 and 13....^xh4+
Black has good play.
12.. .1.xf6 13 h5
A very unaesthetic move in the
King’s Gambit. The pawn advances
not with any attacking or positional
aims, but merely to avoid capture.
Nevertheless, White can be pleased
with his compact centre. Black’s next
move attempts to undermine it.
13.. .d5
This aims to break up the white 16 <53x64?
centre and so open more lines for the This is the critical moment in the
well activated black pieces. Other game. The natural move is 16 fxe4!,
moves don’t seem particularly promis¬ keeping the centre. Spassky probably
ing, e.g. 13...gxf3 14 .^xf3 .^h4-(- rejected it because he was afraid of the
(14...±g5 15 M «f6 16 «e2) 15 '^d2 sacrifice 16....^xd4!?, which certainly
Wf6 16 '^c2 Sg3 17 ^d2 ±g5 18 ±e2 looks very dangerous. However, it
with advantage to White. seems that if White is vigilant he can
14*62 hold his position together after the
14 e5 is less good, e.g. 14...gxf3 15 sacrifice and then exploit the extra
±xf3 (Black has a strong initiative af¬ piece. But let’s look at the variations:
ter 15 *xf3 M4+ 16 -^fl i.g4 17 17cxd4^xd4 18*dl.
«xd5 «g5) 15...±h4+ 16 '^d2 with
unclear play or 14...Ah4-(- 15 '^d2 (15
^fl ±f5!? 16 f4 ±g3) 15...i.f5. Black
seems at least okay in these variations
due to his more active pieces, espe¬
cially the dark-squared bishop, which
who has no white rival.
14.. .<i>f8
The black king is safe here as long
as the f-file remains inaccessible to
White’s rooks. Also, it doesn’t harm
the co-ordination of Black’s pieces,
since the king’s rook has found an ac¬ Now Black has a choice:
tive role on the g-file. a) 18...*g5 19 ^fl! ±e6 20 *d2

47
The King’s Gambit

Sd8 21 #xg5 Sxg5 22 Scl c6 23 ^c5

and White should win.
b) 18...g3 19 ^f3 (two variations
demonstrate the strength of Black’s
attack against inaccurate play: 19 0-0
«h4 20 ^f3 ^xf3+ 21 Sxf3 «h2+ 22
■^fl .^h3 23 .^xh3 g2-(- 24 .^xg2 Sxg2
and White will be mated; or 19 #a4
±g4 20 ^e5 b5! 21 «a3+ «d6! 22 Scl
[or else the fork on c2 is decisive]
22.. .b4 23 «a5 SgS 24 Sc5 Sxe5 25
Sxe5 ^c2+ 26 '^fl «d3+ 27 '^gl
«e3+ 28 ^fl i.e2 mate) 19...i.g4 20 White has succeeded in castling ‘by
^de5!, and it appears that White can hand’. He now stands better in the
defend successfully, when his extra centre, but it is difficult to break
piece will give him winning chances. through the obstacles on the kingside
Assuming that the above analysis is and get at the black king. Meanwhile,
correct, it seems that Black’s opening Black is preparing counterplay on the
experiment with 7...f3 is unsound. On queenside.
the other hand, it is no surprise that 24.. .C6 25 Shgl 'h-'h
Spassky had no wish to enter these Here Spassky offered a draw. He
sharp lines without pre-game analysis. might have tried 25 ^xg4, since if
16.. .1.g5 Black tries to regain her pawn with
25.. .41.3 26 ^g3 Axg4 27 Af3 Ae6
see following diagram
28 Shgl ±xa2, she faces a withering
The sacrifice on d4 now seems bad: attack after 29 ^f5.
16.. .Axd4 17 cxd4 ^xd4 18 and But Black can ignore the loss of a
White is on top. So Black provokes a pawn and continue her attacking
weakening in White’s centre by pre¬ build-up against White’s king with
venting 0-0-0. 25.. .Ae6 or 25...b5. It is a pity that the
17 f4 i.h4+ 18 *d2 a5 19 ^ecS game was cut short. The Spassky of
i.f6 20 Bael Sb8 21 *c1 ^e7 22 the 1960s would never have agreed a
<5165 23 5le4 i.e7 24 Sdl draw here!

48
 Kieseritsky Gambit (3 thf3 gS 4 h4 g4 5 ^eS)

Summary
The Kieseritzky is an enterprising variation that sets Black some difficult prob¬
lems. However, theoretically speaking, Black seems to have at least equal
chances in almost every variation.

1 e4 e5 2 f4 exf4 3 g5 4 h4 g4 5 4^5

5.. .d6
5.. .^f6
6 Ac4 d5 7 exd5 (D)
7.. .Ag7 - Game 12
7.. .1.d6 8 d4
8.. .0.0 - Game 13
8.. .^h5 - Game 14
6 d4 d6 7 ^d3 (D)
7.. .^xe4 8 .^xf4
8.. ..^g7 - Game 15
8.. .#e7 - Game 16
7.. .f3 - Game 17
6 ‘Sixg4
6.. .Ae7 - Game 11
7 ^f2
7 ^xf6+ «xf6 8 ^c3 (D)
8.. .c6 - Game 9
8.. .Ae6 - Game 10
7.. .1hc6 - Game 8

49
 CHAPTER THREE

other Gambits after

3 af3 g5 and 3...ac6

1 e4 e5 2 f4 exf4 3 Game 22 looks very strong for Black.

According to Korchnoi and Zak ‘in However, Game 23 is played in ro¬
answer to 3...g5 White has only one mantic style, with a heart-warming
means of obtaining a completely equal victory for sacrifice over petty de¬
game, and that is the Kieseritzky fence. In fact. White’s attack after 13
Gambit.’ The games in this chapter .^eS! looks devastating. But before
would appear to confirm this state¬ you start planning to carry out this
ment. Here you will find some fa¬ attack in your own games, remember
mous, enterprising and attractive sacri¬ that first you have to tread through a
ficial lines dating back to the golden minefield of positions that are better
age of the King’s Gambit, but none for Black.
that pass the modern test of analytical In Game 24 we see a selection of
soundness. unsound gambits after 4...g4. It is diffi¬
The Allgaier Gambit 4 h4 g4 5 4ig5 cult to know which is the worst, but
(Game 18) looks highly suspect for this dubious distinction should proba¬
White. bly go to the Lolli Gambit.
Games 19 and 20, the Philidor and The Pierce Gambit is made to look
Hanstein Gambits, show just how like a forced loss in Game 26. How¬
much White suffers when he fails to ever, when White tries the Pierce
undermine the black pawn chain with Gambit with a different move order in
4 h4! g4 5 ^e5! White already looks Game 25 he has a great success. Mi¬
uncomfortable after 4 .^c4 Ag7. At chael Adams seems bemused to be
least after 4...g4 he can cheer himself faced with the ancient attack and re¬
up by sacrificing a piece with the good acts too passively. This shows that on
old Muzio Gambit 5 0-0 (Games 21- the right occasion a bold choice of
23). Here I have to be cynical and opening can unnerve even the most
point out that 9...#f5! in the notes to steely opposition.

50
 other Gambits after 3 ^fS g5 and 3...^c6

(the standard idea to free his pieces;

Game 18 Black is a piece up and therefore
Neffe-Bronstein doesn’t begrudge returning one pawn)
Wrexham 1995 8 .^xd5-t- '^eS (8...'^g7 is also possible)
9 d4 ^f6 10 43c3 ^h5! (an excellent
1 e4 e5 2 f4 exf4 3 g5 4 h4 g4 move which defends f4; Black is not
5 '^igS prepared to defend passively - he
wants to attack!) 11 0-0 c6 12 J.b3
.^g7! (now the threat to the d4-pawn
gains time to bring another defender
to the f4-pawn) 13 e5 Hf8 14 ^e4
#xh4 15 43d6-i- ^d7 and Black has a
winning attack, with threats of 16...g3
or 16...^g3 or 16...f3. Of course, he is
also still a piece up! This variation is
analysis by Chabelsky, quoted from
Bangiev.

The Allgaier Gambit in its pure

form. A version via the Fischer De¬
fence with ...d7-d6 and d2-d4 thrown
in was considered in Game 5, while
the so-called Hamppe-Allgaier - i.e.
the Allgaier with 4ic3 and ...^c6 al¬
ready played - is analysed in the notes
to move five in Game 26. None of
these versions is theoretically water¬
tight, but in practice they can all prove
tricky. b) 7 d4 f3 8 gxf3 (Black is clearly
5...d5 better after both 8 ‘53c3 .^b4 9 gxf3 d5
Bronstein’s choice, and he knows and 8 ±e3 d5 9 ^c3 ±b4 10 Wd2 ^f6
about these things! However, accep¬ - Estrin) 8...d5 9 M4 ^f6 10 e5 ^h5
tance of the sacrifice with 5...h6 is 11 fxg4 ^xf4 12 «f3 '^g7 and White’s
critical. Then after 6 ^xf7 ‘^xf7 play had been refuted in Gunsberg-
Bird, London 1889.
see following diagram_
c) 7 ^c3!? (perhaps the best try)
White has a choice of three follow-up 7.. .d5 8 d4 f3 9 ^xd5 10 ^xf6 is
moves: Bosboom-Teichmann, Ramsgate 1984,
a) 7 .^c4-(- (this is the normal move and now 10...#xf6 looks at least equal
but, judging from the following varia¬ for Black, as 11 e5 #f5 12 .^d3 fails to
tion, it seems bad for White) 7...d5 12.. .fxg2 13 Sgl «f3!

51
The King’s Gambit

6 exd5 h6 7 #62+ ±e7 8 ^eA- f5 queens is now forced. However, this

exchange doesn’t necessarily mean the
end to White’s attack in the King’s
Gambit.
13 lfxe8+?3xe8
The alternative was 13...'^xe8,
when if 14 ^b5? ^xd5 15 c4 c6 wins
for Black. However, White can do
better with 14 ±xf4 ±d6?! 15 ±b5-l-!
'^d8 (if 15...±d7 16 0-0) 16 0-0 with
good play. Therefore, Black should
answer 14 .^xf4 with 14...'^d8. Black
has an extra piece, but White has an¬
9 -S^bcS!? noying pressure and can slowly build
A spirited approach, but can it up his game with 0-0-0, .^d3 etc.
really be sound? White gives up a 14 i.xf4 i.d6
piece to disturb the black king, but Perhaps Bronstein believed that this
Black has a lead in development and refuted the attack, as 15 .^xd6-(- 4lxd6
the white rooks are far away. There is is hopeless for White. But Neffe finds
no good reason why Black should an elegant reply.
suddenly find himself mated. 15 i.d3! i.xf4 16 0-0
The pusillanimous move 9 ^f2 was Now White regains one piece and
played in another Bronstein game. In maintains his initiative.
Duz-Khotimirsky-Bronstein, Moscow 16...*g7 17 axf4 Sf8 18 5le2!
1954, White was soon defeated after
9.. .^f6 10 d4 0-0 11 .^xf4 ^xd5 12
Wd2 ^xf4 13 Wxf4 '^gZ 14 ±e2 Se8
15 0-0 ±d6 16 «d2 Wxh.4.
A lot of history (chess and other¬
wise) is evoked by these two games.
Duz-Khotimirsky, whose best years
were before the Russian Revolution,
beat both Lasker and Rubinstein at St
Petersburg in 1909; and he was 75
years old when he played the 30-year-
old Bronstein. Bronstein was 70 him¬
self when he faced the youthful Neffe Another unexpected move. The
in Wrexham. How times change! knight joins in the action. Soon every
9.. .fxe4 10 'txe4 11 Wg6+ <i18 white piece is attacking the black king
12 dA'VleSl and there are few defenders in sight!
Just in time before White plays 13 18...c6!
.^xf4, attacking h6. The exchange of After fifty years of international

52
 other Gambits after 3 fhf3 g5 and 3...fhc6

chess, Bronstein has a calm head in a Hanstein Gambits, so why enter dan¬
crisis. He understands that instead of gerous sacrificial variations?
trying to rush his queenside pieces 5h4
over to the king, he must weather the This move distinguishes the
coming storm by breaking White’s Philidor from the Hanstein 5 0-0 (see
hold on the centre. the next game).
19 Safi Sxf4 20 4ixf4 cxd5 21 5...h66d4d6
‘Sih5+ *g8 22 i.g6 -Side 23 Sf6 fheA- I have changed the move order here
24Sf7 for the sake of clarity. In fact the game
According to analysis by Nigel began as a Fischer Variation: 3...d6
Davies, this is a mistake. He claims and after 4 l.c4(?) h6 5 d4 g5 6 h4 l.g7
that White should play 24 Sf4, plan¬ transposed to the Philidor. White
ning 25 and 26 .^xd5, when could (and objectively should) avoid
White has three pawns and an attack this line. This is easily done: after the
for the piece. So Black’s best reply Fischer 3...d6 play 4 d4 g5 5 h4!, not
would be 24...^d6, when 25 Sf6 ^e4 giving Black time to solidify his king-
draws by repetition. side with ...h7-h6 and ....^g7. And af¬
24...-Sice 25 Sc7 ‘Sixd4 26 Sg7-H ter 3...g5, play 4 h4 g4 5 ^e5 with a
*f8 27 Sf7-H *g8 28 Sg7-H *f8 29 Kieseritzky. The point is to oblige
Sf7+ *g8 Black to play ...g5-g4 immediately. If
Bronstein is happy to take the draw. you fail to force Black to weaken him¬
Instead he would have had winning self with ...g5-g4 then there is no hope
chances by running to the queenside for an advantage. In fact, as we shall
with 29...'i’e8! White has no good way see, it is Black who normally gets a
to exploit the discovered check. Nev¬ stronger attack.
ertheless, Davies believes that White 7 0-0!?
would have reasonable practical
chances after 30 ^f4 ^d6 31 Hh7-(-
■^fg 32 h5 etc.
30 Sg7+ 34-72
A highly interesting game.

Game 19
Yoos-Hjartarson
Reykjavik 1996

1 e4 e5 2 f4 exf4 3 <af3 g5 4 i.c4

i.g7!
Personally I think that Black does This is better than the old line 7 c3
well to avoid the complexities of the 4ic6 8 14153? (8 0-0 would transpose to
Muzio (Games 21-23) after 4...g4 5 0-0 the main game), when Black has an
etc. He is better in the Philidor or undoubted advantage after 8...#e7!

53
The King’s Gambit

Zak and Korchnoi analyse 9 0-0 ^f6 complications, as otherwise White

10 hxg5 hxg5 11 ^xg5 (the only con¬ simply takes on f4 with a good game)
sistent move) and now ll...^xd4! is 11 ±x{7+ ^xl7 12 «c4+ ±e6 13
very strong, e.g. 12 ^x{7+ '^dS 13 #xb4 f3 14 gxf3 gxf3 15 Sfl ^c6 16
cxd4 ^xe4 and both 14 ^f3 ^xd4-t- 15 Sxf3+ '^g8 17 «xb7 ^xd4 18 ^xd4
^xd4 #h4 and 14 ^xf4 ^xd4-(- 15 ±xd4 19 Sg3-i- '^h7 20 e5!? and now
^e3 ^xe3-(- 16 #xe3 ^xg5 are hope¬ 20.. .d5 looks best with obscure play.
less for White. If this were not Was Pavlovic bluffing, or did he have
enough, Black also has a decisive at¬ a new idea against 7...^c6? Had Tuk-
tack after the more mundane makov forgotten the theoretical rec¬
11.. .^xe4, e.g. 12 ^xf7 Sh7! 13 Sxf4 ommendation?
^xd4! 14 mi ^g3. b) 7 ^c3 has the obvious drawback
Apart from 7 0-0!? and 7 c3 two that the d4-pawn can no longer be
other moves are possible: supported with c2-c3. This is an espe¬
a) 7 #d3. This is dismissed by the¬ cially risky way for White to play. We
ory because of the simple developing have to delve in the archives to find an
7.. .^c6! Now after 8 hxg5 hxg5 9 example: 7...^c6 8 ^e2 #e7 9 #d3
flxhS ±xh8 10 e5 (threatening 11 ±d7 10 ±d2 0-0-0 11 ±c3 (Zak points
Wh7; the whole idea of 7 #d3) out that Keres’ suggestion 11 0-0-0 fails
10.. ..^g7! leaves White with no good to ll...C^f6!, when 12 hxg5 is met by
way to continue his attack since he is 12.. .^xe4) Se8 12 d5 (not a pretty
behind in development, e.g. 11 ^c3 move to have to make) 12...‘53e5 13
^h6 12 exd6 cxd6 13 ^d5 '^f8 14 ^xe5 dxe5 14 0-0-0 4116 and White
^xg5 ’Iii^xg5 15 .^xf4 #h4-(- and White had negligible compensation for the
has little compensation for the piece, pawn in Anderssen-Neumann, 1866.
Rosenthai-Neumann, 1869. Equally 7.. .‘Slc6 8 c3 -Slfe
good is 10...'^f8!, when after 11 #h7 If 8....^g4!? 9 #d3!? with unclear
±g7 12 #h5 ^h6 13 exd6 the piece play - Yoos.
sacrifice 13...^xd4! 14 ^xd4 .^g4! 15 An important question is whether
#h2 #xd6 proved decisive in the after 8...#e7 White is obliged to
game Remakulus-Brglez, Correspon¬ transpose into the note at move seven
dence 1983, as White’s king is trapped above with 9 #b3. Perhaps 9 #d3 is
in the centre. White resigned after 16 better, keeping the queen involved in
^e2 Se8 17 ^d2 ^f5! It is easy to the defence of the centre.
work out that there is no defence to A final possibility is 8...g4!? Accord¬
18.. .<53g3 or 18...<53d4. ing to Estrin White gets the advantage
This all looks very convincing, but after 9 ^el! f3 10 gxf3 #xh4 11 f4 g3
7 #d3 was repeated in the game Pav- 12 C^f3. This verdict was challenged in
lovic-Tukmakov, Lugano 1986. That the game Hughes-J.Littlewood, Eng¬
game continued 7...g4 8 ^gl ^c6 9 land 1992, which continued 12...#h5
^e2 ^ge7 10 ^bc3 (this looks better 13 f5 ^f6 14 mi g2!? 15 '^xg2 Sg8 16
than 10 .^xf4 d5) 10...^b4 (stirring up <^12 .^d7, and after 17...0-0-0 Black

54
 other Gambits after 3 fhf3 g5 and 3...fhc6

had good play. Even better for Black is 10...fxg6

14...d5!, as after 15 exd5+ ^e7 16 Sel
0-0! White is in deep trouble -
J.Littlewood.
9 hxg5 <2^h5!?

11 ihh27\
Hjartarson gives 11 <53bd2 as un¬
clear. However, Black has a sound
extra pawn, a wedge on f4 and argua¬
Yoos had planned to answer bly the safer king. And what is
9.. .<2ixe4 with 10 .^d5!?, when White’s plan? An attempted break¬
10.. ‘2ixg5 11 .^xf4 is unclear, rather through with e4-e5 would lead, after
than follow the ECO recommenda¬ the exchange ...d6xe5; d4xe5, to the
tion of 10 Sel d5 11 .^d3 hxg5 12 weakening of the a7-gl diagonal,
.^xe4 dxe4 13 Sxe4-i- '^fS, which which would put White’s king in
looks bad for White. Unfortunately peril. White’s compensation rests in
for him Black got his novelty in first! the possibility of gaining space on the
10 g6l? queenside with b2-b4 etc., and the fact
An interesting sacrifice. White gives that he can respond to the develop¬
up the pawn in such a way that the h- ment of the bishop on c8 with ®b3,
file remains blocked and his king is hitting both b7 and threatening .if7-i-.
therefore safe from attack by the rook However, the plan of b2-b4 can be
on h8. After Black’s reply the scope of met in similar fashion to the game,
the bishop on c4 is increased and while Black can prepare the develop¬
Black can no longer spirit his king ment of his bishop with ll...'#e7. In
away to safety on the kingside. How¬ all. Black’s chances must be preferred.
ever, 10 g6 also straightens out Black’s 11.. .Ef8 12 b4
wrecked kingside pawn structure, so it Switching play to the queenside. In
is not a natural move. Nevertheless, it his earlier calculations. White had
is difficult to suggest an alternative as probably thought that he could play
after 10 gxh6 Sxh6 11 Wb3 ®d7 12 12 .^e2 here, missing the combination
•^igS? ^xd4! 13 cxd4 .^xd4-i- 14 Sf2 11.. .^g3 13 Exf4 ^xd4! 14 Sxf8-i-
^g3 Black’s attack wins (variation by '^xf8 15 cxd4 .^xd4-i- and wins by
Hjart arson). forking on e2.

55
The King’s Gambit

^2...a6 13 a4 i.d7 14 i.a3 ihg3 15 open the e-file to force the win. How¬
Sel «h4 ever, this proves none too easy.
Now Black’s attack on the kingside 18.. .'i'h5!
begins to look dangerous, so White Hjartarson is an excellent defender.
gambles on a quick breakthrough in Of course, the exchange of queens is
the centre. anathema to White so Black gains
16 e5! dxe5 17 b5 time to bolster e5.
19 «b3 Ef5! 20 <23df3?
White is nonplussed by his oppo¬
nent’s defence. Flexible thinking was
required. Since the e5-square is heavily
fortified. White should have looked to
an easier target. There are few black
defenders on the queenside, so 20 bxa6
bxa6 21 Wb7 was correct, when after
21.. .Ec8 22 .^xa6 the passed pawn be¬
comes the most important feature of
the position.
20.. .e4! 21 i.e6?
17.. .<2^e7? It still wasn’t too late for 21 bxa6
Black misses 17...f3!, e.g. 18 <53x13 bxa6 22 Wb7.
Wh\+ 19 ^12 <53e4+! 20 ^e3 l'xg2 21 21.. .axb5! 22 i.xe7 *xe7 23 i.xf5
±xf8 (21 <i>xe4 #12!) 2l...l'f2+ 22 «xf5
‘i’d3 <53g3! and in view of the threat of Now the mobile black centre
23.. ..^f5+ White is in deep trouble - pawns, two bishops and the ridiculous
Hjartarson. However, it was by no white knight on h2 give Black a strong
means easy to see this variation during initiative.
the game, and even at the end White 24 <23d2 e3 25 <23df1 <23e4 26 <23f3 g5
can still complicate with 23 .^e6!? 27 <53xe3 fxe3 28 Sxe3 i.e6 29
(clearing c4 as an escape route for the *54+ *d8 30 Eael Exa4 31 *b2?
king) 23....^xe6 24 .^xg7. Hjartarson’s The last chance was 31 Wbl, but
blunder reminds us that the King’s 31.. .1.d5 32 Exe4 i.xe4 33 Exe4 Sal
Gambit experience is an unpleasant 34 *xal *xe4 35 *a8+ ^c7 36 *g8
one even for strong grandmasters. .^f6 wins for Black (Hjartarson).
Even if a line is theoretically bad it can 31.. .Ea2 32*b1 Exg2+ 0-1
still work wonders in practice against
a surprised, bewildered or complacent Game 20
opponent. Jonkman-L.B.Hansen
18 <23d2 Wijk aan Zee 1994
Suddenly White has an excellent
position: the enemy king is trapped in 1 e4 e5 2 f4 exf4 3 ^f3 g5 4 Jic4
the centre and he only has to break i.g7 5 0-0

56
 other Gambits after 3 thf3 gS and 3...fhc6

White avoids 5 h4. His king will

now be safer on the kingside, but on
the other hand so will the black king.
5...d6 6d4 h6

pieces undisturbed he soon has the

better game, e.g. 8 ‘53a3 ‘53f6! 9 Wd3 0-0
7 c3 10 ^62 d5! 11 exd5 ^xd5 12 i.b3
White strengthens the d4-square. <53de7 13 Sael .^f5 (Chigorin). There¬
The alternatives are equally unpromis¬ fore White has to do something active.
ing; But what?
a) 7 .^e6!.^ 8 .^xe6 fxe6 9 e5 First, he could consider attacking
<53c6! and Black gains a clear advantage the f4-pawn with 8 g3. However, this
by undermining the white centre rebounds after 8....^h3!, e.g. 9 gxf4
(variation by Rabinovich). Wd7! (this is much better than seizing
b) 7 g3 (this attempt to break up the the exchange with 9....^xfl, when all
black kingside fails as White’s centre is the dynamism disappears from Black’s
unstable; in fact, it is the white king- position - rapid development and an
side which is more fragile) 7...J.h3 8 attack on White’s centre and kingside
Sf2 <53c6! (counterattacking against d4; is called for!) 10 Sf2 ^f6 11 Wei 0-0-0
if now 9 gxf4 g4 etc.) 9 .^b5 ‘53f6! 12 .^bS She8 (completing the mobili¬
(developing with an eye on e4) 10 d5 sation of all the black pieces; now e4 is
a6 (Estrin) and Black has a big advan¬ coming under fatal pressure) 13 ‘53bd2
tage after 11 .^a4 b5 or 11 dxc6 axb5 gxf4 14 '^hl <53xe4! 15 <53xe4 d5 and
as the white centre has lost all its cohe¬ Black conquers the centre and thus
sion. gains a clear advantage (analysis by
7...ihc6\ Glaskov).
The careless 7...4116? would allow 8 Second, White could try and attack
e5! dxe5 9 ^xe5 0-0 10 1^53, when the g5-pawn with 8 h4. This trans¬
White threatens both 11 <53xf7 and 11 poses to the Philidor game above, in
<53g6 and 7...‘53e7 is also inaccurate, as 8 which 8...‘53f6! proved good for Black.
g3 g4 9 ^h4 f3 10 ^a3 0-0 11 i.f4 A third option is 8 ®b3, attacking
gives White an initiative for his pawn. f7. This is well answered by 8...''i'd7!,

__ —
The King's Gambit

planning 9...4ia5 to deprive White of ening 21 4lfe5 or 21 4lxb6. After

his good bishop. 20.. ..^xf3 21 4lxb6 looks good for
Since the game move also proves in¬ White, e.g. 21....^d4-h 22 cxd4 Wxd4-H
sufficient, it seems that Black has a 23 S^hl ^xg2+ 24 S&xg2 Wxe4+ 25
least a small advantage in the diagram Sf3 g4 26 Sell Wxel 27 ^xc6+ S&f8 28
position. White’s misfortunes can be Sxf4 and White should win. Or if
traced all the way back to the fourth 20.. .Wc7 21 <53xb6! is strong.
move, when he failed to undermine Probably Black should answer the
the black kingside with 4 h4! pawn thrust with 18...0-0, but then
8.. .«e7?! after 19 b7 the passed pawn gives
The immediate 8....^g4, planning White compensation for his material
9.. Md7, may save a tempo on the and positional deficits.
game continuation. In any case, this line was White’s
9 b5 10 i.d3 i.g4 11 i.a3 ITd? only chance, as the game continuation
See the note to Black’s eighth move. is hopeless.
The queen moves out of the way of a Note that if Black hadn’t squan¬
threatened pin after 12 e5. dered a tempo with 8...We7 and
12 <2^bd2 ihe7 13 'tfa4 b6 14 i.b4 11.. .Wd7 he would have already had
ihbl 15«a6 <2^c5?! time to castle kingside before 15 Wa6.
A very logical move. Black makes a Then 15...^c5! really would have
pseudo-sacrifice of his worst placed been crushing. Therefore, the possibil¬
piece to wreck the white centre and ity of 18 b6 doesn’t change the verdict
unleash the bishop on g7. However, that the Hanstein is a poor choice for
there is a tactical drawback to this White.
move due to the fact that Black’s king 18.. .0.0
is still in the centre. No doubt Black was relieved to
16 dxc5 dxc5 17 .^xc5 bxc5 play this move!
19 *33 c6! 20 *xc5 Eac8!
White has temporarily regained his
pawn, but there is no good way to
defend c3.
21 b6 axb6 22 *xb6 Jixc3
The dust has cleared and Black is a
pawn up with a strong pair of bishops.
Hansen’s technique now makes short
work of his opponent.
23 Sadi &cd8 24 *b3 ^g7 25 e5
^d5 26 h3 *a7-l- 27 *h1 i.f5 28
i.xd5 cxd5 29 *b2 SaS 30 Sal f6 31
18 i.c4? *b3 fxe5 32 *xd5-h *f7 33 *xf7-h
White should have tried 18 b6!, e.g. Sxf7 34 ^b3 i.d3 35 Sfel e4 36 Sadi
18...cxb6 19 i.b5 ^c6 20 ^c4, threat¬ 0-1

58
 other Gambits after 3 fhf3 gS and 3...fi^c6

White resigned as he loses after Gambit) 7 Wxf3 ^f6 8 Wxf4 ktl 9

36...i.c2 37 Sd2 ^xb3 38 axb3 exf3. ^c3 0-0 10 d3 c6 11 i.b3 i.e6 12 i.d2
i.xb3 13 axb3 ^bd7 14 Sf3 White
Game 21 built up a strong attack in Auerbach-
Chigorin-Davidov Spielmann, Abbazia 1912.
St Petersburg 1874 6 irxf3 irte
6....^h6 is dubious as Black does
1 e4 e5 2 f4 exf4 3 g5 4 i.c4 nothing to oppose White’s strong cen¬
g4 5 0-0 tre, e.g. 7 d4 Wf6 8 e5 Wf5 9 ^c3 and
The Muzio Gambit, where White White has good chances.
frequently jettisons a huge amount of Another suspect try is 6...We7 7 d4
material, even by King’s Gambit stan¬ <53c6.
dards.

Now the most accurate move is 8

5...gxf3 <5303, when in view of the threat of 9
Here Black has also tried 5...d5, <53d5 Black is virtually forced into
which looks logical as this freeing 8.. .‘53xd4, after which 9 Wd3 ‘53e6 10
move is often the antidote to white <53d5 Mc5+ 11 '^hl b5 12 i.b3 i.h6 13
sacrifices in the King’s Gambit. How¬ .^d2 etc. gave White a dangerous ini¬
ever, after 6 .^xd5 the black defensive tiative in Steinitz-Anderssen, London
idea ...Wf6 is no longer effective, as 1862. In a more recent game. White
there is no threat to win an unde¬ preferred 8 .^xf4, but this was refuted
fended bishop on c4 with ...Wd4-i-. by some cold-blooded defence:
Therefore, Black has to develop nor¬ 8.. .<53xd4 9 Wd3 ^e6 10 ^c3 kg? 11
mally, which means that White isn’t <53d5 <53xf4!! 12 Sxf4 (or 12 ^xe7
compelled to stake everything on a <53xd3 and Black wins) l2...We5! 13 c3
second sacrifice on f7. With two (also hopeless for White are 13 Sxf7
pawns for the piece. White can play in '^xf7 14 <53xc7-h '^e7 15 <53xa8 Wd4-i-
positional style. After 6...gxf3 (6...c6? 7 and 13 %3 Wd4+ 14 S^hl ke5 15
.^xf7-h '^xf7 8 <53e5-h gives White a ^c7-i- kxc7 16 kx{7+ S^dS 17 kxgS
very superior version of the Lolli i.xf4 18 Wh4-h '^c7 19 Wxf4-i- d6)

59
The King’s Gambit

13.. .^h6 14 Safi 0-0 15 ^hl d6 16 Games 22 and 23.

‘Slf6-i- '^hS. Here White had no com¬ 8.. .1.h6 9 ^c3
pensation for the piece in Friedman- 9 J.d2 <53e7 10 <5303 transposes to
Shipman, Chicago 1989. the game (10 .^c3? Wc5-h).
7 e5 9.. .<23e7 10 i.d2 <23bc6!
The slow 7 d3?! is not to be rec¬
ommended as Black is given time to
organise his defences. Morais-Boino,
Portugal 1993, continued 7...J.h6 8
^c3 ^e7 9 e5 WxeS 10 i.xf7-i-!? ^d8!
(the acceptance of the offer with
10.. .'^xf7 gives White some attacking
chances, e.g. 11 .^xf4 .^xf4 12 <53e2!
<53bc6 [or 12...<53g6 13 <53xf4 <53xf4 14
Wg3] 13 <53xf4, planning Sael etc.,
after Black has side-stepped the discov¬
ered check on the f-file. As we shall see
in the main game, the best place for The black king will be safer on d8
Black’s king in this variation is d8. than on the kingside. Hence 10...0-0?
White has therefore lost time by driv¬ would be a grave mistake, giving
ing it there) 11 l.d2 <53bc6 12 Sael White a strong attack after 11 Sael
®f6 (once again Black is helped by the Wc5-i- 12 '^hl followed by ‘53e4, .fi.c3
bishop on f7, which is now attacked etc.
and so must move again, thereby los¬ 11 Sael Iff5 12 ^d5 *d8 13 i.c3
ing more valuable time) 13 .^b3 d6 It seems that White has no promis¬
and White had little or no compensa¬ ing continuation. A key position is
tion for the piece. reached after the alternative 13 We2
b5!

7...irxe5 8 d3
This doesn’t seem sufficient even Now White has two ways to pursue
for equality. 8 .^xf7-i- is considered in his attack:

60
 other Gambits after 3 fhf3 gS and 3...fi)c6

a) 14 <2lxe7 ®c5+ (but not This adds another defender to e7

14.. .bxc4?? 15 <5^xc6+ when Black will through the bishop on f6.
be mated on e7) 15 '^hl Wxe7. Black 15 g4 ITge 16 i.xg5 «xg5 17 h4
retains his booty and should win by irxh4 18 «xf4 d6 19 <23f6 <23e5?
beating off the white attack, e.g. 16 Black could have won with 19...Ef8!
Wh5 Wg5 17 Wxf7 (if 17 i.c3 then 20 Ee2 i.f5! 21 gxf5 ®xf6 - Rabi¬
17.. .5.8!! 18 h4 Wg6! refutes White’s novich. However, this game was
play) 17...bxc4 18 ^xf4 (if 18 ^c3 Sf8 played in an age when the King’s
19 i.f6+ Wxf6 20 Se8+ Sxe8 21 Gambit usually led to spectacular vic¬
itxf6+ ^e7 22 Wxh6 cxd3 23 cxd3 tories for White.
Sb8 and Black’s material advantage 20 Exe5! dxe5 21 WxeS i.xg4?
should be decisive - Zak) 18...Wg6 19 Black could have held the draw
i.xh6 Wxf7 20 Sxf7 i.a6 and Black with 2l...i.e6! 22 WM+ ^d5 23 i.xd5
wins. Wg3+ 24 i.g2-H Wd6 25 Wxd6+ cxd6
b) 14 .^xf4 .^xf4 (if 14...‘53xd5? 15 26 <53xe8 ‘i’xe8 according to Golom-
.^xli6! and White has a winning at¬ bek and Cafferty, e.g. 27 .^xb7 Eb8 28
tack) 15 Sxf4 Wg5 (if 15...We6 then 16 .^c6-i- '^f8 29 b3 Sc8 picking up the
Wf2 maintains the attack, while c2-pawn.
15.. .Wxf4 16 <53xf4 bxc4 17 dxc4 is un¬ In the game White now wins in
clear) 16 ^xe7 ^xe7 17 Sxf7 bxc4 18 style.
Sxe7 Wg6 and Black wins as the white 22 ird4-i- *c8
attack cannot be strengthened.
13.. .Ee8
Both 13...Eg8 and 13...Ef8 also look
good for Black.
14i.f6
Instead 14 Wc2 was played in
Keene-Pfleger, Montilla 1974, and a
draw by repetition was agreed after
14.. .We6? 15 Wf3 Wf5 16 We2. Instead
Keene gives 14...d6! 15 <5316 (15 .^f6
i.e6) 15...Ef8 16 g4 Wg6 17 h4 as good
for White, but Black wins after
17.. ..^xg4 18 Wxg4 (18 <53xg4 Eg8 19 23 i.e6-H!
S^hl Wxg4) 18...Wxg4-i- 19 ^xg4 Eg8 A beautiful move which exploits
20 .^f6 Exg4-i- 21 '^f2 Eg6 etc. the pin on the black bishop to win
White’s other try is 14 <53f6, but ac¬ control of the crucial d7-square. Such a
cording to Zukertort Black wins after move is difficult to see rather than to
14.. .5.8 15 g4 Wg6 16 h4 d5 17 i.xd5 calculate, as it is not often a good idea
i.xg4 18 Wxg4 Wxg4-H 19 ^xg4 Eg8 to put a bishop en prise on a square
20 i.f3 f5 21 i.f6 S&d7! which is heavily defended. Of course,
14.. .1.g5! the Fritz program took less than a

61
 The King’s Gambit

second to find this move: a computer 22 #f3-h and White wins) 14 J.xf6
has no human prejudices! .^xf6 15 ‘53d5 #xg4-i- 16 #xg4 .^xg4
23...*b8 24 ^d7+ *c8 25 <2^c5+ 17 ^xf6 (or 17 Sxf6-i- s£g7 18 Sf4 [18
‘£>58 26 ^a6+! bxa6 27 #54 mate Safi Sg8! 19 <53xc7 <53d7 is very good
1-0 for Black] 18...1.h5 with a small ad¬
vantage to Black - Korchnoi) 17....^h3
Game 22 (Korchnoi gives 17...<53c6 as equal) 18
Leisebein-Baer Sf3 ^6 19 ^d5-i- i.f5! 20 Sxf5-i- s£e6
Correspondence 1996 21 Sh5 Sag8-i- 22 s£hl Sg7 23 ^xc7-i-
Sxc7 24 65+ with equality according
1 e4 e5 2 f4 exf4 3 ^f3 g5 4 i.c4 to Sapi and Schneider. Perhaps White
g4 5 0-0 gxf3 6 #xf3 #f6 7 e5 can even claim a small advantage in
#xe5 8 i.xf7-h ^xf7 9 d4 #xd4+ the endgame?
b) 12...d6! (this looks best; Black
immediately returns the knight on f6
to gain counterplay along the g-file) 13
.^xf6 .^xg4 14 #g2 Sg8 (threatening
15.. .1.f3!, but not 14...i.g7? 15 ^g5+
S£g8 16 #xg4) 15 s£hl i.f5 16 Wd5+
(Estrin stops his analysis here and
claims that White is slightly better)
16.. .s£xf6 17 ^3.

Here 9...#f5! is a major alternative

and may be the only playable move! A
critical position is reached after 10 g4
#g6 (not 10...#f6 when after 11 J.xf4
d6 12 4lc3 there is no good answer to
13 ^d5 or 13 ^e4) 11 i.xf4 ^f6 12
^e5 and now:
a) 12...^e7?! was analysed by Sapi
and Schneider in the BCM, September
1988. Their analysis went 13 ‘53c3 d6 At first glance this position appears
(instead 13...'£>g8.^ leads to an over¬ to be uncomfortable for Black, as after
whelming white attack: 14 Sael d6 15 17.. .<53c6 (most other moves, e.g.
i.xf6 i.xf6 16 ^d5 ^d7 17 <53xc7 Sb8 17.. .C6 or 17...<53d7 or 17...1.h6 meet
18 Wd5+ '£>g7 [18...#f7 allows the with the same response) 18 Sxf5-h!
pretty finish 19 Se8-h <53f8 20 fixf6! #xf5? 19 ^e4-i-! s£g6 20 Sgl-i- wins
#xd5 21 Sexf8-i- s£g7 22 ^e8 mate] 19 the black queen. However, Black can
<53e6-i- ^h6 20 Sxf6 <53xf6 21 g5-i- '£>h5 jettison the bishop on f5 and emerge

62
 other Gambits after 3 ti^f3 gS and 3...^c6

with good chances after 18...'^e7! Ae7 18 Axc7 ^c6 19 ^ d6 20 i.xd6

Then White has some attacking ^d8 21 Axe7+ ^cxe7 22 Sd6-i- ±d7
chances for the piece after 19 Sel+, 23 Wh3 We8 24 4lf6 and wins) 17 Safi
but I doubt very much if it is enough Ag7 18 Sf7 Axe5 19 Wxe5 ^bc6 20
to save the game. Wxh8! and White wins, as in Glaskov-
10i.e3irf6 11 ^c3 Muratov, USSR 1973.
This transposes to 11 .^xf4 4le7 12 b) 13...d6 14 Ag5 ^bc6 15 g4 Sg8!?
^c3 lines after Black’s reply, but it (15...Wxfl-i- 16 Sxfl-i- ^e8 17 ^d5
gives Black two extra ways to go looks overwhelming) 16 h4! Sxg5 17
wrong. hxg5 Wxfl-i- 18 Sxfl-i- '^g7 19 WB
n...^e7l gives White a strong attack.
It was bad to accept the third piece, 13 Eael Ag7 14 ^4 Wf5 15 <2^d6!
as ll...fxe3? 12 Wh5+ '^g7 13 Sxf6 1-0
^xf6 14 %5+ ^f7 15 Sfl Ae7 16 Leisebein gives the variation
^d5 would be a massacre. Also bad 15...cxd6 16 Sxe7 ^c6 17 Sxg7-i- ^xg7
was ll...d6? 12 4ld5 Wf5 13 g4! Wxg4+ 18 Ah6+ ^xh6 19 Wxf5 and White
14 Wxg4 Axg4 15 fixf4+ and next wins.
move White either captures on c7 or
g4 with check. Game 23
12i.xf4*g8? Yoos-Kirton
This loses. The standard 12...<5^f5 Saskatoon 1994
and the inferior 12....^g7? are exam¬
ined in Game 23. However, judging 1 e4 e5 2 f4 exf4 3 ^f3 g5 4 i.c4
from the outcome of that game it may g4 5 0-0 gxf3 6 irxf3 *16 7 e5
be that after 9...Wxd4-i- Black already WxeS 8 i.xf7-i- *xf7 9 d4 irxd4+ 10
has a lost position! i.e3 ITf 6 11 i.xf4
12...Wf5? is also bad after 13 We2!
and now:

11...4le7
The alternative ll....^g7? seems
a) 13...<^e8 14 Ae5 We6 15 Sf6 Wg8 wholly bad. Two examples are:
16 Wh5+ ^d8 (or 16...<5^g6 17 Sel a) 12 Wh5+ %6 13 i.xc7-i- ^f6

63
The King’s Gambit

(13...1.f6 is the only chance) 14 Wd5+

^f8 15 i.d6+ ^e8 16 Sel+ <i’d8 17
We5 1-0 Lebedev-Normant, Corre¬
spondence 1987.
b) 12 ^c3 ^e7 13 ^d5 ^xd5 14
WxdS-i- We6 15 ^62+ S&g8 16 Sael
®xd5 (16...‘53c6 17 Sxe6 dxe6 fights
on) 17 Se8-i- .^f8 18 .^h6 and mates,
Smirnov-Tikhonov, USSR 1954.
12<2^c3<2^f5
If 12....^g7? then 13 ‘53d5 would
transpose to the Smirnov game in the
last note. a) 16...i.d7 17 gxf5 Wf6 18 Wh5-(-
^g8 (18...S&e7 19 ^d5-i-) 19 Sgl-i-
.^xgl 20 Sxgl-i- '^f8 21 .^h6-i- and
White wins.
b) 16...^d4 17 Wd5+ S&g7 18 i.xd6!
is strong as 18...Wxd6 (18....^xd6 19
Wxd4-f-) 19 Wi7+ S&h6 20 Sf6-i- wins.
c) 16...^h4 17 W<i5+ ^g7 18 i.cl2!
<53c6 19 J.C3-I- ‘53e5 20 .^xe5-i- dxe5 21
Wxe5-i- '^g8 22 Sf6! with an over¬
whelming attack
d) 16...^c6 17 gxf5 i.xf5 18 i.xd6
.^xd6 19 Wxf5-i- Wxf5 20 Sxf5-i- '^g6
13 i.e5!! 21 5f3 .^c5 22 <53e6 (22 Sg3-i- forces a
An incredible novelty in a well- draw by perpetual) 22....^b6 23 Safi
known position. The two known with dangerous threats to the black
moves are 13 ‘53e4 and 13 ‘^idS: king and a guaranteed draw with 24
a) 13 ‘53e4?! is supposed to fail, e.g. Sg3-h if he wants it.
13.. .Wg6 14 g4 i.e7 15 '^hl <53h4 16 However, it is hardly worth look¬
We3 '^g8 17 .^e5 b6! and the threat of ing at these variations if 13 .^e5 is as
18.. ..^b7 refutes White’s attack. strong as it appears to be.
b) 13 <53d5!? is a much better try. 13...irxe5
After 13...Wg6 Sapi and Schneider ana¬ The alternatives are no better;
lyse 14 Sael J.C5-I- 15 .^e3! as strong a) 13...i.c5-i- 14 S^hl Wxe5 15 Sael
for White. Also very interesting is 14 Wf6 16 Wh5+ ^g7 17 Sxf5 %6 18
<53xc7!, e.g. I4...i.c5-i-15 '^hl d6 16 g4. Sg5 wins the black queen and keeps a
huge attack.
see following diagram_
b) 13...'tb6-i- 14 S^hl d5 (if 14...d6
Now Black has a wide choice, but 15 Wh5+ S^gS 16 %5-i- <^f7 17 Sxf5-i-
everything seems bad for him: i.xf5 18 Wxf5-i- S&e8 19 Wc8-i- S&e7 20

64
 other Gambits after 3 thfS g5 and 3...^c6

^d5+ ‘4>f7 21 Sfl+ and mates quickly) Black has a rook and three pieces for
15 ^xh8 (possibly not the best) the queen - none of which are devel¬
15...Wc6 (this looks ridiculous, but oped, unlike the king! It is inconceiv¬
what else?) 16 ^xd5 ^a6 17 Sael ^g8 able that the black king will survive
18 ^c3 and wins. the attack of the queen and rook, e.g.
14 ^5+ *g8 17...<^g6 18 Sfl and 19 «f7 or 19
An important moment. The king #f5-t- will be decisive next move.
can advance forwards and defend the The only other move for the black
knight, but a massacre seems inevita¬ king is 14...'^g7, but this loses at once
ble: after 15 Wg5-h 'if7 16 Sxf5-h. It there¬
a) 14...<^e6 15 Sael ^e3 16 «f7-t- fore appears that Black is lost after 13
^d6 17 Sf6-h ^c5 18 h4+ ^d4 ^e5.

Black is defenceless, e.g. 15...#g7 16

Safi h6 17Sf7.
16 IfgS-t-Ifge

19 Sd6-i-!!
A real problem-like move, discov¬
ered by Fritz. The point is to clear the
f-file for the queen to check on f3.
Black is mated in one move after 17Sxf8-h!
19.. .'^xc3 20 Wb3 or in two moves This final sacrifice forces an imme¬
after I9...i.xd6 20 ^e2-t- ^e4 21 «f3 diate win.
or in three moves after the alternative 17.. .'1^x18 18 Sf1-h';^g8
19.. .«xd6 20 ^b5-t- ^e4 21 «f3-t- ^e5 If 18...'^e8 19 We5+ wins. More re¬
22 Sxe3. sistant was 18...'^g7, but 19 We5+ '^h6
Actually, Fritz tells me that Black (19...<^g8 20 WeT) 20 «xh8, intending
can struggle on to a mate in six by giv¬ 21 Sf6 or the crude 21 WxcS, is deci¬
ing up all his pieces with 19...^d5 20 sive.
Sxd5-t- «xd5 21 «xd5-t- etc. I9'te7! 1-0
b) 14...‘^f6 15 Sxf5-t-! (this is much If 19...«g7 20 «e8-t- mates next
better than 15 Sael, when 15...#d4-h move. A pretty game which could be
16 '^hl d5! is none too clear) the death knell for Black in the
15.. .«xf5 16 ^d5+ «xd5 17 «xd5. 9.. .#xd4-t- Muzio.

65
The King’s Gambit

move. After 7...^e6 Keres recom¬

Game 24 mends 8 d4!? c6 9 Axf4 cxd5 10 exd5
Leien-Marzec followed by 11 0-0. This second piece
Los Angeles 1991 sacrifice looks highly dangerous for
Black) 8 0-0 (bad for White is 8 «c3
1 e4 e5 2 f4 exf4 3 g5 4 ±c4 Wh4+ 9 ^fl i.c5) 8...i.d6 9 d4 ^xd4
g4 10 #h5 .^e6 11 .^xf4 .^xf4 12 ^xf4.

5 ±xf7+? Now 12....^xc4 led to unclear play

The Lolli Gambit, whereby White in Charousek-Marco, Vienna 1897,
offers the bishop rather than the after 13 We5+ ^f8 14 «xh8 i.xfl 15
knight. There are also some interest¬ Sxfl m 16 «xh7 «xf4 17 Sxf4
ing alternatives but, unfortunately, ^e2-t-. However, when I showed the
they all seem to end in total defeat for diagram position to the Fritz program
White: it came up with 12...^f3-t-!! which
a) 5 ^c3 is the McDonnell Gambit. seems to refute White’s play, e.g. 13
Black has to be wary. 5...gxf3 6 #xf3 #xf3 (or 13 Sxf3 Wd4+ 14 ‘^hl Axc4
and now: and Black is ready to castle queenside)
13...#d4-t- (the point is to rule out 14
Wc3) 14 ^hl i.xc4 15 Sadi (15 Sfdl
We5) 15...«e5 and White is lost.
a2) 6...d6 7 d4 (after 7 0-0, 7..Ae6 is
supposed to be a good defence for
Black. The advantage of 7 d4 first is
that 7...^e6 can be answered by 8 d5
and 9 .^xf4 with good compensation
for the piece - Keres) 7...^c6 8 Axf4.
This has transposed to Fedorov-
Adams, Game 25. In doing so. White
can be pleased that he has avoided
al) 6...d5!? 7 ^xd5 ^c6 (the best ...d7-d5! lines.

66
 other Gambits after 3 fhf3 g5 and 3...^c6

b) 5 d4 is the Ghulam Kassim At¬ material after 7...#xg4 8 ^xg4 d5! 9

tack. White’s wants an attacking set¬ exd5 ^d4 (Steinitz-Hruby, Vienna
up similar to that in Game 25, but, as 1882).
in the McDonnell Gambit above, a
quick ...d7-<l5 seems to give Black ex¬
cellent chances: 5...gxf3 6 WxB d5!
(this move is also the antidote to 6
Axf4 or 6 0-0) 7 .^xd5 ^f6 8 0-0 c6!

c2) 7 d4 ^xe5 8 dxe5 .^c5 9 .^xf7-t-

<^f8 10 We2 f3 11 gxf3 «h3-t- 12 <4>el
gxf3 is winning for Black (Bilguer).
c3) 7 i.xf7-t- '^e7 8 ^xc6-t- dxc6 9
i.xg8 axg8 10 Wei g3 11 d4 f3 12 h3
After 9 i.b3 WxdA+ White’s posi¬ .^g4 13 We3 Sg6 gave Black a winning
tion would be collapsing, so he has to attack in Dublin University-
throw more wood on the fire. How¬ Cambridge University, Correspon¬
ever, neither 9 .^xf7-t- '^xf7 10 #xf4 dence 1892.
i.g7 11 e5 af8 12 exf6 <^g8! (Zak) nor c4) 7 ^xf7 (the only challenging
9 ^c3 cxd5 10 excl5 i.g7 11 i.xf4 0-0 move) 7..Ac5 8 Wei g3 9 ^xh8
12 i.g5 <53bd7 13 ^e4 b5 14 a4 i.b7 15 10 Wdl ^f6 11 ±e2 (if 11 d4 d5 12
^xfe-K ^xf6 16 i.xf6 «xf6 17 «xf6 exd5 .^g4 13 .^e2 ^xd4 and Black has
i.xf6 18 Sxf6 b4 (ECO) offers White a winning attack - Csank) ll...d6 12
any hope. In the first variation he is a c3 .^g4 13 h3 (or 13 d4 0-0-0 14 ^f7
piece down, with Black’s king per¬ Sf8 picking up the knight with a
fectly safe; in the second, his vulner¬ strong initiative) 13...^e5 14 d4 f3 15
able pawns will soon be picked off by i.xf3 ^xf3 16 gxf3 g2-t- 17 ^e2 i.xf3-t-
the black pieces in the endgame. and Black wins, Goncarenko-
c) 5 ^e5 (the Salvio Gambit) Alekseev, Correspondence 1963, as 18
5...#h4-i- 6 ‘^fl ^c6! This move has <^xf3 gxhlW-t-19 «xhl «xe4-t- skew¬
been known for more than a hundred ers the white queen.
years and seems to refute White’s idea: As we shall soon see, the Lolli
Gambit is also inadequate. This means
see following diagram_
that the only way for White to get
cl) The great World Champion reasonable chances is with the main
Steinitz once played 7 #xg4, losing line Muzio 5 0-0!

67
The King’s Gambit

5.. .*xf7 6 ^e5+ *e8 7 'txg4 ^f6 8 seems to be sufficient for equality,
lfxf4 d6 9 ^f3 though probably no more: 8 #xh4
The knight has to retreat, demon¬ gxhlW 9 ^c3 and now:
strating that White’s sacrifice has
failed. The consistent 9 0-0 simply
leads to a lost position, e.g. 9...dxe5 10
Wxe5+ ^{7 11 Wc3 ^c6 12 e5 «d4+
13 #xd4 ^xd4 14 Sxf6-t- '^gS and
Black wins.
9.. .«e7
The simple move 9...Sg8, threaten¬
ing 10...Sg4, was a very strong alterna¬
tive.
10 ^c3 <S^c6 11 0-0 ag8 12 <S^d5
<^xd5 13 exdS Sg4 14 #xg4
Here 14 We3 #xe3-t- leaves White a) 9...^c6 10 Wh5 ^d8 (Black
with a lost endgame, but the rest is a should try 10...^xe5, though after 11
massacre. «xe5-t- ^e7 12 «xh8 «xh2 13 ±e3
14.. ..^xg4 15 Sel <S^e5 16 <S^xe5 White is better - Schmid) 11 .^gSi? (11
dxe5 17 d4 <;^d7 18 dxeS WcS-h 19 ^f2 ^f6 12 Wh4 ag8 13 «xf6 Wxh2+
.^e3 WxdS 20 h3 .^c5 21 hxg4 14 ^el i.g7 15 «f4 «xf4 16 i.xf4 d6
i.xe3+ 22 Sxe3 WcS 23 Sael Se8 is clearly good for Black) 11....^e7 12
24 <;^h2 1^X02 25 e6+ <;^c8 26 Sle2 0-0-0 and White has a dangerous initia¬
^6 27 e7 '»xg4 28 af3 Sxe7 29 tive.
axe7 '»h4-t- 0-1 b) 9...i.b4 10 ^f7 i.xc3+
(10...<&£f7 11 «h5-t- ^f8 12 Wf5-t- ^g7
Game 25 13 «g5-t- ^f8 14 «f5-t- ^e8 15 «e5+
Fedorov-Adams with a draw) 11 bxc3 '^xf7 12 #h5-i-
European Team Ch., Pula 1997 and White has at least a draw.
c) 9...d6! 10 ^xf7 and:
1 e4 e5 2 f4 exf4 3 <^f3 g5 4 d4 cl) 10...<^xf7 11 «h5-t- <^g7 12
The Rosentreter Gambit. #g5-t- (12 '^f2, threatening to trap the
4...g4 5 .^xf4!? queen with 13 .^g2, also deserves at¬
This leads to play similar to that of tention. Then 12...h6!! 13 .^g2? ^f6 is
Game 26 below. In fact the transposi¬ the incredible suggestion of Fritz)
tional possibilities are pretty bewilder¬ 12.. .'^f7 13 Wh5-t- '^g7 with a draw, as
ing! 13.. .'^e7 14 .^g5-t- is highly dangerous
The alternative was 5 ^e5 Wh4-t- 6 for Black.
g3 fxg3 7 #xg4 c2) 10...i.e7 11 Wh5 ^f6 12 ^xd6-t-
'^d8 (12...‘^d7? allows mate in six: 13
see following diagram
Wf5-t- ^c6 14 d5-t- ^xd6 15 ^b5-t- ^c5
Now after 7...g2-l-? White’s initiative 16 «f2-t- ^b4 17 i.d2-t- ^a4 18 b3

68
 other Gambits after 3 fhf3 gS and 3...‘^c6

mate!) 13 ^f7+ with perpetual check 7 <S^c3 <S^c6 8 .^c4

(Levenfish quoted in Zak). Now we have a reached a position
However, Black can avoid all these from Mortazavi-Miles (see the note at
variations with the less greedy move seven to Game 26 below), but
7..Mxg4!, when after 8 ^xg4 d5 9 with the moves Axf4 and ...d7-d6
^3e3 dxe4 10 hxg3 ^c6 11 ^b5 Ad7 thrown in. White is planning 0-0 etc.
(ECO), White has some compensation to start an attack aimed principally at
for the pawn since Black’s structure f7, so Adams forces the exchange of
on the kingside is dislocated, but it is queens. However, according to Fe¬
not enough. dorov in Informator 69, Black could
have snatched the d-pawn: 8...^xd4! 9
i.xf7+ <^xf7 10 «h5-t- ^g7 11 0-0 ^f6
12 M6-t- ^g8 13 «g5-t- ^f7 14 «h5-t-
^e6 15 Wh.3+ ^e7 16 «h4 ^f5! and
Black should win. Of course, this
variation by no means exhausts all the
tactical resources available to White in
the position. I’m sure most players
would be too terrified to enter this
variation as Black, despite Fedorov’s
assurances (especially if they were
playing Fedorov). One possible im¬
5.. .gxf3 6#xf3 d6?! provement is the calm 11 0-0-0, attack¬
After 6...^c6, 7 d5 looks like a good ing the knight and with ideas of 12
answer (7...^d4 8 Wdi). ad3. Then ll...<5^e6? 12 ±e5+ would
However, according to theory be awkward for Black, while ll...^c6
Black can gain the advantage with 12 e5! keeps up the initiative. In any
6.. .d5! Then 7 exd5 ^f6 8 i.b5+ c6 9 case, it is no surprise that Black
Ae5 .^g7 10 dxc6 bxc6 11 .^xc6+ (11 ducked the challenge in the game.
0-0 is similar, e.g. 11...0-0 12 .^xc6 [12
i.d3? <53bd7 13 «xc6 ^xe5 14 dxe5
.^d7 15 Wnb ^g4 is very good for
Black] 12...<53xc6 13 «xc6 ±e6 14 «f3
^g4 15 .^xg7 '^xgT) ll...^xc6 12
mxc6+ ±d7 13 «f3 0-0 14 0-0 ^e8 is
better for Black (ECO). Play could
continue 15 .^xg7 ^xg7 16 c3 Wb6
etc., when in the middlegame the
black bishop will prove more valuable
than the three white pawns. I wonder
what improvement Fedorov had in
mind?

69
The King’s Gambit

^xf6 11 0-0 <S^xd4? 15...i.xb2!, e.g. 16 ^xc7-t- <^d7 17

Despite the exchange of queens, 4lxa8 Axal 18 4lb6-i- axb6 19 Sxal
White maintains a dangerous initiative (Fedorov).
with ideas of capturing on f7 and 16 c3 h5 17i.h4!
overrunning the black king’s defences
after 12 e5. Black therefore returns the
piece and submits to a worse endgame.
The critical variation is ll....^e7 12 e5
(both 12 i.h4 ^g4 and 12 ^b5 <^d8
13 e5 ^e4! are nothing for White)
12.. .dxe5 (Fedorov gives 12...^d7 13
i.xf7-h ‘^dS 14 Sadi with unclear
play) 13 dxe5 (13 AxeS ^xe5 14 dxe5
Ac5-h 15 ‘4'hl ^g4 leaves Black better)
13.. .<5^a5!

This fixes the h-pawn on a vulner¬

able square and eyes f6. It is now ap¬
parent that White has a clear advan¬
tage. The black f7- and h5-pawns are
split and vulnerable and the weakness
of the f6-square is more important
than White’s own hole on e5.
17...Sde8 18.ic2 .ieS
Although e5 is a good square for
any black piece, 18...^e5 was more
natural, planning ...c7-c6. Instead Ad¬
The point is that 14 exf6 is an¬ ams intends to utilise the g-file for his
swered by 14....^c5-t-! and then rooks, which only leads to a further
15.. .^xc4. worsening of his chances.
White can maintain the pressure 19 Sf2 Shg8 20 g3 Sg4 21 ^3
with, 14 Axf7-(- '^f7 15 ^e4, but the Sg7 22 ^f5 Sh7 23 a3 a6 24 Sdl
position is by no means clear. b5
12Sxf6i.e6 13i.d3? A bid for counterplay on the queen-
Of course he avoids strengthening side.
Black’s pawn structure with 13 Axe6? 25 ?3d4 -53a5 26 a4 c5 27 -S3f3 *07
fxe6. However, according to Fedorov 28 axb5 axbS 29 .id3 4k:4 30 Sal
13 ^d5 was better, when White has a Sb8 31 Sa7+ Sb7 32 Sxb7-i- *xb7
clear advant^e after 13...0-0-0 14 c3 33 b3 ^5 34 ^xe5
^c6 15 Safi etc. White adds the two bishops to his
13.. ..1g7 14 Sffi ^c6 15 ^d5 0-0-0 other positional advantages. The e5-
Black could have equahsed with pawn will be fatally weak.

70
 other Gambits after 3 fhf3 gS and 3...‘^c6

34.. .dxe5 35 ±xb5 <S^xb3 36 Af6 c4 Game, 1 e4 e5 2 ^c3 ^c6 3 f4 exf4 4

37 i.xe5 h4 38 i.e8 h3 39 af4 Sh5 ^f3, which is in fact the most com¬
40 ±g7 ^c5 41 ±xf7 Zh7 42 ±d4 mon move order.
axf7 43 ±xc5 *c6 44 axf7 ±xf7 Gallagher points out that in the
Black has defended tenaciously and King’s Gambit move order White has
forced play into an opposite-coloured the extra possibility of 4 d4!? Theory
bishop endgame. However, White condemns this move, but not Joe! In
now wins with some accurate play. his book, he analyses this move all the
45 .^d4 .^g6 46 e5 <;^d5 47 ^f2 <;^e4 way through to a rook and pawn end¬
48 e6 <;^d5 49 e7 <;^e6 50 .^c5 <;^d5 game 28 moves deep. King’s Gambit
51 .^b4 *e4 52 .^a5! ^f7 53 .^c7 aficionados will be pleased to know
Now all is ready to advance the g- that White wins the race to queen! In
pawn. Black’s blockade crumbles. summary, one variation of Gallagher’s
53.. .';^d3 54 g4 ';^xc3 55 '1^63 '1^54 after 4 d4 that seems satisfactory for
56 <;^d4 <;^b5 57 <;^e5 <;^c6 58 <;^f6 Black, but no more, is 4...d5 5 exd5
.^e8 59 .^e5 1-0 «xd5 6 i.xf4 i.g4 7 ^c3 (if 7 i.xc7
The g-pawn marches through. Even then 7...Sc8! followed by 8....^xf3 is
if Black could somehow take the e- good for Black) 7...i.b4 8 ±e2 0-0-0 9
and g-pawns for his bishop, White 0-0 Wd7 10 d5 i.xc3 11 dxc6 «xc6 12
would win with the bishop and ‘right’ ^e5! «c5-t- 13 ^hl ±xe2 (taking the
rook’s pawn. queen is bad after 14 .^xg4-h) 14 Wxe2
^xe5 15 .^xe5 ^f6 16 .^xf6 gxf6 17
Game 26 Sxf6 with approximate equality.
Polasek-Karolyi 4...g5 5 d4
Prague 1988 A major decision. White could en¬
ter Allgaier type lines with 5 h4 g4
1 e4 e5 2 f4 exf4 3 ^f3 ^c6 4 ^c3 (forced) 6 ^g5 h6 7 ^xf7 '^xf7 8 d4.

For the sake of clarity I have The difference is that the queen’s
changed the order of moves in this knights are out. Now 8...f3!? is critical
game. It actually began via the Vienna (also possible is 8...d5, but this seems

71
The King's Gambit

stronger) when Gallagher-Hresc, Ge¬ Two other moves should be consid¬

neva 1991, continued 9 ^c4-t- d5 10 ered here:
^xdS-t- '^g7 11 gxf3. a) 7 #xf3!? was tried in Mortazavi-
Miles, London 1994. If now 7...^xd4
8 ±x{7+ ^xf7 9 Wh5+ ^g7 (Black
could try to win with 9...'^e7. How¬
ever, this seems highly dangerous after
10 ^d5-h [also worthy of attention are
10 We5+ ^e6 11 «xh8 and 10 «h4-t-
^f6 11 e5] 10...<^d6 11 i.xf4-t-) 10
#g4-t- '^f7 seems a forced draw by
repetition. If White plays for an ad¬
vantage with 11 0-0 then 11...^f6
should be good for Black who after all
has two extra pieces.
In the game White achieved a good Miles in fact played to win with
position after ll...^b4 12 ^e3 ^f6 13 7.. .d5! 8 ^xd5 (the problem with 8
^c4 We7 14 We2. He castled queen- exd5 is that 8...^xd4 9 WeA+ We7
side and then began a decisive attack forces off the queens) 8...^xd4 9 #xf4
on the kingside. However, Black’s i.d6 (9...^xc2-i- 10 ^fl ^xal? 11
play is not altogether logical. Having ^xc7+ is bad for Black) 10 Wi2 (10 e5
played 8...f3, he should have seized the is possibly a better try) 10...^c6 11
chance to disrupt the smooth build-up i.f4 ^e5 12 0-0! (or else 12 i.b3 i.e6
of White’s game with ll....^e7! Then and White can resign) 12...^xc4 13
after 12 0-0 (the best answer to the Wd4 f6 14 WxcA and now according to
threatened check on h4) White’s king Mortazavi 14...c6 ends White’s com¬
has been forced to live in the airy pensation for the piece. However,
wastes of the kingside rather than in White can carry on attacking with 15
comfortable retirement on the queen- ^xfe-t-l? ^xf6 (15...«xf6.> 16 i.xd6!
side. Gallagher assesses the position as «xd6 17 «f7-t- <^d8 18 Sadi wins) 16
unclear after 12....^xh4 13 f4. How¬ e5 etc. with unclear play. Instead the
ever, after the plausible 13...^f6 14 game continued 14...Axf4 15 ^xf4 c6
.^e3 Se8 I think that Black has a clear 16 e5! and White had a dangerous ini¬
advant^e. His king is safe, his pieces tiative. It seems that 7 Wxi3 offers rea¬
are mobilised and the formidable- sonable practical chances for White,
looking white centre is in fact vulner¬ even though it feels suspect. However,
able (if 15 «d3 ^b4). it is rather spoilt by the fact that Black
Therefore, 5 d4 looks a better try can force a draw with 7...^xd4.
for White. b) 7 Axf4 is suggested by Gallagher.
5.. .g4 6 i.c4 If now 7...fxg2? 8 .^xf7-t-! ‘4>xf7 9
The Pierce Gambit. #h5-t- '^g7 10 Sgl wins. And if
6.. .gxf3 7 0-0 7.. .1.g7? 8 0-0 i.xd4-t- 9 <4>hl i.xc3?!

72
 Other Gambits after 3 /i^f3 g5 and 3...4ic6

10 i.xf7+! ^xf7 11 'td5+ ^e8 he simplify to a endgame where he has

(ll...<^g7 12 2x13!) 12 Wh5+ ^e7 13 two or three pawns and a strong cen¬
e5! gives White a decisive attack - tre as compensation for the piece. And
Glaskov and Estrin. However, why finally, he has lost control of the im¬
not use the f-pawn to disrupt White’s portant dark squares c5 and e5.1 think
plan of 0-0? After 7...f2-h! 8 '^xf2 Ag7 that 7 0-0 has been refuted by this
the white king is badly placed on the f- move.
file. I think that Black is doing nicely.

8 .^xf4 .^c5 9 i.xf7-h

7 0-0 <S^xd4! Gallagher suggests that 9 '^hl is a
Other known moves include 7...d5 better try. However, Black has many
and 7...d6, but I believe this to be the good continuations, for example 9...d6
strongest. Black exploits a tactical fea¬ 10 .^e3 (White has to try and attack
ture of the position to win White’s d- down the f-file; 10 gxf3 .ke6 is hope¬
pawn, as after 8 #xd4?? #g5! both less) 10....^e6!? (returning the extra
9...#xg2 mate and 9...^c5 winning material to seize the initiative) 11
the queen are threatened. There would .^xd4 fxg2-t- 12 '^xg2 .^xc4 13 J.xh8
be no adequate defence against both Wg5+ 14 ^hl i.xfl 15 «xfl 0-0-0 16
threats, as 9 Wi2 ^c5 would still win ®xf7 ^h6 17 #f6 #g4! and since 18
the queen while 9 Sf2 .^c5 10 .^xf4 #xh6 #f3 is mate. Black can play
i.xd4 11 i.xg5 ^e7 12 Sdl (12 gxf3 18.. .5g8 next move with a decisive
Sg8) 12....^xf2-t- 13 '^xf2 fxg2 would attack along the g-file.
leave White a lot of material down. 9.. .*xf7 10.^63 *e8!
White therefore loses his central d- Black is happy to return one piece
pawn. The loss of a mere pawn may in order to break the attack. Less clear
not seem vital when it is considered is 10...«f6 11 ^d5.
that White has gambited a whole 11 .^xd4 ^xd4+ 12 l'xd4 Wfei
piece. However, the d-pawn was es¬ This gains time as 13 e5 would
sential for White’s plans. Now he can exchange queens and win easily.
no longer hope to overrun Black with White’s only hope is a middlegame
a pawn storm in the centre. Nor can attack.

73
The King’s Gambit

13 Wd3 ^e7 14 SxfS l^eS 15 ^dS See the last note. Now the white
c6! queen is forced to a passive square as
This drives the knight from its cen¬ 24 WdA Hfl-t- mates.
tral post and prepares a hole on c7 for 24 #51 Sc8
the king. Black completes development and is
16^f6-t-*d8 17^xd7!? now ready to assume the initiative.
The alternative 17 Sdl sets some 25 Sb3 #62 26 Sel #12 27 Sxb7
nasty traps, for example 17...d5? 18 Sc2 28 Sgl Sg8 29 Wfl Sf8?
exd5 .^f5 (winning a second piece but So far Karolyi has played excel¬
...) 19 «d2! «xf6 20 dxc6-t- ^c7 21 lently, but here he misses an immedi¬
«f4-t-! ^c8 22 cxb7-t- ^xb7 23 fid7-t-! ate win with 29...#xfl 30 Sxfl figxg2
and Black has to give up his queen 31 fif7 (or 31 Sf8+ <^e7) 31...figd2! 32
with 23....^xd7 24 #xf6 or be mated. fibxd7-i- '^c8 and White has no de¬
Also bad for Black is 17...#e6 18 fence against a back-rank mate that
^xd7 «xd7 (18...<^c7 19 ^f8! threat¬ doesn’t cost a rook. As played. Black
ens mate on d8, while 18....^xd7 19 should still be winning easily enough,
Sf6! drives the queen from the defence but he makes it an excruciating expe¬
of d7, e.g. 19...«g4 20 h3! «xdl-t- 21 rience for both players by allowing
Wxdl and White has a dangerous at¬ the game to drift into a laborious
tack) 19 Wfl ^d5 20 exd5 cxd5 21 technical endgame. The remaining
Sfd3! and White threatens 22 fixdS or moves were:
22 ®f6-t-. However, Black has a simple 30 Wxf2 Sfxf2 31 Sxa7 Sxb2 32 a4
reply to 17 fidl: 17...d6! and, since .^f5 33 Sg7 Sa2 34 h4 Sxa4 35
capturing on d6 gives a lost endgame, *h2 Sxh4+ 36 *g3 aff4 37 Sal
the white attack is at an end. Shg4-h 38 Sxg4 Sxg4-h 39 <^3 Sd4
17.. .1.xd7 ISSdl ^d5! 40 Sa5 ^d7 41 <;^e3 Se4+ 42 ^3
Black avoids 18...#e6 19 fif6! ®g4 ^d6 43 Sa2 ^xd5 44 <;^g3 <;^e5 45
20 h3. A less straightforward path is Sb2 <;^f6 46 Sa2 <;^g5 47 Sb2 ad4
18.. .«c7 19 «c3 fig8 20 fifd3 ^e8 21 48 ^h2 Sd3 49 Sa2 Sc3 50 Sb2
Sxd7 #xd7 22 fixd7 '^xd7, though .^e4 51 Se2 ^f4 52 Sf2-l- ^e3 53
Black should win ‘on points’. Sf7 Sc2 54 Sg7 Sc6 55 Sa7 Sc2 56
19exd5 cxd5 Sg7 .^g6 57 Sa7 <1^2 58 Sa4 Sb2
Avoiding the pitfall 19...#xd5? 20 59 Sf4-h <;^e1 60 <;^g1 Se2 61 Sg4
c4! mc5+ (20...«xd3 21 Sfxd3) 21 312! Se4 62 Sg3 <;^e2 63 Sa3 Sc4 64
We? 22 Sfd2. Sa2+ Sc2 65 Sa4 Sd2 66 Sf4 <;^e3
20 Se3 Wde 21 c4 HKcS'. 67 Sf3+ <;^d4 68 Sg3 .^e4 69 Sg5
This threatens a pin with 22...Se8, *e3 70 ag3+ *e2 71 ag7 ac2 72
and so forces White to move his king, Se7 *e3 73 ag7 ScH- 74 *h2 Sal
when back-rank mate themes emerge. 75 ag3+ ^2 76 ag7 aa2 77 ad7
22 <;^h1 Sf8 23 cxd5 #55! Se2 78 Sdl .^xg2 79 Sd2 .^f3 0-1

74
 other Gambits after 3 fhf3 g5 and 3...^c6

Summary
After 1 e4 e5 2 f4 exf4 3 ^f3 g5 4 ^c4 ^g7 neither the Philidor Gambit 5 h4 h6
6 d4 d6 7 0-0 ^c6 8 c3 ^f6 (Game 19) nor the Hanstein Gambit 5 0-0 d6 6 d4 h6
7 c3 ^c6 (Game 20) is satisfactory for White. By omitting 4 h4 (to force 4...g4)
White allows his opponent to set up a solid wall of pawns on the kingside,
which frustrates all his attacking aspirations. The Muzio Gambit 4 .lc4 g4 5 0-0
is a lot of fun, but this may also be imsoimd for White (see the notes to Game
22). White’s other alternatives after 3...g5, such as the Allgaier, Lolli and Pierce
Gambits are also unsoimd, so White should prefer the Kieseritzky (Chapter 2).

1 e4 e5 2 f4 exf4 3

3.. .g5
3.. .^c6 - Game 26

4 h4 g4 5 ^g5 - Game 18
4 d4 - Game 25

4.. .g4
5 0-0 gxf3 6 #xf3 m 7 e5 #xe5 (D)
8 d3 - Game 21
8 i.xf7+ ^xf7 9 d4 #xd4+ 10 i.e3 #f6 11 ^c3 ^e7
12 i.xf4 (D)
12.. .‘^g8 - Game 22
12.. .^f5 - Game 23
5 .^xf7-i- - Game 24
5h4(E>;
5 0-0 - Game 20
5.. .h6-Gdwe 19

7...mxe5 12±xf4 5h4

75
 CHAPTER FOUR | ±±^±K±±±

Cunningham Defence £m£m'...

(3 ±e7)

1 e4 e5 2 f4 exf4 3 <5if3 ±e7 play 4....^h4-i-. Short is in no mood for

The Cunningham 3...^e7 is a very compromise!
solid response to the King’s Gambit. The alternative is 4 .^c4, after
Black avoids creating weaknesses in which Black’s check on h4 is less em¬
his kingside pawn structure with 3...g5 barrassing, as the white king has a ha¬
or 3...d6 4 d4 g5, so his king should be ven on fl. This move will be consid¬
safe on the kingside. His counterplay ered in Games 29-31.
rests in the ...d7-d5 advance to under¬
mine White’s centre and gain freedom
of action for his pieces. An important
question is whether Black should play
....^h4-i- to force the white king to give
up castling. This check is very tempt¬
ing, but the loose placement of the
bishop on h4 makes this move prob¬
lematic. Of course, 4....^h4-i- should be
stronger after 4 ^c3 (Games 27-28)
than after 4 .^c4 (Games 29-31), since
in the former case White has to play 5
‘^e2, which shuts in his bishop on fl. 4....ih4+
Piket accepts the challenge. A more
Game 27 solid alternative was 4...^f6, when
Short-Piket two moves are worth looking at:
Madrid 1997 a) 5 e5 ^g4 6 d4.

see following diagram

1 e4 e5 2 f4 exf4 3 <53f3 .ie7 4 <53c3
A bold move, which dares Black to This is an important position. Black

76
 Cunningham Defence (3 fhf3 Jie7)

now has: move order, but 6...d6 may still be

Black’s best option here.
b) 5 d4 d5 6 .ld3 dxe4 7 ^xe4 <53c6
8 i.xf4 0-0 9 c3 ^xe4 10 i.xe4 i.h4+
11 ^fl i.g4 12 #d3 ^h8 and White
had some advantage in Lukin-
Faibisovich, Leningrad 1967.
5 <^e2 d5
Again the most aggressive approach.
Piket wants to attack the ridiculously
placed white king as quickly as possi¬
ble and therefore opens lines in the
centre.
al) 6...^h4+ (this check is now Black has other, quieter moves such
critical) 7 ‘^e2 ^e3 8 #d3! (this looks as 5...^e7 and 5....kg5, but then White
better than 8 ^xe3 fxe3 9 <^xe3 d6 10 can carry out a plan of d2-d4, ‘^f2 and
exd6, as in Bangiev-Egin, Simferopol .^c4 (perhaps .^d3), unravelling his
1985, when 10...0-0! gives Black a dan¬ kingside and probably emerging with
gerous attack) 8...0-0 9 g3! <53xfl a slight advantage due to his better
{9...^e7 10 gxf4 ^xfl 11 Sxfl gives pawn structure. If Black tries a more
White a huge centre) 10 .^xf4! and gradual pawn attack with 5...c6, then
White regains his piece next move he has to reckon with Hebden’s idea
with a good game. of 6 d4 d5 7 #d2!, when in Gallagher-
a2) 6...^e3 7 i.xe3 fxe3 8 i.c4 d6 9 Faure, Geneva 1989, White obtained a
0-0 0-0 10 'td3 ^c6 11 exd6 i.xd6 small advantage after 7...dxe4 8 ^xe4
(this is better than ll...cxd6, when ^f6 9 #xf4 ^xe4 10 #xe4-i- #e7 11
White had a clear advantage after 12 ‘^d3! #xe4-i- 12 ‘^xe4. As Gallagher
Sael in Spassky-Holmov, Leningrad remarks. White’s king is well central¬
1963) 12 #xe3 (if 12 ^e4 Black can ised for the endgame!
try to hold on to his extra pawn with 6 -?^xd5
12....1f4!?) 12....1g4 looks a Black can play even more aggres¬
little dangerous after 13 ^e5 .^xe5 14 sively with 6....^g4 7 d4 f5. Then Gal-
dxe5 .^xc2 15 e6, but may be playable) lagher-Jacobs, Cdella 1985, continued
13 ^g5?! .Ih5 and Black’s two bishops 8 #d3 ^e7 9 ^xf4 ^bc6 10 c3 #d7
compensate for White’s extra centre 11 e5 g5. Now White fell for a trap
pawn and pressure along the f-file. with 12 g3? gxf4 13 gxh4 ^xe5! In¬
a3) 6...d6!? aims to lead play into stead, Gallagher recommends 12 ^h3
variations examined later after 4 .^c4. as better for White. This seems cor¬
Thus after 7 exd6 #xd6 8 .^c4 0-0 9 rect, e.g. 12...h6?! (12....1xf3-i- is better,
0-0 #h6 we have reached the diagram but then 13 gxf3 0-0-0 14 f4! prevents
at move nine in Game 31. White can ^xe5 and leaves White with a strong
try 8 ^b5 to exploit Black’s irregular centre) 13 g3! (trapping the bishop)

77
The King’s Gambit

13.. .<5ixe5 14 dxeS #xcl3+ 15 ‘^xcl3 11 Wxf4l'e6 12<^e3!

l.xf3 16 Sgl g4 (16...Sd8+ 17 -^63) 17
gxh4 gxh3 18 ^xh3 looks winning for
White in view of the two bishops and
fairly useful extra pawn (but of course
he should be careful to answer
17.. .5.8+ with 18 ‘^c4 rather than 18
<^c2 ±dl mate!).
7 ^3xf6+ Wxf6 8 d4
Here 8 d3!? .Ig4 9 #d2 is a similar
and more cautious version of the game
continuation. However, unless the
piece sacrifice of Game 28 proves good
for Black (which seems doubtful) Short was very pleased with this
White has no reason to avoid gaining move, which introduces two threats,
space by pushing the pawn two the obvious 13 ^xh4 and the sneaky
squares rather than one. 13 .^c4!, when 13...#xc4 14 #xg4-i-
8.. .Ag4 9 Wd2 picks up the other bishop next move.
White clears the d 1-square for his Since ll....^xf3 12 gxf3 is positional
king with gain of time by attacking capitulation, Piket decided on an all or
the f4-pawn. More commonly seen is nothing attack.
9 c3, which gives Black the extra op¬ 12.. .g5 13 «3xg5 .ixgB 14 Wxg5 f5
tion of 9...c5 (9...^c6 10 Wd2 trans¬ 15h3!
poses to the game). However, in Ama- A nonchalant move. Short says that
son-Wedberg, Randers 1985, this he thought 15 e5} ^xe5 16 dxe5 Wb6-i-
turned out badly for Black: 10 dxc5 17 <^14 #f2 mate was best avoided.
#e7 11 HtdSl ^d7 12 i.xf4 ^f6 13 Probably this is the way the game
#e5 ^xe4 14 ‘^e3! (a favourite haimt might have ended 150 years ago.
for the white monarch in this varia¬ 15.. .<53xd4
tion, as Short’s king also ends up on 15...#xe4+ 16 <^f2 i.dl 17 #f4
this square in our main game) #c2+ 18 #d2 #a4 19 i.d3 is entirely
14.. .1.xf3 15 i.b5+ -^18 16 1il(xe7+ hopeless for Black - Short.
^xe7 17 ‘^xf3 <53xc5 and White had a 16 cxd4 &xd4 17 hxg4
sizeable advantage in view of his The correct capture. 17 ‘^xd4
bishop pair, queenside pawn majority would be too outrageous, even though
and the imcomfortable position of the White may still be winning, e.g.
black king. 17.. .5d8+ 18 #xd8+! or 17...#xe4+
9.. .-?^c6 10 c3 0-0-0? (17...Wb6+ 18 ^c3) 18 ^c3 #e5+
This is totally bad. Black had to (18...#c6+ 19 1.C4) 19 <^c2 and the
preserve his kingside clump of pawns king evades the checks.
with 10...g5, for which see the next 17.. .6xe4+ 18 ^2 fxg4 19 &h6
game. &f8+

78
 Cunningham Defence (3 fhf3 ±e7j

lems that Short’s unexpected opening

had set him.
23 *g3 md3+ 24 *xg4 1-0
Black resigned since the checks soon
dry up, when White wins on points.

Game 28
Gallagher-Klovans
Oberwart 1993

1 e4 e5 2 f4 exf4 3 -?^f3 ^el 4 -?^c3

.ih4+ 5 *e2 d5 6 -2^xd5 7
20 *g1?? <53xf6+ Wxf6 8 d4 .ig4 9 Wd2
Short sets the scene in his Sunday 10 c3 g5
Telegraph chess column: ‘I had seen This is much better than Piket’s
that 20 ^g3 mu 21 i.e3 Sg8 22 Wc5 10...0-0-0 in Game 27.
b6 23 #c3 was the end of the excite¬
ment, but I became obsessed with the
idea that Black might be able to give
an (imaginary) check with his queen
on the b8-h2 diagonal. I decided to
play the safer move.’
It only remains to add that Piket
was in almost fatal time pressure with
20 moves to go and that Short was
trembling uncontrollably. Yes, the
King’s Gambit is not for the faint
hearted.
20...axf1+ 21 *xf1 SelH- 22 4>f2 11 *d1 0-0-0 12 *c2 SheSi?
We2+? Standard theory gives 12....^xf3 13
Black could have forced perpetual gxf3, when the strong white centre
check with 22...Se2-i-! 23 ‘^g3 Sxg2-i- and bishop pair give White the advan¬
24 ‘^xg2 me.2+. This would have been tage. Gallagher also mentions 12...#h6
an amazing finish: Black is two rooks 13 ^xh4 #xh4 14 g3! with advantage
and a bishop down, but forces a draw to White. No doubt Joe was hoping
with his last piece! Certainly a re¬ for this when he was rocked back with
minder never to give up hope. 12...She8.
Though in fairness to Piket (and 13.id3
Short) it should be remembered that White sensibly declines the offer.
Black was desperately short of time. Klovans later demonstrated in Infer-
And, of course, he wouldn’t have been mator 58 that Black has a vicious at¬
short of time but for the novel prob¬ tack after 13 <53xh4 Sxe4!, e.g. 14 ^f3

79
The King's Gambit

(or 14 ^cl3 <5lxd4+! 15 cxd4 Sexd4 16 pieces in the centre.

#c3 gxh4 17 b3 h3 18 gxh3 ^f5 with a 17 <£>61 &c6 18 Wb3 a5 19 a3 a4 20
dangerous attack for Black) 14...^f5! Wb5 <i’b8 21 ±e2 Wd8 22 Sfl Sb6
15 .^d3 g4 16 <53el (16 Sel looks bet¬ 23 #c5 &c6 24 Wb5 &b6 25 Wa5
ter, though Black has a strong initia¬ Sb3 26 We5 f6 27 Wc5 &b6 28 Wa5
tive after 16...Se3 17 Sxe3 fxe3 18 Wd6 29 <i’a2?!
Wxe3 gxf3 19 Wxf3 i.xd3+ 20 #xd3 A mistake. Klovans suggests that 29
Sxd4!) 16...^xd4-i-! 17 cxd4 Sexd4 18 Sa2 would have been unclear.
#c3 #e6 19 i.xf5 #xf5+ 20 ^b3 29.. .We6+ 30 *b1 Wd6?
#b5-i- 21 ‘^c2 Sc4 winning the white Black misses the chance of
queen and keeping up the onslaught. 30.. .5xe4!, when White has to grovel
13....ixf3 14 gxf3 <53xd4+! with 30 .kd3 as 31 fxe4? #xe4-i- 32
Black has to do or die, since slow ^a2 #e6+ 33 ^bl Htxel 34 Sgl ±{2
play leaves White with a clear posi¬ would win for Black.
tional plus. 31 *a2 We6+ ’/2-’/2
15cxd4 Sxd4 The surprise value of ...She8 gained
Klovans an easy draw as Black, but I
doubt if the experiment should be re¬
peated

Game 29
Gallagher-Neussner
Loosdorf 1993

1 e4 e5 2 f4 exf4 3 -?^f3 4 .ic4

.ih4+
This check is not so attractive in
this position, when compared to 4
16 Wc3? ^c3, since White has cleared fl for his
White hastens to break the pin on king. Nevertheless, it is a fighting
the d-file, but why not 16 a4!, e.g. move which sets difficult problems for
16.. .5ed8 17 Sa3 bringing the queen’s both players.
rook into the defence. It is hard to The quieter 4...<53f6 is the subject of
believe in the strength of Black’s at¬ Game 31.
tack, since besides the piece sacrificed 5*f1
the bishop on h4 makes little contri¬ The Cunningham Gambit 5 g3?! is
bution to the game. virtually refuted by 5...fxg3 6 0-0 d5!
16.. .6e6 (instead 6...gxh2-i- plays into White’s
Now the rook on al remains out of hands. After 7 “^hl d5 8 .^xd5 ^f6 9
the game and Black’s initiative com¬ i.xf7+ <^xf7 10 ^xh4 Sf8 11 #f3 ^gS
pensates for the missing piece in view 12 d3 an unclear position is reached.
of the awkward congestion of white White has a strong centre but his king-
 Cunningham Defence (3 i^f3 ±e7)

side is denuded of pawn cover) 7 ±xd5 7.. .1.g4 8d3 0-0

^f6 8 ^xf7+ (8 4ixh4 ^xd5 favours The best move. Three alternatives
Black) 8...<^xf7 9 e5 (9 ^xh4 lfd4+) in descending order of inferiority are
9...^h3 10 exf6 ^xfl 11 #xfl gxh2+ 8.. .^xe4?? 9 We2; 8...^h5? 9 i.xf7+
12 ‘^hl ^xf6 and Black had a winning and 8...<Si\c6?! 9 .&xf4.
material advantage in Krejcik- 9 Wd2
Schlechter, Vienna 1918.

9...4^h5!?
5.. .d5 An important moment. By delaying
The only good move. Black wants ....&xf3 for a move Black avoids the
to develop his king’s knight to f6 variation 9....&xf3 10 gxf3 ^h5 11
without dropping the bishop on h4 Wg2! ^c6 12 Wg4 Wg5 13 Igl «xg4
and 5...Ae7 wastes too much time af¬ 14 fxg4!, which is good for White ac¬
ter 6 d4. cording to Gallagher. The question is,
6 .ixdS can White exploit this delay by play¬
Almost universally played, but 6 ing 10 ^xh4 to avoid ...Axf3 next
exdS is an interesting alternative, after move? The answer seems to be ‘No’:
which Black has to retreat his bishop. 10 ^xh4 #xh4 11 1^12 ^g3+ 12 ^el
Perhaps best play is 6...±f6 7 d4 g5 (12 '4'gl ^e2-l- 13 '4'fl W6! is very
(7...^e7 8 ^c3 ^g6 9 ^e2 Wd6 10 good for Black) 12...m5 13 Igl ^e2
Wdl looks better for White) 8 ^c3 14 h3 and Black has the choice be¬
±f5 9 h4 h6 etc. with unclear play. tween 14...^xgl 15 hxg4 'B^hl 16 Wl
6.. .<53f6 m4+ (not 16...f3 17 ^f2!) 17 W{2
This is the point of Black’s last Whl 18 Wfl #h4-(- with a draw by
move. He can now develop his knight repetition and 14...<?:\xcl!? 15 hxg4
immediately, as 7 <53xh4 ^xd5 #a5-i- 16 <53c3 <?i\xb3 17 cxb3, which
shouldn’t trouble him. looks better for Black.
7 .ib3 10 M -?^c6?
This retreat is possible as 7...<53xe4? Now Black goes wrong. He had to
8 We2 wins a piece. The alternative 7 play 10....^xf3! 11 gxf3 ^c6 12 'i^g2
.^c4 is examined in the next game. (no better is 12 ^d5, e.g. 12...^d4 13

81
The King’s Gambit

^xf4? ^g5) 12...^d4 13 Wg4? (after ens 10 ^xh4.

both 13 <?^e2 ^xb3 14 axb3 f5 and 13
^d5 ^xb3 14 axb3 f5 Black is slightly
better) 13...g6 14 ^d5 (14 .^xf4? ^xb3
15 axb3 f5 16 exf5 Sxf5 is bad for
White) 14...<53xb3 15 axb3 f5 and
White is in trouble.
11 4^xh4!
Now this move is perfectly possible,
which means that the f4-pawn is very
vulnerable.
11...®xh412'tff2Wxf2+
Of course, 12...^g3+? 13 ‘^gl just
loses a piece now that el is defended 9....ixf3 10 .ixf3 Wd4 11 #62 0-0-0
by White’s knight. This seems to lose. Black should try
13 *xf2 -?^d4 14 -?^xb3 to mobilise his kingside pawns
After this Black will have to wreck straightaway with ll...g5, when White
his kingside to defend f4. However, replies 12 d3 and now:
the attempt to mount a counterattack
with 14...“^hS 15 <53xf4 (not 15 ^xc7?
Bac8) 15...f5 fails after the simple 16
e5.
15 axb3 g5 16 g3 c6?
Black quickly falls apart after the
game move. The best defensive chance
was 16...fxg3+ 17 hxg3 f6!
17 ^e7+ *h8 18 gxf4 f5? 19 h3
fxe4 20 hxg4 -?^xf4 21 -?^f5 1-0

Game 30
McDonald-Hector a) If now 12...Sg8.> 13 e5! ^d7
Oviedo 1992 (Black loses his queen after 13...#xe5
14 .^xc6-i- or a piece after 13...^xe5 14
1 e4 e5 2 f4 exf4 3 -?^f3 4 .ic4 ^b5 #c5 15 d4) 14 g3 fxg3 15 hxg3
kh4+ 5 *f1 d5 6 .ixd5 7 .ic4!? .^xg3 16 e6 and Black’s king seems to
An alternative to 7 .^b3 in the pre¬ be in the most danger.
vious game. b) Instead 12...<53e5 is better here (or
7..Ag4 8 -?^c3 <53c6 9 ±e2 one move earlier with ll...^e5, when
12 d3 g5 transposes). Then 13 g3! gives
see following diagram
Black the choice of 13...fxg3 or 13...g4:
This is White’s idea. He breaks the bl) 13...fxg3 14 hxg3 g4 (14....1xg3
pin on the knight and thereby threat¬ 15 .^xg5 is excellent for White in view

82
 Cunningham Defence (3 ^f3 Ae7)

of his strong centre and the freedom 18 Wxd4 g2+ 19 *g1 Sxd4 20
his queen’s bishop now enjoys) 15 ±xh5 f3
^xg4 (not 15 ^g2 ^h5!) 15...<5lfxg4 White has a rook trapped but two
16 Sxh4. White has a big positional pieces is a lot of consolation.
advantage in addition to his extra 21 .if4 «^g4 22 .ixg4+ Sxg4 23
pawn. .ig3 &b4 24 b3 Sb6 25 ^d5
b2) 13...g4 14 i.g2 (14 i.xg4? ^fxg4 gxhlW+ 26 <i’xh1 &e6 27 ^e3 Sg8
15 gxh4 f3 would be bad for White) 28 Sfl b5 29 Sxf3 Sf8 30 Sf5 Sa6
14.. .f3. Now a strange situation has 31 Sxb5 &xa2 32 &c5 1-0
arisen in which both players have a A bizarre game.
bishop trapped. Play could continue
15 W{2 fxg2+ (of course this bishop Game 31
couldn’t run away, but it is difficult to Belotti-Loncar
see what else Black can do) 16 ‘^xg2 Mitropa Cup 1995
#xf2+ 17 ‘^xf2 and White picks up
the bishop with a good game, unless 1 e4 e5 2 f4 exf4 3 -?^f3 ^e7 4
Black plays 17...<53xe4+ 18 ^xe4 Ae7.
However, White then has a very
pleasant position after 19 in view
of Black’s weak kingside pawn struc¬
ture.
12d3g5 13 g3! 14Wg2!
Black’s strategy is refuted by this
quiet move, which creates a retreat
square for the bishop on f3 and de¬
fends g3 a second time.
14.. .g4 15.id1
This isn’t normally a square that the
bishop hopes to end up on in the
King’s Gambit. Nonetheless, White is Black spurns the check on h4,
glad that this retreat is available as he which is probably sensible in view of
now wins a piece. the analysis in Games 29 and 30. In¬
15.. .<53h5 stead he develops and looks to equalise
Black has to stake everything on an with an immediate 5...d5. White there¬
attack as 15...f3 loses a piece after 16 fore kicks the knight away.
Wi2. Also inadequate is 15...^xd3 af¬ 5 e5 -?^g4
ter 16 cxd3 #xd3-i- 17 ^e2 f3 18 .^xd3 Less good is 5...^h5 after which
fxg2-i- 19 <^xg2 Sxd3 20 gxh4. Estrin suggests that 6 <Si\c3 d6 7 exd6
16gxh4 ShgS 17 Wf2 g3 #xd6 8 d4 ^c6 9 0-0 0-0 10 i.e2! i.g4
The last gamble, but White now 11 ^e4, planning 12 <53f2, is awkward
also picks up the knight on h5 and for Black.
Black is hopelessly outgunned. 6 0-0

83
The King’s Gambit

White’s other moves are 6 and ^e7 13 d5 gives White the superior
6 d4, which usually transpose into 6 chances) 11 <53h4!, hitting the bishop
0-0 lines. For example, 6 <53c3 d6! 7 and planning an attack along the f-file
exd6 Wxd6\ 8 d4 0-0 9 0-0 Wh6 trans¬ after 12 .^xf4 etc. Black tried ll...#g5,
poses to the game. Or 6 d4 d5 and but the endgame was miserable for
again we reach the game after 7 exd6 him after 12 <53xf5 #xf5 13 #e4! #xe4
fcdh 8 ^c3 0-0 9 0-0 m6. Note that 14 ^xe4.
Black’s correct response to 6 ^c3 is Nevertheless, it may be that 9 #el-i-
6.. .d6: if Black plays 6...d5 (the stan¬ is not White’s strongest move.
dard move against 6 0-0 and 6 d4) then b) In Hebden-Malaniuk, Vrnjacka
he will get a rude shock when White Banja 1991, White preferred 9 ^c3 0-0
replies 7 .&xd5! 10 ^e4! This improves on 10 ^e2
6.. .d5 <53e3 11 A.xe3 fxe3, when Black was
The main alternative was 6...^c6 7 slightly better in Keres-Alatortstev,
d4 d5 8 exd6 A.xd6 and now: USSR 1950. Hebden’s move attacks
the bishop on d6 and thereby under¬
mines the f4-pawn. Black could find
nothing better than to liquidate to a
slightly worse enc^ame: 10...^f6 11
^xd6 #xd6 12 c3 i.g4 13 #d2! (this
move is a Hebden speciality) 13...4ld5
14 i.xd5 #xd5 15 #xf4 i.xf3 16 #xf3
Wxf3 17 Sxf3. The bishop is much
superior to a knight in this type of
position, but Malaniuk’s Russian
technique succeeded in holding the
balance after 17...Sad8 18 .lf4 Sd7 19
a) The check 9 #el-(-!? is awkward Sel ^d8! 19 b3 ^e6 etc.
for Black. If 9..Me7 then 10 Wxe7-i-
'4'xe7 11 ^c3 .&f5 12 ^d5-l-, planning
13 c3, is slightly better for White in
the endgame. Or if 9...^e7 10 h3 ^h6
11 ^e5 g5 12 h4 f6 13 hxgS fxgS 14
^f3 gives White the better chances
according to Estrin and Glaskov. In¬
deed, Black’s kingside looks pretty
flimsy here. Finally, 9...‘^f8 was
played in Illescas-Fernandez, Las Pal¬
mas 1987, when White obtained a
clear advantage after 10 ^c3 .^fS?! (it
was better to play 10...g5, though 7 exd6 Wxd6!
Bhend suggests that 11 h3 <53h6 12 <53e4 This is much better than 7....^xd6,

84
 Cunningham Defence (3 ^f3 Ae?)

when in Gallagher-Reinhard, Eupen remains solid. Fience he puts his queen

Open (rapidplay) 1995, White had a on a square where it cannot be chased
good endgame after 8 We2+! We7 9 away by a white knight.
Wxe7+ <^xe7 10 d4 ±f5 11 ±b3. Black
didn’t put up much of a fight:
11...4ld7?! (ll...Se8 12 <5lc3 c6, prepar¬
ing ...<^f8, was a better try) 12 <53c3 c6
13 <53g5 ^g6 14 ^xf4 ^xf4 15 Sxf4 f5?
(a horrible move, but 15...<53gf6 16
Sel-i- <^f8 17 d5! is pretty awful, as
Black is playing without his king’s
rook) 16 ^e6 <^d6 17 ^xg7 ahf8 18
Sel and Black soon resigned.

This is not very promising, but the

alternatives were no more enticing.
For example, if 10 Wei then 10....i.e6
looks good (11 d5? I.c5+). Alterna¬
tively, White can try 10 4ld5, but
10.. .±d6 11 Sel l.e6 12 h3 c6!? 13
<53b6 axb6 14 .^xe6 fxe6 15 hxg4 Wg6
is slightly better for Black according to
Blatny.
8 d4 0-0 9 <53c3 Whei 10.. .<Sie3 11 i.xe3 fxe3 12 <5ie5?!
This is an important improvement Blatny suggests that 12 4ld5 is bet¬
on 9...c6? 10 h3 <53e3 11 .^xe3 fxe3 12 ter, in order to regain the pawn with
4le5 (as in Hebden-Fassert, Guernsey approximate equality after 12....i.d6 13
1988) when if Black tries to defend f7 Wd3 ae8 14 Sael.
with 12...±e6 then 13 <53e4 chases the 12.. .1.e6
queen away from the defence of the All as planned (see the note to move
bishop. Then after 13...Wd8 14 ^xe6 nine). Fiowever, 12....^f6, to answer
fxe6 15 Wg4 White has a winning at¬ 13 <^d5 with 13...±xe5 14 dxe5 l.e6,
tack. Another variation on this theme may be even better (Shofman-
is 9...<53e3? 10 .^xe3 fxe3 11 <53b5! Wd8 Antoshin, Moscow 1953).
12 <53e5 ±f6 (if 12...±e6 13 ±xe6 fxe6 13 i.xe6 Wxe6
14 Wg4) 13 ^xf7 axf7 14 ^5 and Compared to the variations exam¬
White’s attack is decisive. ined at move nine, where Black had to
With 9...Wh6! Black wants to play answer .^xe6 with ...f7xe6. Black’s
....^e6 and be in a position to answer kingside is rock solid. This means that
.^xe6 with ...Wxe6, when his kingside he can now start to undermine

85
The King’s Gambit

White’s centre, a process that begins has the last laugh).

on the next move. 19...i.g3!
14 Wf3 c5! 15 'txb? cxd4 16 <5:id5
After this Black’s attack soon be¬
comes overwhelming. White had to
snatch the exchange with 16 Wxa8,
though 16...dxc3!? 17 <53(13 <53c6 18
Wh? ab8 19 Wa6 e2 (Blatny) with
ideas of 20...cxb2 and 20...±h4 gives
Black a strong initiative.
16.. .1.d6 17<af3
17 Wxa8 Wxe5 18 <53f4 g5! and wins.
17.. .e2 18 WxaS 6x11#+ 19 ^xfl
Blatny points out that 19 Sxfl <53d7
20 Wc6 <^b6!! wins the knight, as 21 20 «b7 ad8 21 c4 dxc3 22 <53xc3
<53xb6 is met by 21....^h2+ (but not #63! 23 <53e4 'td3-H 0-1
the immediate 20...±h2+ because of After 24 <^gl Wdl-i-! mates or wins
21 '4’xh2 Wxc6 22 <53e7+, and White material after 25 <53el Wxal.

86
 Cunningham Defence (3 fhf3 ^e7)

Summary
After 1 e4 e5 2 f4 exf4 3 k.t7 White has an interesting choice between 4 4ic3
and 4 ^c4.
The assessment of the line 4 4^c3 J.h4+ depends on Black’s piece sacrifice in
the variation 5 '4’e2 d5 6 4ixd5 4if6 7 4^xf6+ Wxf6 8 d4 ^g4 9 Wd2! 4^c6 10 c3
g5 11 '4’dl 0-0-0 12 '4’c2 She8!? 13 ^d3 J.xf3 14 gxf3 <53xd4+ 15 cxd4 Sxd4 in
Game 28. If 16 a4 is good for White then it is difficult to see where Black’s play
can be improved earlier. If Black plays 4...4lf6 White has at least 5 d4 d5 6 J.d3
dxe4 7 <53xe4 with a slight advantage.
After 4 .^c4 .^h4+ 5 '4’fl d5 6 .^xd5 4lf6 (Games 29 and 30) 7 J.c4!? J.g4 8
4lc3 <53c6 9 .^e2 .^xf3 10 .^xf3 Wd4 11 #e2 looks very good for White, so Black
should prefer 4...<53f6 (Game 31).

I e4 e5 2 f4 exf4 3 i.e7

4<53c3
4 ±c4 (D)
4.. ...^h4+ 5 '4’fl d5 6 ..^xdS 4^16 (D)
7 l.b3 - Game 29
7 J.c4 - Game 30
4.. .<53f6 - Game 31
4...i.h4-l- 5 ^e2 d5 6 ^xdS 1 ^xi6+ WxfB 8 d4 i.g4 9 '»d2 lhc6
10 c3 (D)g5
10...0-0-0 - Game 27
II 'txf4 - Game 28

£K£X«H S Em mm. m
mtmtmimt illii Biriii illii PlPl
B ® ® 1 ^
9f Wf
■ 5 B1 • ■ ■i* ii
P ''9 ®
itflill "Bill t Kit SI Bill

4 k.c4 10 c3

87
 CHAPTER FIVE | m±m. mtmt'

Modern Defence (3 dS)

1 e4 e5 2 f4 exf4 3 d5
The Modern Defence is a very solid This is much better than 4...Wxd5,
approach by Black, based on Reuben which loses time after 5 ^c3. The
Fine’s maxim that the antidote to all knight will be well centralised on d5.
gambits is ...d7-d5. After 2...exf4 3 <53f3 Black has experimented with
d5 4 exd5 Black hopes to ex¬ 4.. ..^d6 here, but then the vigorous 5
change White’s d5-pawn for the f4- d4 and 6 c4, seizing space in the cen¬
pawn. Then he should achieve a fluid tre, should give White a good game.
and rapid development of his pieces, as It should be mentioned that Black
White is deprived of disruptive pawn also has the option of transposing into
thrusts such as e4-e5. White has two other variations here. The Cunning¬
distinct responses to Black’s plan. ham is reached after 4....^e7 5 .^c4
First, he can play 5 .^c4 (Games 32- .^h4+ 6 'i’fl (see Chapter 4) while
33), allowing 5...<53xd5, when a quick 4.. .c6 5 d4 .^d6 6 <53c3 is the Nimzow-
..^.xdS should give him a very small itsch Counter-Gambit (see Chapter 7).
positional advantage as he can seize
some space with d2-d4. Second, White
can gamble with 5 .^b5+ (Game 34).
This crosses Black’s plans and prom¬
ises more winning chances, though at
much greater risk. It’s your choice!

Game 32
Gallagher-Van der Sterren
San Bernardino 1992

1 e4 e5 2 f4 exf4 3 <Sif3 d5 4 exd5

 Modern Defence (3 ^f3 d5)

5 i.c4 9...c5? This is a very natural move

The alternative 5 ^b5+ is the sub¬ which aims to liquidate White’s cen¬
ject of Game 34. tre. However, it can be refuted by a
5.. .<Sixd5 concrete tactical variation: 10 ‘?3c3
Two inferior alternatives for Black Wc4 11 Wei! M6 12 ±d6 ±xd4+ 13
should be dismissed here. First, ^hl ad8 14 ^e4 f5?! 15 Wh4! ^c6
5.. ..^d6?! allowed White to force a and now Gallagher uncorked 16 <53e5!!
favourable endgame after 6 We2+! We7 If 16...<53xe5 17 Wxd8-i-, while 16...Wd5
7 Wxe7+ <^xe7 8 d4 ±f5 9 ±b3, plan¬ 17 ‘?3xc6 Sxd6 18 4le7+ wins the
ning 10 c4, in Gallagher-Metzger, queen. In both Gallagher-Balashov,
Lenk 1989. Second, 5...<53bd7 worked Lenk 1991, and Gallagher-Campora,
out badly for Black after 6 d4 <53b6 7 Biel 1990, Black tried 16...±xe5 but
±b5-i- i.d7 8 We2+ in Gallagher- was hopelessly behind on material
Ferretti, Chiasso 1991. after 17 <53f6+ .kxf6 18 Wxc4+ etc.
6 0-0
It is quite possible that 6 .^xdS! is
the correct move here, as the game
continuation is unpromising for
White. If White wants to play ±xd5,
it is best to do so before Black has
played ...±e6, so that Black is forced
to recapture with his queen rather
than with his bishop. However, as¬
suming that Black avoids the tactical
trap discussed below, it seems that
White cannot hope for much advan¬
tage by giving up his powerful bishop. Suitably impressed, I tried to catch
For a consideration of 6 .^xdS, see the Lawrence Cooper with this trap at the
note to Black’s seventh move. British Championship in 1993. I de¬
6.. Ae7 cided to choose a move order which
The alternative 6....^e6 is examined ruled out 6...^e6 or 7...Ae6, so the
in the next game. game went 6 .^xd5 Wxd5 7 0-0 ^e7 8
7 d4 i.e6 d4 0-0 9 .^xf4, reaching the diagram
The main alternative is 7...0-0 with position. Imagine my excitement as
the standard continuation 8 ±xd5 Black’s hand reached for the c-pawn...
tfxdS 9 ±xf4 However, Cooper had obviously read
Gallagher’s book, as he played 9...c6!
see following diagram_
After 10 <53c3 Wd8 White managed to
Gallagher relates how within the build up a promising position, begin¬
space of two years, two grandmasters ning with 11 Wd3, but objectively I
and an international master all fell for feel that chances should be equal with
the same trap against him by playing best play. Black has the bishop pair
The King’s Gambit

and a solid pawn structure to balance pawns. Eventually, White is forced

White’s space advantage. However, into an inferior endgame.
that is not quite the end of the story, 14...g6 15 *g4 h5 16 *h4 f6 17
as the alternative 11 Wd2 was played 5lf3 <53d7 18 .^xf4 g5!
in Gallagher-Hedke, Biel 1992, and
White quickly obtained the advant^e
after some inferior play by Black:
(ll...i.e6 looks safer) 12
d5!? 4lb6?! 13 d6 and White’s
passed pawn gives him a clear advan¬
tage. Here Gallagher tried 14 4le4,
allowing 14...±xb2. This seems need¬
lessly speculative: the simple 14 b3,
restraining the knight on b6 and plan¬
ning Sadi or Sael followed by 4le4,
c2-c4 etc., looks very strong for White.
Whether or not White has any real 19 i.xd6 *63-1- 20 *f2 *xf2H- 21
advantage after 9...c6, he should still &xf2 cxd6
give this line preference to that The endgame is clearly better for
adopted in our illustrative game, in Black in view of the serious weakness
which Black achieves a good position. of the c-pawns. However, Van der
Therefore, White should aim to play Sterren gets tricked.
.^xdS before Black’s ...Jieb-, probably 22 ^d2 SfeS 23 <^^ &ac8 24 c4 d5
6 .^xdS is the most accurate moment, 25 cxd5 &xc2 26 Ae2 <^>18 27 a4 f5
as in McDonald-Cooper in this note. 28 SxeS-i- <ilxe8 29 <iie2 ^f6?
8We2 The precise 29...Sc3! would have
White wants to challenge the knight prevented the white king from ad¬
on d5, but first he must defend his vancing to d3. Hence the king would
bishop on c4. be unable to support the move ?k4,
8.. .0-0 9 ?lc3 5lxc3 10 bxc3 .^xc4 which proves a vital part of White’s
11 Wxc4 i.d6 12 WbS b6 13 ^g5 strategy in the game continuation: the
‘With a good game for White’ ac¬ knight is brought to a strong centre
cording to Gallagher. However, square where it defends d6 and intro¬
Black’s next move seems to refute this duces the idea of ^5-h. White would
verdict. be left without a good plan, as the al¬
13.. .'te7! 14*15 ternative 30 Sfl f4 31 ^4?.? Se3-h
The variation 14 .^xf4 .^xf4 15 would of course fail.
Sxf4 *e3+ 16 Sf2 c5! 17 «lf3 *xc3 is 30 ^d3 &c7 31 d6 &d7
the reason that things look good for As a result of the inacciu-acy on
Black. White therefore tries to attack move 29, Black’s rook has been (Wven
on the kingside, but his queen ends up to a passive position and White’s
being pushed around by Black’s knight and rook have become active.

90
 Modern Defence (3 fhf3 d5)

32 &e1+ <£>f8 33 ^c4 sii'g? 34 SeS was played in Gall^er-Murey, Metz

<ifg6 35 a5 g4? 1990. Now instead of 10 .^xf4 k.e6 11
Overlooking a tactic. The correct 4ic3, which led to a draw, 10 axb3
path was 35...bxa5 36 Sxa5 Sb7. fxe3 11 Sel seems preferable, e.g.
36 &b5! 11.. ..^e7 12 Sxe3 0-0 13 c3 with a su¬
Threatening 37 4ie5+ and therefore perior pawn structure for White.
winning the b-pawn. Now Black, pre¬ b) 7...c5 is an idea of the English
sumably in time pressure, collapses player Gavin Wall. The best response
completely. may be 8 4ic3, when both 8...4ixc3 9
36...&d8? dxc3, intending 10 .^xf4, and 8...4ic6 9
The best chance was 36...‘^?g5. 4ixd5 .^xdS 10 d3 should favour
37 axb6 axb6 38 ^e5-i- <ilg7 39 White. So the critical move is 8....^e7.
&xb6 ^d5 40 &b7-l- ‘i>f6 41 ^f7 &f8 However, 9 We2 ^c6 (9...0-0 drops a
42 d7 5if4+ 43 *c4 5ie6 44 d5 1-0 piece) 10 We4!? may be good for
A pretty finish. White.
8c4!
Game 33 It seems that White is committed to
Hector-Ziatdinov entering a complex sacrificial line if he
Antwerp 1994 wishes to fight for the initiative.
In McDonald-Weill, Douai 1992, I
1 e4 e5 2 f4 exf4 3 ^f3 d5 4 exdS tried to prove that 8 d4 was good, and
^f6 5 .^c4 ^xd5 6 0-0 .^e6 found my opponent to be in a co¬
operative mood: 8...0-0 9 c4 ^3! (this
is the reason that c2-c4 is usually
played before d2-d4) 10 .^xe3 fxe3 11
^c3 f5? 12 c5! Wc8 13 ^d5 .fi.xd5 14
IxdS-h sfehS 15 ^e5 i.f6 16 Wf3 c6?
17 ^g6-i-1-0. Black should have played
11.. ..^g4!?, which looks annoying as
12 Wd3 can be answered by 12...^c6.
For my part, perhaps 9 Sel was bet¬
ter, maintaining the tension and keep¬
ing c2-c4 in reserve. Possibly White
would even dispense with c2-c4 and
This is the main alternative to prefer c2-c3 to defend the d4-pawn.
If now 7 d4?? - a highly natu¬ 8.. .51b6 9 d4!
ral move! - 7...4ie3 8 Jixei .fi.xc4 wins This gambit line is the only way to
the exchange. set Black problems.
7 .^b3 i.e7 9.. .51xc4
Two other moves should be consid¬ If Black declines the offer, e.g. with
ered: 9.. .0-0, then White has some advant^e
a) 7...4ic6 8 d4 4ie3 9 .^xe3 .^xb3 after 10 d5 and 11 .fi.xf4.

91
The King’s Gambit

10i.xf4 c6?! the game Bangiev-Flomin, Correspon¬

In Hebden-Geller, Moscow 1986, dence 1986-87.
Black was soon in trouble after 10...0-0 c2) 16 d5 is recommended by
11 We2 b5 12 ^c3 a6 13 a4! as his Bangiev as being very strong. How¬
queenside was collapsing. Black there¬ ever, after 16....^c5-i- 17 <^hl 0-0-0 18
fore delays castling and spends a move dxe6 Wxe6 19 ^b5 She8 20 Sfel ^d7
bolstering his centre and queenside (rather than Bangiev’s suggested
immediately. However, the game con¬ 20.. .<53d5) things look awkward for
tinuation indicates that Black is taking White, e.g. 21 Wcl Wh6 22 ±xc7
a fatal risk with his king’s safety. Wxb5 23 ±xd8 SxdS.
The critical move is 10...4^b6, re¬ c3) 16 <53b5! <53d5?! (perhaps the best
treating the knight. Then after 11 move is 16...±xe5, when Black should
.^xe6 fxe6 12 We2 <53c6 13 <53c3 Wd7 be able to survive after 17 Wh5-i-!? g6
14 ^e5 an important position is 18 'txe5 0-0-0, e.g. 19 ^xc7 'txd4+ 20
reached. Wxd4 Sxd4 21 <53xe6 Sc4) 17 .^xd6
cxd6 18 Sael is given by both Bangiev
and Gallagher as a small advantage for
White. In fact. Black seems to be los¬
ing, as 18...0-0-0 loses the d-pawn as
well as the e-pawn after 19 Wxe6 Wxe6
20 Sxe6. Holding on to e6 with
18.. .<^e7 looks ghastly after 19 Wg4
etc.
Black should therefore try the al¬
ternative mentioned at move 16 in
note c3), or more sensibly, give the
whole 6...Ae6 line a miss.
a) 14...<53xd4? 15 Wh5+ g6 16 <53xg6 11 We2 b5 12 a4 a6 13 axb5 cxb5
looks decisive. 14 ^c3 ^c6 15 Sadi .^.dB
b) 14...Wxd4+ 15 '4’hl <53xe5 16 Stopping the d-pawn in its tracks.
kxeS Wc4 17 Wh5+ g6 18 ^3 af8 19 Black only needs one more move -
axf8+ ±xf8 20 Wxh7 0-0-0 21 ±f6 castles - and his opening will have
Sd7 (or 21...Se8 22 Wxg6 with advan¬ been a complete success. Unfortu¬
tage to White) 22 Wxg6 or 22 'ffg8 nately for him, he is swept away by a
looks awkward for Black. wave of taaics before he can find time
c) 14...<53xe5 (Gallagher gives this for this vital move.
move as dubious, but doesn’t suggest 16 ^eS! ^6xe5
what Black should play instead) 15 The natural 16...0-0 loses a piece af¬
.^xe5 .^d6 and now: ter 17 <53xd5 'ffxd5 18 4lxc6 Wxc6 19
cl) 16 ±xd6?! cxd6 17 Sael af8 18 Wxe7.
axf8+ ^xf8 19 Wxe6 Wxe6 20 Sxe6 17 ±xe5 f6
Sd8 21 <53e4 and a draw was agreed in At first glance it seems that Black

92
 Modern Defence (3 fhf3 d5)

can escape the worst with 17...0-0, as 24*68+ 1-0

18 ±xg7? <^xg7 19 ^xd5 Wxd5 20 A brilliant attacking game.
Wxe7 We6 isn’t so clear. However,
White has the insidious move 18 Game 34
Ac7!!, when Black loses a piece after Westerinen-Korneev
18...Wxc7 19 ^xd5 etc. Zaragoza 1995

1 e4 e5 2 f4 d5 3 exdS exf4 4 <2^f3

<af6 5 i.b5+
White crosses Black’s plan of recap¬
turing on d5 with the knight.
5...C6
It is logical to dissolve the d5-pawn,
as otherwise White will support it
with c2-c4 and obtain the ascendancy
in the centre.
6 dxc6 <2^xc6
The alternative is 6...bxc6, when 7
18&xf6! ±c4 ^d5 8 0-0 M6 9 ^c2> ±e6 10
This is delightful butchery, but 18 ^e4 ±e7 11 ±b3 0-0 12 d4 ^d7 13
<5lxd5 *xd5 19 Sxf6 was a better *e2, planning 14 c4, proved good for
move order, e.g. 19...gxf6 20 .^xf6 Sa7 White in Spassky-Sakharov, Leningrad
21 .^xh8 and White is material up 1960.
with a strong attack.
18.. .<2^X65
Fritz, with the defensive sangfroid
of a computer, suggests the brilliant
defence 18...gxf6 19 .^xf6 .^f3!! Black
deals with the double threat of 20
<5lxd5 and 20 .^xh8 by buying time to
move the rook to safety. After 20 gxf3
Sg8+ 21 <^hl Sa7! Black can battle
on, though he is still worse. Of course,
there was little chance of the belea¬
guered Ziatdinov finding such a varia¬
tion over the board. 7 d4 i.d6
19 <2^xd5 gxf6 20 dxe5 ^f8 21 ^h1! A tricky alternative is 7...*a5+!.5 8
A quiet interlude in the middle of a <53c3 .^b4. In Belotti-Dutreeuw, Asti
r^ing attack. Black is defenceless, so 1995, Black obtained the advant^e
White takes a move to tuck his king after 9 *e2+ ±e6 10 0-0 0-0 11 *d3
away in the corner. .^d6, as the white bishop proved mis¬
21.. .ac8 22 ^xf6 *a5 23 *h5 ac4 placed on b5. Here the exchange 12

93
The King’s Gambit

Axc6 bxc6 would not solve White’s side pawns are slightly weak.
problems: it strengthens Black on the b) 10....fi.c7 11 c3 ^7 12 .fi.a4 b5!?
light squares by increasing his hold on 13 ^xh5 Wd5 14 ^a3 ^5 and Black
d5 and concedes the bishop pair. had attacking chances for the pawn in
Instead of 9 We2+, Gallagher gives 9 Kinlay-Nunn, New Malden 1977.
0-0 ^xc3 (maybe 9...0-0 is okay) 10 8.. .1.e6 9 ^g5
We2+ Ae6 11 bxc3 as clearly better for The consistent move. The other ag¬
White. However, after ll...Wxc3 12 gressive try is 9 ^5, but this worked
.^xf4 0-0 matters are far from clear: out badly in Hartston-Spassky, Hast¬
White has the initiative and the two ings 1965/66, after 9...0-0 10 ^xc6
bishops, but on the other hand his bxc6 11 .^xf4 ^d5 12 .^g3 f6 13 ^f3
queenside is weak and the d-pawn is Jixg3+ 14 hxg3 Se8 15 sfef2 (this looks
hanging. Black, meanwhile, has every horrible, but 15 0-0 Wb8!?, hitting
piece well entrenched. both b2 and g3, would have been very
unpleasant) 14....^f5 16 Wc4 ‘^?h8 17
^c3 ^e3 18 Wc5 ^g4+ 19 sfegl Wd7
and Black’s build-up quickly became
overpowering.
9.. .0-0 10 5lxe6 fxe6 11 .^xc6 bxc6
12 0-0
Not 12 lfxe6-i-?! <^h8 13 0-0 f3! with
an all-out attack on White’s king,
while White’s queenside is asleep.
12.. .51d5!?
Instead 12...Wc7 leads to a critical
position after 13 4ld2 e5! 14 dxe5 (14
8ire2-i- ^c4 e4 15 4lxd6 Wxd6 16 .^^4
This is the most ambitious move. A Wxd4-i- is good for Black, as the pawn
less risky alternative is 8 0-0, when on b2 drops) 14....fi.xe5 15 4lc4 .^d4-i-
play normally continues 8...0-0 9 16*hl^d5 17We4.
^bd2 (instead of this, Gallagher sug¬
gests that 9 c4 deserves attention)
9...Ag4 10 41c4 and now:
a) 10....^xf3 was played in Renet-
Van der Sterren, Budel 1987, when
White avoided some complications to
emerge with the better enc^ame after
11 Sxf3 Ac5 12 ^xc6 (12 c3 ^xd4!?
13 cxd4 Wxd4-i- 14 Wxd4 .fi.xd4-i- 15
‘^?fl a6 16 .^a4 b5 17 Sxf4 looks un¬
clear) 12...Wxd4-i- 13 Wxd4 .^xd4-i- 14
‘^?hl bxc6 15 .^xf4 and Black’s queen-

94
 Modern Defence (3 fhf3 d5)

Glaskov claims that White is better 15.. .Wxf8 16 Wxe3 e5! 17 ^c3 (not 17
here in view of the structural weak¬ dxe5? .^c5) 17...exd4 18 We6-i- 'i?h8 19
nesses in Black’s position. However, ^4 .^f4 20 Wxc6 Se8 and Black has
Gallagher continues 17....^c5 and sug¬ dangerous play for the pawn.
gests that Black has enough aaivity to Therefore, it seems that White
compensate for the weaknesses. I must made the corrert choice in the game.
admit that I would prefer to be Black 13.. .*h8 14 ^c3 ^xc3 15 bxc3 f3
here. The f-pawn has a strong cramp¬ This is the only way to maintain
ing influence on White’s position. the initiative.
Black will be the first to get a rook to 16 fixf3 fixf3 17 gxf3 Wf8 18 Wei
the open e-file, and the e3-square could fie8 19 Wf2 h6
become a strong outpost for a knight
or bishop.

A good moment to take stock.

White is two pawns up, which is a
13 «xe6-i-! serious material advantage. However,
According to established theory his kingside is fragile and the bishop
Black’s last move is bad because of 13 on cl has no effective squares. If
c4, attacking the knight. So what had White is to achieve a safe and promis¬
Korneev prepared? We shall investi¬ ing game he has to turn the inert
gate: clump of pawns in the centre into a
a) The immediate 13...Wh4? loses af¬ fighting force.
ter UWxe6+m7 15Wxd6. 20i.d2fie6 21 fiel?
b) 13...f3 is interesting, e.g. 14 A feeble move. White should seize
Wxe6-i- ^h8 15 cxd5 t2+ 16 ^hl Se8 the initiative with 21 c4, when after
17 Wg4 Sel 18 ^d2 cxd5 19 b3 Wa5 21.. .5f6 22 Sfl Wf7 23 d5! cxd5 24 c5
when White has an extra piece but is etc. he has some advantage.
tied up. Nevertheless, I don’t trust this 21.. .fife 22fif1 «f7 23«e2
for Black. This allows Black an outside passed
c) 13...^e3! 14 i.xe3 fxe3 15 Sxf8-i- pawn and good winning chances. It
(it seems best to deflert the black was better to play 23 a4, e.g. 23... «g6-H
queen from the d8-h4 diagonal) 24 ^hl Wxc2 25 a5.

95
The King's Gambit

23...Wxa2 24 We4 a5 25 <i>h1 a4 26 threatening to queen, as 38 Sxh6-(- can

i.e1 a3 27 i.g3 i.xg3 28 hxg3 Wb2 be answered by 38...'^?xh6, when there
29 «e8+ 'ih7 30 «e4+ *h8 31 is no 39 Wh4 mate - the g-pawn is in
We8+ <^h7 32 We4+ fig6 33 <i>g2 a2 the way.
34fih1!! 38We1?
White collapses under the pressure.
He had to continue to counter Black’s
threat to queen with further ideas of
perpetual check. Two moves suggest
themselves. First, 38 g5! when if
38.. .alW 39 Sxh6+ gxh6 40 We7+
forces perpetual as before. Black could
try 38...Wb5 instead of queening, but
then 39 f4 Wd5 40 '^?f3 should be okay
for White.
Alternatively, 38 Wf5! would step
beyond the obstructing g-pawn and
Just when the game seems to be reintroduce the idea of 38...alW 39
over, White discovers an amazing de¬ Sxh6-(- '&ch6? 40 Wh5 mate. So Black
fensive resource. If now 34...alW then would have to make do with a draw
35 Sxh6-(-! gxh6 (not 35...'^?xh6?? 36 with 39...gxh6 40 Wi7+ Sg7 41 WB+
Wh4 mate) 36 We7+ 2^7 37 We4+ '^?g8 42 Wc8+ '^?f7 43 Wd7+ and the
'^hS 38 We8-(- etc. and Black cannot only end to the checks is with
escape perpetual check. 43.. .‘i’g6?? 44 Wf5 checkmate.
34...Wa3 35 fia1 1^52 36 Sh1 Wb8! 38.. .Wxc2+ 39 ^g3 Wb3 40 Wal
Black finds the best winning at¬ fie6 41 fih2 Wb8-i- 42 <ih3 Wf4 43
tempt, attacking the g3-pawn and forc¬ fif2 fie2! 0-1
ing White to advance his g-pawn. A nice touch. After 44 Sxe2 Wxf3-i-
37 g4 Wb2 Black regains the rook and then puts
Back again, and this time really his queen on bl to force the win.

96
 Modern Defence (3 ^f3 d5)

Summary
After 1 e4 e5 2 f4 exf4 3 <53f3 d5 4 exd5 <53f6 White has to make a choice between
5 ±c4 and 5 ±b5+. If White opts for 5 J.c4 <53xd5 he should now play 6 ±xd5!
(see the notes to move seven, Game 32) with chances for a very slight edge after
6.. .«xd5 7 0-0 ±e7 8 d4 0-0 9 i.xf4 c6 (but not 9...c5?) 10 ^c3 «d8 11 «d2 etc.
On the other hand, 5 ±b5-(- c6 6 dxc6 <53x06 7 d4 is completely unclear (Game
34).

1 e4 e5 2 f4 exf4 3 ^f3 d5 4 exd5 ^f6

5^c4(D)
5 k.h5+ - Game 34
5.. .^xd5 6 0-0i.e7
6... J.e6 - Game 33
7dA(D) - Game 32

5kc4 6 0-0 7d4

97
 CHAPTER SIX I
Bishop and Mason Gambits
(3 ±c4 and 3 ^c3)

After knight. And thirdly, in several varia¬

1 e4 eS 2 f4 exf4 tions of the King’s Knight’s Gambit,
White has two other plausible ways White has to be ready to give up cas¬
to play instead of 3 ‘5if3: the Bishop’s tling in any case, so why should he be
Gambit 3 J.c4 and the Mason Gambit particularly afraid of 3...Wh4-(-?
3^c3. If the reader remains sceptical about
the merits of 3 J.c4, remember that
The Bishop’s Gambit (3 J.c4) the great Bobby Fischer ‘refuted’ the
The King’s Bishop’s Gambit has long King’s (Knight’s) Gambit (see Chapter
been unfashionable. Ordinary club 1), but nevertheless persevered with 3
players are probably frightened off by J.c4 himself. For Bronstein’s view on
the queen check at h4, which certainly the Bishop’s Gambit, see the Introduc¬
looks powerful at first glance. Interna¬ tion to this book.
tional players, on the other hand, are Here we shall concentrate on the
discouraged by the ease with which modern approach to defending the
Black can achieve ...d7-d5, exploiting Bishop’s Gambit, which involves
the position of the bishop on c4 to 3.. .c6 4 <53c3 <53f6 or the equivalent
gain time to open up the centre. 3.. .<53f6 4 <53c3 c6. White then has the
And yet several points can be raised choice of 5 J.b3, as played by Short in
in favour of the Bishop’s Gambit. Game 35, or the enterprising but risky
First, in the King’s (Knight’s) Gambit 5 d4 (Games 36 and 37). One move
3 ‘53f3, the bishop almost always goes earlier, Piket’s 3...c6 4 43c3 d5 looks
to c4 at some point, so why not play it dangerous, but Ivanchuk succeeds in
there immediately? Second, by delay¬ drawing its fangs in Game 38. Men¬
ing <5313 White takes the sting out of tion should also be made of 3...‘53f6 4
Black’s pawn advance ...g7-g5 and <5303 J.b4!?, a little-known but promis¬
...g5-g4, since it no longer attacks a ing idea for Black, which is examined

98
 Bishop and Mason Gambits (3 ^c4 and 3 ^c3)

in the notes to Game 36. Various Instead 3...<5316 4 43c3 c6 would

other ideas for the attack and defence transpose. This is the way that Games
are considered in the notes to Game 36 and 37 actually begin.
35. Here we shall consider some of
Black’s other possibilities:
The Mason Gambit (3 <^c3) and a) 3...d5 is often recommended as an
other Third Moves easy way to equalise after 4 J.xd5 43f6
T can only conclude that White is just 5 <5303 J.b4 6 ‘53f3 J.xc3 7 dxc3 c6 8
asking for it in this variation,’ wrote i.c4 'txdH 9 ^xdl 0-0 10 i.xf4 ^xe4
Joe Gallagher in Trends in the King’s 11 flel. However, I agree with Berry
Gambit. Nevertheless, the Mason that White’s two bishops and advan¬
Gambit (Games 39 and 40) has the tage in development should give him
element of surprise, and may catch out some edge. If White wishes he can
an opponent accustomed to rattling even avoid this variation with 4 exd5.
off 20 moves of a Spanish (Ruy Then 4...'53f6 5 <5303 c6 6 d4 cxd5 7
Lopez). For example, in Game 39 Ab3 would transpose to our main
Spassky’s enterprising opening proves game here. Alternatively, White can
too much for his solid, bookish oppo¬ try 7 J.b5-i- ‘53c6 8 J.xf4 .kd6 9 ^gel,
nent. However, as far as I can tell as in Bronstein-Tseshkovsky, USSR
Spassky never dared to repeat the Ma¬ 1978, though Bronstein claims that
son Gambit experience. Once in a life¬ Black can equalise after 9...J.xf4 10
time is enough. Two other third ^xf4 0-0 110-0 WhG etc.
moves for White are briefly consid¬ b) 3...'i'h4-i- (that scary queen
ered in the notes to Game 39. check!) 4 <^11 d5 (instead 4...d6 5 ‘53c3
i.e6 6 c6 7 ^f3 *67 8 d4 i.xc4 9
Game 35 ®xc4 g5 was played in Fischer-Evans,
Short-P.Nikolic USA Championship 1963/64, and
Wijk aan Zee 1997 now Keres recommends 10 h4! g4 11
<53el with some advantage to White) 5
1 e4 e5 2 f4 exf4 3 JicA c6 J.xd5 (Andrew Martin has champi¬
oned the move 5 exd5!?, claiming that
White has a clear advantage after
5.. .1.d6 6 ^c3 ^e7 7 <53e4 0-0 8 d4
<53d7 [8...‘53f5 looks better] 9 <53xd6
cxd6 10 Ab3. White has the two bish¬
ops and can expand with c2-c4)
5.. .J.d6! 6 ‘53c3 <5367 7 d4 f6 (this rules
out 8 e5 and prepares a kingside offen¬
sive that would justify the position of
the queen on h5) 8 ‘53f3 ®h5 9 ®el
<53bc6 10 <5362 g5 11 c4 with an unclear
position according to Estrin and Glas-
The King’s Gambit

kov, though 11...g4 looks strong for tioned that after 4 '53f3 g5 5 d4 J.g7
Black. White can also try 6 ‘53c3 (rather than
c) 3...‘5lc6!? is annoying for White, 6 c3) when 6...d6 7 ^d5, as in the
as the natural response 4 ‘53f3 trans¬ game Pillsbury-Schlechter, Vienna
poses to the Hanstein Variation of the 1903, is interesting.
King’s (Knight) Gambit after 4...g5 5 Assuming that White has no wish
d4 (bad is 5 h4 g4 6 ^g5 “^leS! 7 J.b3 to transpose to the Hanstein, then 4
h6 8 d4 hxg5 9 dxe5 J.g7, as in Mieses- d4 should be investigated. Korchnoi
Chigorin, Vienna 1903) 5...J.g7 6 c3 gives the sharp variation 4...<5316 5 e5
d6 7 0-0 h6. As was seen in Chapter 3, d5 6 i.b3 ^e4 7 i.xf4 lfh4+ 8 g3
this seems favourable for Black. Berry, ^xg3 9 i.xg3 'te4+ 10 ^f2 Wxhl 11
in an article in Chess Monthly, has sug¬ <53c3. Now he believes that the black
gested 7 Wb3 as an interesting way to queen is doomed, e.g. Il...<53e7 12 We2
avoid the transposition. He then gives h5 13 flel h4 14 i.f4 ^g6 (14...h3 15
the sacrificial continuation 7...We7 8 <53x113 Wxel-(- 16 '^xel Sxh3 17 ‘53xd5
^xg5 «xg5 9 i.xf7+ <^f8? 10 0-0 ^a5 looks good for White) 15 ‘^e3 <53xf4 16
11 «a4 ^xf7 12 i.xf4 «h5 13 i.xd6+ '^xf4 and after 16...g6 17 ^f3 Ah6-(-18
14 J.xc7. White will recover a ^g5 i.xg5+ 19 <^xg5 Sh5+ 20 ^f4
piece with 15 e5 or 15 J.xa5, after White wins. However, Berry (quoted
which he will have a couple of pawns from an article by Tim Wall in the
and attacking chances for one piece. British Chess Magazine) claims that
Even the Fritz computer, which is Black is better after the improvement
usually contemptuous of sacrifices, 16.. .g5-(-! 17 '^e3 (not 17 '^xg5 J.h6-i-
thinks that White is better here. 18 <^16 Sg8) 17...g4. This seems cor¬
However, it seems that Black’s play rect, e.g. if White tries to trap the
in the above variation can be greatly queen with 18 Axd5 then there fol¬
improved with 9...'^d8! Then 10 lows 18...i.h6-h 19 <^d3 i.f5-i- 20 <^c4
J.xg8 loses after 10...Wxg2 11 flfl i.e6! 21 lfxg4 i.xg4 22 i.xhl 0-0-0
Wxe4-(-, when going to the d-file leads and although White has a nice pawn
to mate, e.g. 12 <^d2 13 <^dl centre. Black’s extra exchange and the
14 ^c2 We4+! 15 ^d2 We2, two bishops give him winning
while if 12 '^f2 then 12...<53xd4! is chances. Perhaps 18 ‘53xd5 is best, e.g.
crushing. The consistent reply is 10 18.. .1.h6+ 19 ^d3 i.f5+ 20 <^c3 and if
0-0, planning 11 J.xf4. This also seems Black castles either way he loses the
bad, as after 10...Ah3 11 Sf2 <53xd4! 12 bishop on f5 to a fork. Nevertheless,
cxd4?! (White has to try 12 Wxb7, but Black would be undoubtedly better.
then 12...Sc8 looks better for Black as Therefore 3...^c6 seems a good try.
13 cxd4 Axd4 is still very bad for As Wall remarks, some practical
White and 13 Axf4 fails to 13...‘53f3-(-!) tests are required before a final verdict
12...i.xd4 13 «xh3 i.xf2+ 14 ^xf2 can be reached on the obscure varia¬
Wc5+\ Black picks up the bishop on cl tions examined above.
and wins. Finally, it should be men¬ 4^c3

WO
 Bishop and Mason Gambits (3 ^c4 and 3 ^c3)

In T.Wall-Ferguson, Rotherham some subtle play beats off the black

1997, White tried 4 '^el. That game attack: 9...i.xf4 10 ^xf4 fle8+ 11 ^fe2
went 4..Ae7 5 <5if3 d5 6 exd5 and ^g4 12 ^xd5! i.e6 13 h3! i.xd5 14
White emerged with the better hxg4 i.xg2 15 Sh2 i.f3 16 «d3 flxe2+
chances after 6...cxd5 7 Ab5+ 8 17 flxe2 J.xe2 18 Wxe2 and White
d4 <5if6 9 J.xf4. Instead Wall recom¬ stands better. This verdict was con¬
mends 6...<53f6 7 0-0 0-0 8 dxc6 ^xc6 9 firmed in a couple of correspondence
c3 J.g4 10 d4 J.d6, when White’s games by the King’s Gambit expert
fourth move is looking a bit silly. Steve Berry, one of which continued
4.. .'S3f6 18.. .Wh4+ 19 '»f2! «xf2+ (there is
This position is more often reached nothing better since f7 is attacked
through the move order 3...‘53f6 4 <53c3 twice) 20 '^xf2 with the better end¬

c6. game for White in Berry-Day, Corre¬

5i.b3 spondence 1974.

The more risky 5 d4 is the subject Berry believes that both 12...‘53h6 or
of Games 36 and 37, after the transpo¬ 13.. .<5306 could be improvements for
sition mentioned in the last note. Black, but neither of these seem par¬
5.. .d5 6 exdS cxd5 7 d4 .&d6 ticularly impressive, e.g. a possible line
The more double-edged 7....^b4 was after 13...'53h6 is 14 <53df4 J.xb3 15
played in Morozevich-Anand, Mos¬ axb3 «h4+ 16 ^d2 (not 16 g3 «xg3+)
cow (rapidplay) 1995. That game con¬ 16.. .<53c6 17 c3. Or if 12...<53c6, 13 0-0
tinued 8 ^f3 0-0 9 0-0 i.xc3 10 bxc3 may be a good answer (13...J.e6 14
«c7 11 Ifel ^c6 12 ^4? (better is 12 <53ef4). Fischer was famously accurate
<53e5! fle8 13 J.xf4 with unclear play - as an analyst, so it is difficult to disa¬
Wall) 7...<53e7 13 i.xf4 ^3 14 ±d2 gree with him!
Wc7 15 <53e5 <53f5 16 ^14 l.e6, when
Black has kept his extra pawn and
should win.
8 «3f3
A critical moment. The old move is
8 ‘53ge2, planning to regain the pawn
on f4. Then Spielmann-Bogolyubov,
Marisch-Ostrau 1923, continued 8...0-0
9 0-0 g5! 10 ^xd5 ^c6 11 c3 ^xd5 12
±xd5 ‘53e7 13 Ae4 f5 with advantage
to Black.
However, this is not the end of the
story. Fischer analysed this variation 8.. .«3c6 9 0-0 i.e6 10 «3g5 h6
in the American magazine Chess Life This weakens the kingside. The
(April 1964) and concluded that after simple 10...0-0 was better, or perhaps
8.. .0.0 White can snatch the pawn 10.. .flc8, threatening ll...‘53xd4. In
back immediately with 9 J.xf4. Then either case Black would have had a

101
The King’s Gambit

sound position, though White would fixes the weak pawn on b3.
have some advantage in view of the 19«g6«e7
two bishops after 11 <5^xe6. Necessary to prevent 20 Sf7.
11 ^xe6 fxe6 12 i.xf4 i.xf4 13 20 Wh5
Sxf4 0-0
The immediate 13...''ii^d6 was better,
interfering with the smooth develop¬
ment of White’s game. Then after 14
®d2 0-0 15 flafl the white queen
would be on a less threatening square
than in the game. The sacrifice 14
flxf6 would be unsound.
MWdSWde 15 Safi
White completes the mobilisation
of his pieces. He has the advantage in
view of the weaknesses in Black’s
pawn structure, in particular the If the black knight were on d7 in
backward pawn on e6 and the hole on this position (see move 15) Black could
e5. Furthermore, if he can bring his now play 20...b5, restraining any c3-c4
bishop on b3 into active play then it breakthrough by White and gaining
will prove the best minor piece. play on the queenside. However, as
15.. .'S3h7? things stand in the game the e5-square
Black wants to lessen White’s pres¬ is undefended, which means that
sure on the kingside and therefore cor¬ White could respond 21 We5!, plan¬
rectly offers the exchange of rooks. ning 22 Sxf8-(- followed by <53f4 etc.,
However, the move chosen decentral¬ with a clear advantage.
ises the knight and, as will be seen, is 20.. .^g5
the prelude to an incorrect plan. He The knight heads for the outpost
should play 15...‘53d7, keeping the square on e4, but, as the last note indi¬
knight in the centre and keeping cates, the correct role for this piece
watch over the e5-square. would have been the defence of the e5-
16 ^e2! square. An interesting alternative was
White defends his rook and clears 20.. .Wd6, keeping up the fight for e5,
the way for 17 c3 and 18 ±c2, with a since after 21 Sf7?! ^g5! 22 axb7? (22
winning attack. Black’s reply is there¬ flxf8+) 22...Sxfl+ 23 ^xfl «c6 24
fore forced. Se7 ‘^f8! Black would win a rook.
16.. .«3a5 21 h4 ^e4 22 «e5
He must eliminate the white bishop White’s queen dominates the centre
at the first opportunity. and pressurises the e6-pawn. Its power
17 c3 ^xb3 18axb3 a5! far outweighs the knight on e4 and it
A good move which prevents cannot be challenged as 22...Wd6?
White gaining space with 19 b4 and loses to 23 fixe4.

102
 Bishop and Mason Gambits (3 ^c4 and 3 ^c3)

22.. .5f6 23 Sxf6 gxf6 #b3+.

This solves the problem of the e5- 37...<i'c7 38 #a5+ <i'c6 39 'i'g2 h5
square but weakens the kingside. If 40 #a8+ 'i'd6 41 #d8+ 1-0
Black could maintain a centre with Black resigned, as he loses another
pawns on d5, e6 and f6 he would have pawn after 41...'4’c6 42 #e8+ '4’d6 43
a good game, but unfortunately for #xh5. The knight is still sitting pretty
him his knight can be undermined on e4, but what did it do?
with c3-c4, forcing the dislocation of Short’s opening choice proved a
his centre. Perhaps the passive great success, since Nikolic failed to
23.. .‘5ixf6 24 ‘53f4 Se8 would have find the correct middlegame strategy.
been a better defence.
24 *f4 <ih7 25 c4! Game 36 j
This flanking blow reduces Black’s Westerinen-A.Kuzmin
centre from a compact mass into a Moscow 1989
litany of isolated pawns and weak
squares. 1 e4 e5 2 f4 exf4 3 i.c4 'SlfO 4 4103
25.. .5g8 26 cxd5 exd5 27 Sc1 #e6 c6
According to Short the best defence A critical alternative is 4...J.b4!?
was 27...^d6\, when 28 ‘53g3 flg6! is Then McDonald-Law, British Cham¬
only slightly better for White. Taking pionship 1997, went 5 e5 d5! 6 J.b5+
this analysis further, immediate action c6 7 exf6 cxb5 8 fxg7 (perhaps this
by White would now allow Black to should wait) 8...flg8 9 #e2+ Ae6 10
escape with a draw. For example, 29 <53h3?! (this unusual move looks infe¬
^f5 We2 30 ^e3 ^f5!! 31 Sc7+ <4‘h8 rior to 10 4113, but Black was doing
32 Wxf5 Wxe3+ 33 '^fl ®e4! and very well after 10...'53c6 11 d4 #f6 in
White has nothing better than a draw Chandler-Emms, London 1997)
with 34 flc8+ ‘^h? 35 flc7+ <4‘h8 36 10.. .'»h4+! 11 #f2 '»xf2+ 12 ^xf2
Sc8+ etc. However, the simple 29 (forced, as 12 ^xf2 .ic5+! 13 ^f3
'^h2! would leave Black facing a most flxg7 is bad) and now the simple
unpleasant defence. 12.. .flxg7 must be good for Black.
As played White breaks through 5d4?!
and begins to pick up the loose black It may seem odd to criticise such a
pawns. natural move, but this allows Black to
28 Sc7+ Sg7 29 Sxg7+ ^xg7 30 develop his bishop aggressively to b4,
#c7+ <if8 31 «3f4 itfB 32 #58+ when White already has to start think¬
<ie7 33 #xb7+ 'id6 34 b4! ing about how to save the game!
An excellent way to clarify the posi¬ 5.. .1.b4!
tion. If34...'»xf4 35'»b8+. Undoubtedly the best move. After
34.. .axb4 35 «xb4+ <*>06 36 #a4+ 5.. .d5 White can play 6 exd5 cxd5 7
^^b6 37 g3! J.b5+, when the bishop is more active
And this consolidates the kingside, than it would be on b3 (this is the rea¬
as 37...‘53xg3 loses the knight to 38 son why Westerinen prefers 5 d4 to 5

103
The King’s Gambit

±h3). g4 19 'tf2 i.xe4 20 Sxe4 WfS 21 Sel

6e5 ^d7 was better for Black in Wester-
More or less forced, in view of the inen-Hector, Gausdal 1989. The white
threat to e4. Note that if White had bishops are ineffective, the white king
played 5 ^b3 rather than 5 d4 he faces a dangerous onslaught from
could now have answered 5...^b4 Black’s mass of pawns, and the white
with 6 e5, when the knight has no rook on hi is much harder to bring
good square, since e4 is inaccessible. into the game than the black rook on
Here things are different since after a8, which can enter the fray after
6...?3e4! ...'^c7. Nevertheless, after a hard
Black is threatening both 7...^xc3 struggle, Westerinen won this game!
and 7...Wh4+. In the other excerpt he faced similar
7*f1!? problems, but this time was less fortu¬
nate.
b) 10 'te2-(- ^f8 11 ^f3 i.g4 12 'te4
i.xf3 13 'txf3 'tc7 14 ±d2 ^d7 15
^gl c5! 16 i.d5 Sb8 17 Sfl ^f6 18
i.e4 g5! 19 i.f5 Sg8 20 #12 h6 21 h4
#c6 22 #f3 (in view of Black’s build
up on the kingside. White feels obliged
to offer the exchange of queens, which
shows his game has entirely gone)
22.. .#xf3 23 Sxf3 and White had no
compensation for his pawn in Wester-
inen-Adams, Manila Olympiad 1992.
This is a Westerinen speciality. 9i.a3
There are two alternatives. First, 7 White has to play for the attack, at
Wf3, which turns out badly for White whatever cost in material. The alterna¬
after 7...d5 8 exd6 0-0 9 ^ge2 'th4-(- 10 tive 9 Sbl d5 10 exd6 0-0 would be
g3 ficg3 11 hxg3 Wg4 12 Wxg4 .^xg4 very bad for him. Now at least Black’s
13 .^d3 Se8 (Keres). Second, 7 Wh5, king will remain in the centre.
which is considered in the next game. 9.. .b5!
7...?3xc3 8 bxc3 Axc3 The best move, but two other
Probably not best, though Black moves are worth considering:
seems to have a guaranteed draw. Af¬ a) When I first saw this position I
ter 8...d5! 9 exd6 .^xd6 two excerpts thought that Black could refute the
from the Finnish Grandmaster’s attack with 9...d5, e.g. 10 exd6 Axal
games demonstrate the problems that 11 #e2-(-(?) .^e6 12 .^xe6 0-01! and
White faces: wins. However, Tim Wall pointed out
a) 10 #13 'tf6 11 ^e2 i.e6 12 i.b3 that 11 d7+\ instead of 11 #e2-(-
g5 13 Sbl i.f5 14 i.a4 'te7 15 ^g3 looked dangerous. When we analysed
i.g6 16 i.d2 ^d8 17 Sel 'td7 18 ^e4 this move with Luke McShane on a
 Bishop and Mason Gambits (3 Jlc4 and 3 ^c3)

train to Sandwell and Dudley, we bishop on a3 and thereby securing the

came to the conclusion that after right to castle. Black would then be
ll.-.'txd/ (ll...<^xd7 12 '»g4+) 12 winning, since nothing would remain
Wxal Black is in serious trouble, e.g. of White’s attack on the king.
12.. .b5 13 d5!?, attacking g7, or 12...b6
(intending 13....^a6) 13 '^f2!? Aa6 14
.^b3 and White is ready to develop his
kingside with ‘5lf3 and Sel etc. Black’s
king looks very vulnerable.
Instead of the greedy 10....^xal
Black could try 10...0-0, but then 11
Sbl b5 12 d7 wins the exchange and
looks good for White.
b) Another alternative to the game
move is the immediate 9....^xal, as in
Rahman-Formanek, New York 1993.
This seems wrong, since after 10 .^d6 10i.d6!!
b5 11 Wg4 g6 it is difficult to see how The only move but nevertheless a
Black can defend against the simple 12 striking one. The bishop avoids being
Ab3!, threatening 13 Wxf4 and then shut out with 10...b4 and clamps
mate on f7. For example, 12...a5 13 down on the d-pawn, making it im¬
«xf4 Sf8 (if 13...f6 14 exf6 and the f- possible for Black to free his game
pawn kills Black) 14 Wh6! (Black was with ...d7-d5. It seems highly unlikely
hoping for 14 AxfS d5, though this that Black will ever be able to remove
should also lose after 15 exd6) 14...Sg8 or challenge the bishop, since his own
15 Wxh7 Sf8 16 Wg7 and wins. Or if dark-squared bishop is a long way
12.. .f5 13 Wxf4 (threatening 14 Wh6 from the kingside and none of his
and 15 'B^g7) 13...h6 14 ^h3! (not 14 other pieces can easily approach the
Wg3 'Vfg5) 14....^a6 (what else?) 15 d6-square. This means that Black can¬
Wg3 g5 16 W{3 (threatening mate) 16... not hope to secure the right to castle
g4 17 Wxf5 and wins. In the game and, as will be seen. White’s queen can
Rahman played the weaker 12 Wxf4, join in the attack and seriously
when 12...bxc4 13 Wh6 We7 (the only threaten the black king.
defence against 14 WgT) 14 Axe7 ‘4’xe7 10...bxc4
led to a strange material balance. I The bishop was the more threaten¬
imagine that White is at least equal ing piece; hence Black captures it be¬
and maybe much better. However, in fore the rook. An interesting alterna¬
view of the strength of 12 .^b3 this tive was 10...h5!?, depriving the white
position is only of curiosity value. queen of its natural attacking square
Black’s idea in our main game is on g4 and preparing the exchange sac¬
more sensible. He plans to answer 10 rifice ...Sh6 and ...Sxd6 in some lines.
Ah3 with 10...b4, blocking off the If White loses his nerve with 11 Sbl

" 1^
The King’s Gambit

then he gets a bad game after ll...bxc4 enough counterplay to force a draw.
12 ^xb8 d5! 13 exd6(?) ^g4! etc. So he ihe2 i.xd4 15 ^xd4
13...#66 14
has to continue in enterprising style '»xd4^6Wg^m^+ M *f2 «xc2-(-
with 11 .^b3 .^xal 12 <^f3. Now if i8*fi
Black tries to save his bishop then he White cannot evade perpetual
falls under a decisive attack, e.g. check, e.g. 18 ^f3 'td3-(-19 ^g4 'tf5-(-
12.. .1.c3 13 M b4 14 WfS Sf8 15 20 ■^h4?? (20 <^f3) 20...'»h5 mate. A
^g5 .^a6+ 16 ‘4’f2 .^xd4+ 17 ‘4’f3 wins most unusual game.
for White. But Black has a better de¬
fence which involves bringing
knight to e6 to paralyse White’s at¬
the I Game 37
Westerinen-Pakkanen
tack: 12...^a6! 13 ®d3.> ^c7 14 WfS Helsinki 1992
^e6 and the knight on e6 blocks the
attack. So is best answered by 1 e4 e5 2 f4 exf4 3 ^cA- ^f6 4 ^c3
13 Wxal, when 13...<2^c7 14 ‘4’f2 c6 5 d4 i.b4 6 e5 ^e4 7 «h5
is not at all clear.
11 #94 g6
Virtually the only move, as ll...Sg8
12 ^3 i.xal 13 '»xh7 218 14 '»xg7
wins. Also, ll...g5 fails to 12 <2^h3!,
planning 13 'S3xf4 gxf4 (else 14 <2^h5
and 15 destroys Black) 14 Wg7.
The game Rut-Connors, Correspon¬
dence 1989-91 (did it really take them
three years?), continued 12....^d2 13
^f2 c3 14 ^e4 i.a6-(- 15 ‘4’f2 l.e3+ 16
^el ^6 17 '»xg5 l.d2+ 18 ?3xd2
cxd2-i- 19 ‘4’dl and Black resigned. A clever idea: White rules out one
These variations reveal the theme of nasty check (7...Wh4-t) and threatens
White’s onslaught. He wants to attack one of his own on f7. If Black re¬
the black rook from g7, when it will sponds with 7...0-0 then 8 ‘5lge2, in¬
have nowhere safe to go. Then Black tending 9 0-0, looks good for White.
will not only lose his rook but will Black’s reply in the game is therefore
also be mated, since his king has no critical.
way to escape from the back rank. 7.. .g6! 8 #13
Here 8 Wh6?! looks a little too far¬
This seems better than 12 Wxf4, fetched even for Westerinen. If Black
with the same idea of Wh6 and Wg7, accepts the offer immediately with
since the bishop on c3 is attacked. 8.. .<2^xc3 then there are wild complica¬
12.. .1.xa1 lal^he tions, e.g. 9 bxc3 .^xc3-i- 10 '^dl ixal
Now White to move would win 11 ^7 nf8 12 i.a3 d6! 13 ±xd6 ^d7
with 14 Wg7. However, Black has just 14 i.xf8 ^xe5! 15 i.c5! l.g4-(- 16 ±e2

106
 Bishop and Mason Gambits (3 Ac4 and 3 ^c3)

±xe2+ 17 ^xe2 ^g4 18 '»f8+ 19 this is the best White can do.
Wxf7+ ^c8 20 We6+ Wd7 21 '»g8+ 10 hxg3!
'ifd8 22 Wfe6+ with a draw! However, Now everything goes smoothly for
the usual antidote in such positions White.
seems very good for Black: 8...d5! ru¬ 10...'txh1 11 i.xf4i.xc3
ins White’s plans.
8.. .«h4-t-
If Black now plays 8...d5 then we
are following Keres’s analysis to 7 Wf3
in Game 36, the sole difference being
that the black pawn is on g6 rather
than g7. Can White exploit this? The
answer seems to be ‘Yes’, as after 9
exd6 0-0 10 ^ge2 'th4-(- 11 g3 fxg3 12
hxg3 Wg4 White has the strong move
13 'i!e3! with the possible variations:
a) 13...^f6 14 ±62 i.xd6 15 0-0-0
with attacking chances. Black is defenceless. If 11...0-0 12
b) 13...Se8?! 14 axh7! ‘^xh7 15 <2^e4, planning 13 <2^f6-(- etc., when
±xf7 (threatening mate on h6 and at¬ both the black king and queen will be
tacking the rook on e8) 15...'i!h5 16 in danger of being trapped.
i.xe8 WhU 17 ^gl i.f5 18 i.d2 12 i.xf7-(-! *xf7 13 e6+\
±xd6 19 0-0-0 with a clear advantage Played in Morphy style.
to White. Both of these variations 13...<4>xe6
would be impossible if the black g- If 13...dxe6 14 i.d6+ ‘^g7 (14...‘^e8
pawn were still on g7. 15 Wfb) 15 i.e5-(- ■^g8 16 ^6 Wh6 17
9 4>f1! Wxh8-i- ‘4’f7 18 Wxc8 destroys Black.
This is better than 9 g3?, when after 14 .i.e5! «h5 15 «f6-(- 4>d5 16
9.. .fxg3 10 .^xf7-(- (no better is 10 «d6-(-^e4 17 bxc3! Sf8-(-
Wxf7+ ‘^d8) 10...-^e7 11 hxg3 Wxg3-(- There is no answer to 18 flel-l-.
12 Wxg3 ^xg3 13 ah3 ‘^xf7 14 axg3 18 i.f4 Sxf4-(- 19 '»xf4-(- ^d5 20 g4!
d5 Black achieved a winning endgame 1-0
in Westerinen-Ernst, Helsinki 1991. A nice finishing touch to a very
9.. .^g3-(-? pretty game. If the black queen moves
A serious mistake. In his analysis of to safety, it is mate in two.
the Westerinen-Ernst game mentioned
in the previous note, Ernst recom¬ Game 38 ^
mends 9...d5! 10 exd6 <2^xc3 11 bxc3 Ivanchuk-Piket
i.xd6 12 '»e4-t We7 13 Wxe7-(- ■^xe7 Linares 1997
14 ^e2, which he assesses as equal. I
think that Black has an edge. Anyway, 1 e4 e5 2 f4 exf4 3 ±c4 c6 4 <2lc3
it is clearly a waste of the first move if d5

107
 The King’s Gambit

queen plans to drop back to either f4

or g3 to shelter the king. For example,
after 10...^bd7 the reply is 11 Wg3.
Black can then either agree to an ex¬
change of queens, when the endgame
is much worse for him (why is ex¬
plained below) or retreat his queen,
when he loses valuable time.

A very aggressive approach, but al¬

most certainly inferior to 4...<2^f6.
5 exd5 'th4+ 6 -S-fl f3!?
Short admitted that he was ‘more
than a little concerned about this idea’
when he essayed the Bishop’s Gambit
against Nikolic (see Game 35). Indeed,
at first sight it seems very strong: the
charging pawn uncovers an attack by 10...'tf2+?
the queen on the bishop on c4 and Short recommends 10...Wg4-(-, but
prepares to almost completely denude White is better after 11 ^3. The
the white king with ...f3xg2-(-. Never¬ game move is rather defeatist: Black
theless, as the Russian proverb says forces an endgame where White no
‘one man in the field isn’t an army.’ longer has to worry about his exposed
Black’s only active piece is his queen king. In fact. White has excellent win¬
and it is against the logic of chess for ning chances due to his enormous lead
White to suddenly find that he has a in development. All the white pieces
bad position. Although of course chess can be quickly mobilised and brought
isn’t always a logical game...! to key points in the centre. Mean¬
7 d3 fxg2-t 8 *xg2 ^f6 9 We2+\ while, it will be a long time before the
This is the first indication that all is black rook on a8 will have any bear¬
not well with Black’s position. He ing on the game.
now has to give up his castling rights, 11 'S>xf2 -53g4-(- 12 >S>g2 ^xe5 13
as after 9...^e7 10 d6 wins a piece. i.f4 ^g6?!
9.. .<^d8 10 #65! According to Short, Black’s best
Ivanchuk finds an excellent way to chance was 13...<2^xc4, removing one
solve the problems of the position. of the dangerous attacking pieces.
The queen takes control of the centre Nevertheless, one can sympathise with
and rules out Black’s developing move Piket, who clearly did not like the idea
10.. ..^d6. Furthermore, if attacked the of being left with all his pieces on the

108
 Bishop and Mason Gambits (3 Ac4 and 3 ^c3)

back rank after 13 moves! Black also loses a knight after

14 i.g3 f6 15 ^f3 i.b4 16 ^d4 Ad7 34...<^xd6 35 ±g3+.
17 ^e6+! 35 Sxc7-t ^d8 36 Sf7 1-0
A wise transaction. White acquires
the two bishops and a gigantic passed Game 39
pawn. Spassky-Furman
17.. .1.xe6 18 dxe6 Tallinn 1959
White’s opening surprise has been a
marvellous success. Piket knows a 1 e4 e5 2 f4 exf4 3 ^c3
huge amount of modern opening the¬
ory but has been completely unable to
adjust to the demands of this archaic
gambit.
18.. .^e7 19 Shfl ^c8 20 ^e4 'S>e7
21 i.h4
Threatening 22 iS^xfb!
21.. .5.8 22 c3 i.d6

White’s other third moves can be

dismissed quickly:
a) 3 Wf3 turned out badly for White
in Paoli-Prins, Venedig 1949, after
3.. .^c6! 4 '»xf4 i.d6 5 '»e3 ?3f6 6 i.e2
We7 7 ^c3 .^c5, when Black already
had a clear advantage.
b) 3 M.e.2 was played three times by
23 4>h1! Tartakower at the great New York
A clever retreat, clearing the g-file 1924 tournament and achieved a
for an attack with the rooks. highly creditable lVi/3 score against
23.. .b5 24 i.b3 ^a6 25 a4! Capablanca, Alekhine and Bo-
White plans an attack on both sides golyubov. pavid Bronstein contends
of the board and in the centre. Black’s that Tartakower was the greatest
pieces are so disorganised that he can¬ player of all time since he could play
not resist an onslaught on such a wide any opening successfully!) However, 3

front. .^e2 doesn’t stand up to modern ana¬

25.. .^c7 26 axb5 cxb5 27 d4 a5 28 lytical scrutiny. According to Csom,
af3 a4 29 ±a2 Sa6 30 Sgl ^e8 31 Black can get the advantage with
Sf5 a3 32 Sxb5 g5 33 Sb7-t ^c7 34 3.. .d5!, e.g. 4 exd5 ^f6 5 ^f3 ‘^IxdS 6

^xd6 axd6 c4 'S3e7 7 d4 ^g6 8 ^c3 i.d6 9 h4 h5

109
 The King's Gambit

10 ^e4 i.f5. opens lines for his pieces as a prelude

Spassky’s choice of the Mason to a direct attack on White’s king. In¬
Gambit in our illustrative game was stead 4...d6 5 ^f3 .^g4 6 d4 etc. would
good psychology, despite its dubious transpose to the Barle-Portisch game
reputation. Grandmaster Furman had of the last note. However, 6 <51(15! is
made a deep study of well-known better, when 6....^xf3-(-?! (6...Wd8!? -
opening variations. If something was Korchnoi) 7 gxf3 ‘4’d8 8 d3!, as in
going to unnerve him, this was it! Keres-Kunerth, Correspondence 1936,
3.. .'»h4-(- is good for White according to
This check is much more disruptive Korchnoi.
here than after 3 .^c4, since the white 5 5lxd5 i.d6
king is forced to go to e2, where it After 5....^g4-(- 6 <5lf3 Black could
shuts in the bishop on fl. Since the transpose back into our game with
white king’s bishop is often the magic 6.. ..^d6 7 d4. However, he has the
wand for White’s attack, this is not a additional option of 6...<5lc6!.>, which
good state of affairs. Furthermore, the is discussed in the next game.
entombed white bishop denies the 6 d4 i.g4+ 7 5lf3 <5lc6
king a shelter on fl. Black can also consider two other
Also possible was 3...<5lc6, when 4 knight moves:
d4 ^4-1- 5 <^e2 d6 6 ^f3 i.g4 7 i.xf4 a) 7...5lf6 8 <5lxf6-(- gxf6 9 c3 was
(or 7 ?3d5 0-0-0 8 ‘^d3 Whb 9 i.xf4 played in Bronstein-Alatortsev, USSR
Whs - this square is safe now that Championship 1945. After 9...1.xf3-i-
<2^xf4 is impossible - 10 c4 fS! and 10 gxf3 c5 11 dxc5 .^xc5 12 Wei
White’s king was looking uncomfort¬ (stopping 12...Wf24-) 12...Wg5 13 Wd2
able in Kavalek-Stein, Tel Aviv 1964) Wh4, Bronstein made a winning at¬
7.. .0.0-0 proved good for Black in tempt with 14 Wxf4, when after
Barle-Portisch, Portoroz 1975. White 14.. .Wf2-t 15 ^dl ^c6 16 ±d2 Sd8
tried 8 ‘4’e3, but Black gained a strong Black had a dangerous attack. How¬
attack after 8...Wh5 9 i.e2 g5! 10 'S3xg5 ever, Bronstein won on time at move
^f6 11 h3 i.xe2 12 Wxe2 Wg6 (with 33. Since in our main game Furman
the threat of 13...^h5) 13 d5 loses on time after 31 moves, this
(something has gone wrong for White really is a tricky opening to face!
if he has nothing better than this anti- b) Another alternative for Black is
positional move) 13...^e5 14 ^f3 7.. .<5le7. Then theory gives 8 ^xe7?
.^h6! etc. Wxe7 9 e5 f6 10 ±xf4 fxe5 11 dxe5
However, it is worth checking on <5^c6, as in Ashikhin-V.Zhuravlev,
h4 as soon as possible, because after Yurmala 1964, with a big advantage to
3.. .^c6 White is given the chance to Black. White’s opening has failed if he
chicken out of the Mason Gambit has to exchange knights on e7 unless
with 4 <2^f3! he gets some large compensating ad¬
4 4>e2 d5 vantage. Instead, 8 <5lxf4 looks critical,
The most energetic response. Black when violent attacks by Black seem to

tro
 Bishop and Mason Gambits (3 Ac4 and 3 ^c3)

fail, e.g. 8...i.xf3+? 9 <4>xf3 g5? 10 g3 quickly defeated: 10...Wh6 11 ^xe7+

'»h6 11 ^h3 '»h5+ 12 g4 Wg6 13 Axe7 12 c3 f6! 13 e6 (trying to keep
<5lxg5. the centre blocked; Black would have
had a huge attack after 13 exf6 .^xf6
and 14...ahe8+) 13...f5 14 Wa4 Wxeb-t
15 ‘4’f2 Wh6! (now there is no good
answer to the threat of 16...f4) 16 ‘4’gl
ahe8 17 ±h5 ±xf3 18 gxf3 '»e3+ 19
<^g2 '»d2+ 20 ±i2 ±h4 21 Shfl ad6
22 Sadi '»f4 23 i.xh4 '»xh4 24 i.xc6.>
(a blunder but in any case Black’s at¬
tack is overwhelming) 24...Se2-(- 0-1.
10...^f5?
Bewildered by a multitude of possi¬
bilities, Furman goes wrong. After the
game, he claimed that 10...Ab4! would
have been very strong. This seems
play 8 c3!, when things are not at all correct, e.g. 10....^b4 11 a3 (11 g3 Wh5
clear. For example, 8 c3! 0-0-0 9 ‘4’d3 is bad for White) ll...^xd5 12 cxd5
and now: axd5 13 ■^e3 (forced) 13...i.xf3 14
a) 9...'th6?! 10 ^c2 ^ge7 11 ^xe7+ Wxf3 g5! etc.
^xe7 12 ±63 (not 12 e5 i.xe5). 11 exd6 ^fxd4-(- 12 4>d3!
b) 9...'»h5 10 -^02! (avoiding 10
^xH i.xf4 11 i.xf4 ^f6 12 h3 lhe8)
10.. .^f6 (10...f5?! 11 e5!?) 11 ^xf6
gxf6. In both cases with unclear play.
8.. .0.0-0 9 i.xf4
It is doubtful that White’s king
could survive the attack after 9 exd6
lxd6 10 c4 ^f6!, e.g. 11 i.xf4 and
now;
a) ll...ae8+ 12 ‘4’d3 ^xd5
(12...i.f5-(- 13 ■^d2 'tf2-(- 14 -^cl) 13
cxd5 .^xf3 14 Wxf3 Hxd5! or, perhaps
better, With some precise moves Spassky
b) ll...'S3xd5!, e.g. 12 Axd6 He8-(- 13 demonstrates that Black’s piece offer is
±e5 ^xe5 14 dxe5 IxeS-i- 15 ‘4'd2 unsound.
Wh6+ 16 ‘^c2 ^e3+. 12...'th5 13 ±e2 ^e6 14 i.g3 cxd6
9.. .^ge7 10 c4 15 b4 She8 16 Sel ^c7 17 ^c3
In Lyell-Flear, British Champion¬ Wh6 18 I'd ?ixd5-(- 19 cxd5 Se3-(-
ship 1989, White tried to improve 20 i.d3 Wf6+ 21 4>c2 Sxel 22
with 10 Ag3. However, he was i.xe1 i.xf3 23 dxc6 i.xc6 24 i.c3

111
The King’s Gambit

l'f2+ 25 i.a4+ 26 <^b2 l'h4 27 i.xh6 She8 14 ^d3 WhS 15 '»f4 i.f5-t
i.xg7 ^b8 28 g3 l'g4 29 i.f6 Sc8 16 ^d2 Se4 17 ^3 'txh6-(- 18 ^dl
30 Sc1 Se8 31 b5 1-0 ^xd4 19 cxd4 Sdxd4-(- 20 <5lxd4
Sxd4-(- etc.) 8...f5 9 '^e3 .^xf3 10 gxf3
Game 40 ^f6 11 <5lxf6 2xd4 12 Wei Wxf6 with
C.Horvath-J.Horvath a clear advantage to Black.
Budapest 1995 b) However, 7 c3 is an interesting
attempt for advantage, when the best
1 e4 e5 2 f4 exf4 3 ^c3 '»h4+ 4 reply is 7...0-0-0 (7...f5 is also worth
'S>e2 d5 5 ^xd5 i.g4+ 6 ^f3 ^c6!? investigating, e.g. 8 ^xc7-(- ‘4’d7 9
<2^xa8 fice4. Note that after 7...<5le5 8
d4! Black cannot win the white queen
by capturing twice on f3 followed by
...Wh5-(- and ...Wxdl, as there is a
bishop check on b5 followed by Sxdl,
regaining the queen) 8 Wei Wxel-i-
(8...Wh5 9 <5lxf4 We5 10 d3 f5 is un¬
clear) 9 ‘4’xel Se8 (or 9...f5 10 d3 fxe4
11 dxe4) 10 d4 Sxe4-(- 11 '^f2 with
level chances - variation by Glaskov.
7...<4>d8 8 ?3xa8 ^e5
Here 8...^d4-(- 9 Wf6!? is a
Black’s most aggressive response, tricky alternative which was intro¬
gambiting the rook on a8 for an at¬ duced in the game Jago-J.Littlewood,
tack. The drawback to this idea is that Correspondence 1964-65. That game
the line has been more or less worked went 10 c3 (what about the calm 10
out to a forced draw, when in fact i.e2!?, e.g. 10...i.c5 11 c3 Wa6-(- 12 c4
Black should be looking for more than Wd6 13 <2^xd4 Wxd4-(- 14 '^c2 Wxe4-(-
a draw after the reckless 3 ^c3. Nev¬ 15 d3 Wxg2 16 i.xf4 ±xe2 17 Wd2
ertheless, I don’t think the draw ver¬ ^f6 18 Sael Se8 19 ^c7+ and White
dict of theory was a problem for Josef wins, though of course this is by no
Horvath, who was playing his brother means the whole story) 10..Wa6-(- 11
Csaba here and seems to be in no c4 .^c5 12 b4 <5lf6 13 bxc5 ‘5lxe4 14
mood for fratricide. Wei Se8 and now the game went
7 ^xc7-(-
see following diagram
White does best to accept the offer:
a) 7 d4? works out badly after 15 Wxe4 Sxe4 (if 15...iLf5 16 Wxf5
7...0-0-0, e.g. 8 i.xf4 (8 c3 f5 9 WcB ^xf5 17 i.b2 Wg6 [17...<^d7 18 c6+
<5^f6 10 ^xf6 gxf6 11 .^xf4 fxe4 12 bxc6 19 <5le5-(- '^c8 20 g3, planning
Wxe4 .^h6 gave Black a winning posi¬ .^g2 or Ah3, and White should win]
tion in Keres-Kunerth, Correspon¬ 18 '^c3 and the white king escapes the
dence 1936. A possible finish is 13 attack) 16 '^xe4 <2^xf3.

112
 Bishop and Mason Gambits (3 Ac4 and. 3 ^c3)

9.. .Axf3+
Although 9...^xf3? 10 hxg4 <2^gl-(-
11 Wxhl 12 c3 f3 13 gxf3 ^xf3 14
‘4’c2 would be good for White,
9.. ..^h5!? deserves attention. The criti¬
cal variation is then 10 d4 <2^xf3 11
gxf3 ±xi3+ 12 ■^xf3 'th5+ 13 ■^g2
Wxdl 14 i.d3 ^5 15 i.xf4. The
game Arkhipkin-Klovan, Riga 1974,
continued 15...'S3e7 16 flhfl ^g6 17
.^g3 ^e7. Bangiev claims that this is
unclear or perhaps slightly better for
Now Little wood recommends 17 Black. Certainly this is an interesting
<2^b6! as unclear, though I suspect that material balance.
despite his exposed king the rook and
two bishops give White the better
chances after the plausible 17...axb6 18
gxf3 bxc5 19 fxg4 Wc6+ 20 ‘4’xf4 Wxhl
etc. Instead the game went 17 gxf3?
Wc6+ and Black won.
Panov and Estrin point out the al¬
ternative 15 Wh4+ and claim that
Black has a strong attack after 15... g5
16 ^xg5 ^xc5+ 17 ‘4’xd4 Wf6-(- 18
■^xc5 We7+ 19 <^d4 'te5-(- 20
i.f5+ 21 ^e4+ ‘4’c8 22 ‘4’c2 Wxe4+.
However, White can weather the 11 d3
storm with the seemingly highly risky After 11 d4 Black has no choice but
23 ^b3 '»c2-t 24 ■^a3 Se6 25 '»xf4 to force a draw with ll...Wxf3-l- 12
Z^6+ 26 ‘^b4 '»a4-(- 27 ‘^c3 Wa5+ 28 ‘4’el ^3+ 13 ‘4'e2 #13+ etc. The
<4'd4 'td8+ 29 '^e3 Se6-(- 30 '^f2 and game move, by not attacking the
wins. At move 19 Black should there¬ knight on e5, gives Black the chance to
fore force a draw with 19...Wf6-(- 20 play for a win.
‘4’c5 'te7+ 21 ‘4’d4 'tf6-(- etc. 11...«xf3-t 12 ^e1 '»g3-t
9 h3 The last winning try was 12...Wxhl,
Not 9 d4? since White loses his when 13 i.xf4 ^f3+ 14 ‘^e2! ±c5 15
queen after the continuation 9...<2^xf3 c3 ^f6 16 Wa4 gives a double-edged
10 gxf3 i.xf3-t 11 ■^xf3 ^5-^ 12 <^12 game - Kuindzhi.
Wxdl. 13 ^e2 m3+ V2-V2

113
The King’s Gambit

Summary
The fashionable reply to the Bishop’s Gambit 3 ^c4 is 3...c6 4 <5303 ‘53f6 (or
3.. .‘53f6 4 <5303 c6), when White should prefer 5 ^b3! d5 6 exd5 cxd5 7 d4 (Game
35) to 5 d4?! (Games 36 and 37). Black’s best choice may be the relatively unex¬
plored 3...<5316 4 ‘53c3 .^b4!? (see the notes to Game 36) or 3...‘53c6!?, though the
latter may involve learning a large amount of the archaic Hanstein and other
theory contained in Chapter 3!
Although theory has not yet found a refutation of the Mason Gambit 3 ^c3
(Games 39-40), White immediately loses his ‘birthright’ of a slight opening ad¬
vantage. Nevertheless, this double-edged opening will continue to appeal to
those willing to take risks.
White’s other third move alternatives, 3 ^e2 and 3 Wf3 (Game 39) are not to
be recommended.

1 e4 e5 2 f4 exf4

3i.c4
3 <53c3 Wh4+ 4 <^e2 d5 5 ^xd5 (D)
5.. .M6 - Game 39
5.. .J^g4-i- - Game 40
3.. .C6
3.. .<5316 4 ‘53c3 c6 - see Games 35-37 (by transposition)
4 (D)
4.. .d5 - Game 38
5i.b3
5 d4 i.b4 6 e5 ^e4 (D)
7 “^fl - Game 36
7 Wh5 - Game 37
5.. .d5 - Game35

114
 CHAPTER SEVEN

Nimzowitsch Counter-Gambit
(2...d5 3exd5 c6)

1 e4 e5 2 f4 d5 3 exd5 c6 occurs in Games 44 and 45. Somewhat

In the Nimzowitsch Counter- surprisingly. White can play for an
Gambit Black’s strategy is similar in attack on the kingside, which works
spirit to that of the Modern Defence well in Games 43 and 44. However,
(Chapter 5): he deflects White’s e- Black’s problems in these games were
pawn with ...d7-d5 so that developing largely caused by his adoption of an
moves such as ...^f6 and ...^d6 can be inferior move order, as is explained in
made without worrying about the Game 45.
reply e4-e5. However, in contrast to
the Modern Defence, Black plays very Game 41
dynamically here. Thus the d5-pawn is Boudre-G.Fiear
eliminated with 3...c6, rather than ex¬ Pau 1988
changed for the f4-pawn as occurs in
the Modern Defence. 1 e4 e5 2 f4 d5 3 exd5 c6 4 ^c3
In Games 41 and 42 Black sacrifices An interesting alternative idea is 4
his e-pawn with 4 ^c3 cxd5, hoping We2, which wins the e5-pawn but
to regain it later with a freer game. leaves White with a congested posi¬
However, this line has now been su¬ tion. A possible continuation is
perseded by 4...exf4 (Games 43-45), 4.. .cxd5 5 fxe5 ^c6 6 ^f3 (6 c3 d4! 7
when the f4-pawn gives Black a space ^f3 ^ge7 8 d3 ^g6 9 ®e4 ±c5 10
advantage on the kingside and controls ^bd2 0-0 11 ^b3 f5! was good for
e3, an important centre square. How¬ Black in the old game Alekhine-
ever, the drawback to all this is that Johner, Carlsbad 1911) G...M.c5 (or
White has a much more healthy pawn 6.. .±g4 7 Wf2 i.xf3 8 gxf3 ±e7! 9 ®e2
structure for the endgame. If Black ..i.h4+ 10 <4’dl ^h6, which Bangiev
fails to generate counterplay, he will gives as imclear, though I would rather
suffer in the later stages of the game, as play Black here) 7 c3 d4! (an important

115
The King’s Gambit

move, preventing White’s consolidat¬ This quiet move is the prelude to a

ing 8 d4) 8 d3 ^ge7 9 ^bd2 0-0 10 surprisingly sharp battle. In Gallagher-
^b3 l.b6 11 Ag5 Se8 12 0-0-0 a5 with Sinkovics, Loosdorf 1993, White pre¬
unclear play in Penttinen-Sakovich, ferred 7 ..i.d3 and obtained an advan¬
Lubniewice 1994. tage after 7...‘53c6 8 ^f3 ^xe5 9 ^xe5
However, 4 dxc6 is insipid, after Wxe5 10 0-0 i.e6 11 ®e2 ^e7 12
which 4...^xc6 5 ..^bS exf4 6 ^f3 ..^d6 .i.b5+ ^c6 13 ..i.xc6+ bxc6 14 d3 ^e7
7 d4 ^e7 8 0-0 0-0 was Ree-Short, 15 Wd5 etc., though Black held
Wijk aan Zee 1986. Black has com¬ on to draw. In fact, it appears that
pleted his development smoothly and White can play more accurately. At
the white bishop could prove to be move 11, 11 b3! was a better try, plan¬
misplaced on b5. ning Mhl, perhaps combined with c2-
c3 to open lines for the bishop. Play
could go ll...i.e7 12 i.b2 ^f6 13 ®f3
0-0 (13...0-0-0!? may be best to add to
the defence of d4) 14 Sael and Black’s
position looks awkward.
7.. .?ic6 8 ?ixe5 9 i.e2 f5
This move has been criticised, but I
can’t see how Black can achieve a fully
equal game if he fails to disrupt
White’s build-up. For example, if
9.. ..1.e7 10 0-0 ^f6 then 11 “^lixeS ®xe5
12 looks slightly awkward for
4.. .cxd5?! Black, as 12...®e6 allows 13 ^xf6+
The alternative 4...exf4, which ..i.xf6 14 .i.f3 when the pressure on b7
seems the better move, is examined makes it difficult for Black to develop
below in Games 43-45. his queen’s bishop. The alternative
5 fxe5 d4 12.. .®b5 allows the sacrifice 13 ^d6+
Also possible is 5...^c6, when the .i.xd6 14 ..i.xd6 Wxb2, which looks
game Gallagher-Milovanovic, Liech¬ dangerous for Black after 15 Sbl
tenstein 1990, continued 6 d4 ®h4+ 7 'txa2 16 i.f3 etc.
g3 #xd4 8 .i.b4 9 Wxd4 ^xd4 10 10?ied2 ?ig4 11 4lc4 b5
0-0-0 1.XC3 11 bxc3 ^c6 12 Sxd5 ll.....i.b4+ is met by 12 c3! with
^ge7 13 Sd6 l.e6 14 ^f3. The two ideas of 13 Wa4+.
bishops plus the extra doubled pawn 12 h3! bxc4
must be worth something, and it is no Here 12.....i.b4+ is still dubious be¬
surprise that White eventually won. cause of 13 c3 dxc3 14 0-0! bxc4 15
6 4le4 hxg4, planning 16 Wa4+, or perhaps
Instead, 6 ..i.b5-i-!? is the subject of the immediate check 15 Wa4+ is even
Game 42. better.
6.. .«d5 7 d3 13 hxg4 fxg4 14 dxc4 I'de?!

116
 Nimzowitsch Counter-Gambit (2...d5 3 exdS c6)

Sael with a massive attack.

Therefore, we must conclude that
the whole variation seems dubious for
Black, perhaps as far back as 5...d4.
15 0-0!
Black is now clearly in trouble as
the acceptance of the piece sacrifice
with 15...gxf3 gives White a decisive
attack, e.g. 16 ..i.xf3 Sb8 17 ..i.f4! ®b6
18 i.xb8 ®xb8 19 i.c6+ i.d7 20
Wh5+ with a massacre.
15.. .1.b7 16 »e1 ±e7 17 ^g5 ^f6
A crucial moment. Black is enticed 18 i.d3 0-0
by the idea of checking the white king The immediate 18...®c6 falters after
on g3. Instead, he could have prepared 19 Sf2 g3 20 Se2 0-0 21 ^e6 Sf7 22
to attack the white king where it is ®xg3 etc.
going to be, rather than where it is 19 i.f4 Wc6 20 We6+
currently placed. The manoeuvre Black is forced into an endgame in
14.. .®a5+ 15 ±62 (if 15 ^d2 ^f6) which the d4-pawn is soon lost. Flear
15.. .®b6 was highly interesting. At puts up a tough fight but eventually
first it seems that Black has lost time has to submit to the inevitable.
with the check on a5, but the point is 20.. .»xe6 21 ?ixe6 Sf7 22
that White’s key move - castles - is SffS 23 217 24 i.e5 i.c8 25
rendered dubious, i.e. 16 0-0 d3+ 17 ?ixd4 i.c5 26 c3 a5 27 g3 i.b7 28
<4>hl dxe2 18 ®xe2+ ±e7 and it is by 2ae1 2e8 29 i.xf6 2xe1 30 2xe1
no means clear how White can con¬ 2xf6 31 .ie4 i.xe4 32 2xe4 h5 33
tinue his attack - the loss of the bishop 2e5 i.d6 34 2xh5 i.xg3 35 2g5
on e2 has removed most of the dyna¬ ±f2+ 36 4?g2 i.e3 37 2b5 i.f4 38
mism from his position. After 2f5 i.c1 39 2xf6 gxf6 40 b4 a4 41
15.. .®b6, White could try 16 ‘53e5 but 4?g3 i.d2 42 b5 i.xc3 43 ?ic6 f5 44
then 16.....i.d6!? seems a good answer, b6 1-0
e.g. 17 ^xg4? d3! 18 cxd3 (18 i.xd3 is White played very accurately in the
met the same way) 18.....i.g3+ 19 <4>fl technical phase.
±x%A and in view of the threatened
mate on f2 Black wins a piece. Game 42
However, all is not rosy for Black. McDonaid-Petr
White’s best response is 16 ^g5!, Catfordmi
when 16...^f6 17 ..i.d3! should be
good for him. If 17.....i.d6 18 We2+, 1 e4 e5 2 f4 d5 3 exd5 c6 4
preparing 19 0-0-0 etc. Or if U...±e7 cxd5?l 5 fxe5 d4
then again 18 #e2, answering Whilst preparing for this tourna¬
18.. .®xb2 with 19 0-0 followed by 20 ment game, I examined the Boudre-

117
The King’s Gambit

Flear game above. I wondered why Black has to act fast to exploit
White didn’t develop his king’s bishop White’s backward development. Of
instead of shutting it in with 7 d3, and course, if 14.....^xc2? then 15 d3 traps
so; the bishop.
6l.b5+!? 15 cxd3 .^c5

This move may be the final nail in

the coffin for the 5...d4 variation. It is essential to return the pawn to
6.. .?ic6 unblock the queenside pieces and pre¬
The alternative is 6.....i.d7, but after pare the way for castling kingside by
7 i.xd7+ ^xd7 (or 7...®xd7 8 ^e4) 8 challenging the bishop on c5. What
^e4 ®h4+ 9 ^g3 ^xe5 10 ^f3 White saves White is the enormous strength
has the advantage. of the passed pawn on f7, which
7 «d5 guards the e8-square and so prevents
The only challenging move. the completion of Black’s attacking
8 »e2 ±f5 9 ^g5\ build-up with ...She8.
White pinpoints f7 as the weakest 16...i.xd4 17 d3 h6
square in Black’s position. Now 10 Black has no good continuation and
M.c4 is a threat. therefore plays a quiet move which at
9.. .?ih6 10 0-0-0 11 i.c4»d7 12 least prevents 18 k.%5. However,
‘5ixf7! White finds an equally effective role
The only consistent move. If White for the bishop on e3.
simply develops, say with 12 0-0, then 18i.e3 i.xe3
play could continue 12.....i.c5 13 d3 During the game I was worried
She8 and Black is ready to exploit the about 18...She8!?, which is perhaps
knight on g5 with 14...f6. White Black’s best practical chance. How¬
would then find it hard to prove an ever, White has two winning replies,
advantage. With the game move both of which demonstrate the power
White wins a second pawn but falls of the pawn on f7. The brutal 19
dangerously behind in development. fxe8® is good enough after 19...Sxe8
13 e6 «c7 14 exf7 d3 20 ^xd4 ^xd4 21 ..i.xd4! (but not 21

118
 Nimzowitsch Counter-Gambit (2...d5 3 exd5 c6)

®f2 Sxe3+!) 21...Sxe2+ 22 ®f4 etc., while 7...<4>f8 8 ^e5!, intending 9

23 ^e3 etc., when the black queen is .i.xf4, is a good middlegame for
outgunned by White’s big material White.
advantage. Also sufficient is the more 7i.c4
subtle 19 f8®, which deflects the rook After the alternative 7 dxc6 ‘53bxc6
and should win after 19...Sxf8 20 White does best to transpose to Game
^xd4 ^xd4 21 l.xd4 Sxd4 22 0-0, 45 with 8 .i.c4 etc. 8 d5?! has also been
when White is a pawn up and the tried, but this seems bad: 8...^b4 9
bishop on f5 is awkwardly pinned. .^c4 ..i.f5 (also good for Black is
The game move is entirely hopeless Valvo’s suggestion 9...0-0 10 a3 b5!) 10
for Black. ..i.b3 #b6! (exploiting the weaknesses
19 Wxe3 b5 20 i.e6+ i.xe6 21 created by 8 d5 to prevent White from
«xe6+ »d7 22 «xd7+ <4?xd7 castling) 11 a3 ^a6 12 ®d4 ^c5! 13
The endgame is lost for Black, even 0-0 0-0 14 <4>hl l.d3! and Black was
though he will pick up the f7-pawn. better in Gallagher-Nunn, Bayswater
23 Scl ShfS 24 0-0 2x17 25 2c5 1987.
Avoiding the trap 25 Sxc6?
(planning a fork on e5) 25...Sxf3!
25.. .b4 26 2fc1 2f6 27 d4
Here the simple 27 Sxc6 wins after
27.. .5xc6 28 ^e5+ <4>d6 29 ^xc6, but
for some reason I didn’t want to play
the pawn endgame that results after
29.. .5c8. Nevertheless, the game move
is also decisive. The rest of the game is
rather pointless.
27.. .4^e7 28 2c7+ 4?d6 29 2xa7 4^15
30 2a6+ 4?d5 31 2c5+ 4?e4 32
2e5+ 4i^f4 33 2xf6 gxf6 34 2e6 7...cxd5?!
?ixd4 35 2xf6+ 4?e3 36 ?ixd4 2xd4 This natural move could well be a
37 h3 h5 38 a3 b3 39 2f3+ 1-0 mistake. Black should instead simply
castle, and leave it to White to resolve
Game 43 the central tension by playing d5xc6.
Gailagher-Sorin The reasoning behind this is revealed
Biel 1992 in Game 45.
8 i.xd5 0-0 9 0-0
1 e4 e5 2 f4 d5 3 exd5 c6 4 ^c3 Three years later Gallagher reached
exf4 5 ?if3 i.d6 6 d4 this position again and tried 9 .i.b3.
If 6...‘53f6 then 7 We2+ is irritating. However, this seems to be an inaccu¬
White has the (very slightly) better racy, as Black was able to dispense
endgame after 7..Me7 8 Wxe7+ ‘^xe7 with 9...‘53bc6 and play 9....i.g4! im¬
9 ^e5 “^iixdS 10 ^xd5+ cxd4 11 ±xf4 mediately (of course, after the stan-

119
The King’s Gambit

dard 9 0-0, 9...^g4? would be an¬ kingside onslaught. The insipid

swered by 10 ^xb7). The game con¬ 13.....i.xf3 14 Wxf3 is simply good for
tinued 10 0-0 ^g6! and White could White (he has the two bishops and a
no longer continue in normal style strong centre) so the only other move
with 11 ^e4 as 11...^h4 would be an is 13.....i.h5. Then White does best to
awkward pin. Gallagher therefore continue 14 ®d3, planning a build-up
chose 11 ®d3, but was soon in trouble with 15 M.d2 and 16 Sael. (Notice
after ll...^c6 12 l.d2?! (12 ^d5, that because the black bishop has been
planning 13 c3, safeguards White’s driven back to h5. Black no longer has
centre) 12.....i.xf3 13 Sxf3 #b6, when the option of ....^f5 in reply to Wd3,
Black won the d-pawn, as ^e2 allows which would have been a very awk¬
...^e5, picking up the exchange. ward pin.) If Black attempts to play
White battled on but eventually lost sharply after 14 ®d3 with 14...‘53ge5,
the endgame in Gallagher-Barkhagen, then White has 15 <53xe5 <53xe5 16 #b5
Geneva 1995. 17 <5312 (simplest) and both e5 and
Why did the maestro play 9 ..i.b3? b7 are attacked.
Did he simply get his moves the 14 ?3fg5! h6
wrong way round? The alternative 14.....i.xe4 is exam¬
9.. .?ibc6 10 i.b3 i.g4 11 i.c7 ined in Game 44.
12 c3 13 h3! 15»h5 ?3xd4
Putting the question to the bishop,
as Nimzowitsch would say. Here this
move proves effective, but in a similar
variation with the white bishop on c4
rather than b3 it is a blunder. So be
careful! (For the full story the reader is
referred to the 11th move of Game 45
below.)
An alternative idea is 13 ^f2, as
played in Hebden-Nunn, London
1987. White aims to attack the f4-
pawn as quickly as possible. The game
continued 13.....i.f5 14 ^d3 ‘53a5 15 Thus far the game has followed
^fel ^xb3 16 axb3 ®h4 17 ®f3 Sae8 Westerinen-Motwani, London 1988.
18 ..i.xf4 ^xf4 19 <53x14 M.eA. White In this earlier game Black accepted the
has won a pawn, but Black’s two piece offer with 15...hxg5, but the
bishops give him considerable coun¬ forcing sequence 16 ‘53xg5 <53h8 17
terplay. ^xf7! ^xf7 18 «xf5 m 19 'txf6 gxf6
13.. .1.f5 20 ..i.xf4 ..i.xf4 21 Sxf4 <4’g7 22 Sg4+
The bishop relinquishes the pin on <4’h6 23 Sfl left him unable to find a
one knight and attacks the other, but good defence to the white attack, de¬
this allows White to begin a dangerous spite the exchange of queens. He

120
 Nimzowitsch Counter-Gambit (2...d5 3 exd5 c6)

therefore had to enter an endgame Sorin was played in 1992, evidently

material down after 23...^g5 24 h4 just before publication of Winning
^h7 25 i.c2 ^e7 26 Sel Sf7 27 i.b3 with the King’s Gambitl
Saf8 28 i.xf7 Sxf7 29 Sge4 etc., 18.. .1.d3
which White won easily. Sorin tries a Black has no good way to continue.
more aggressive approach, but unfor¬ After 18...fxg3 (or 18...®h4 19 cxd4)
tunately for him Gallagher was well 19 cxd4 ®xd4+ 20 <4>hl <4>g8 21 ®xf5
prepared. Sf8 22 .i.e3! White defends and re¬
16i.xf7+! mains the exchange up (analysis by
An interesting moment. In his book Gall^her).
Gallagher gives 16 ^xf7 first, when 19 i.xf4! ^g8 20 i.xc7 »xc7 21
16.. .5xf7 17 M.xi7+ would transpose cxd4 i.xf1 22 »xg6 i.b5 23 ^f5
to the game. Is this just a harmless Not only is Black a pawn down but
transposition? Evidently not, as Black he also faces a menacing attack on his
can reply to 16 ^xf7 with 16...^xb3!?, king by the white queen and knight.
e.g. 17 ^xd8 “^^xal 18 ‘53xb7 .i.xe4 and 23.. .4.h8 24 ?id6 »d7 25 ^f7+ ^g8
Black has more than enough for the 26 ^xh6+ ^h8 27 4lf7+ ^g8 28
queen. Therefore, 16 ..^xf7+ first ?ih6+ 4?h8 29 ?if7+ 4?g8 30 ?ig5
seems essential. «xd4+ 31 ^h1 i.d3 32 «f7+ ^h8
16.. .axf7 17 4?xf7 33 »h5+ ^g8 34 «f7+ ^h8 35 Sel
«d8 36 »h5+ ^g8 37 »f7+ >4?h8 38
Se7 Wf8 39 Sxb7
Black has staved off the mating
threats but the loss of a second pawn
makes the endgame hopeless.
39.. .»xf7 40 ‘^^xf7+ 4?h7 41 b4 a6
42 ab6 g6 43 ?ie5 i.f5 44 ab7-h
<^g8 45 ^g4 .^bl 46 a4 ac8 47 a5
ac1+ 48 4?h2 i.d3 49 4?g3 ac2 50
?ie3 ae2 51 4?f3 ab2 52 ab6 4?g7
53 4?f4 4?h6 54 g4 i.b5 55 ?if5+
^h7 56 4ld6 ^h6 57 ^f5+ ^h7 58
18?ig3!! ab7+ 4?h8 59 ?id6 i.f1 60 4?g5 ad2
This surprising move was discov¬ 61 4?h6 1-0
ered by Gallagher and examined in his
book. Here he gets the chance to play Game 44
his analysis in a tournament game. Galiagher-Kelier
And it is good for White! I recall that San Bernardino 1992
John Nunn once remarked that after
writing a book you should try to play 1 e4 e5 2 f4 d5 3 exd5 c6 4 ?ic3
any good new ideas in the interval exf4 5 i.d6 6 d4 lhe7 7 i.c4
before it is published. The game with cxd5

727
The King’s Gambit

Again 7...0-0 is recommended here 16 «f3 ihab 17 i.d2 b6 18 ^f2

(see Game 45). ^xb3 19 axb3 «g3 20 4ld3 «xf3 21
8 i.xd5 ^bc6 9 0-0 0-0 10 i.b3 gxf3!
i.g4 11 i.c7 12 c3 ^g6 13 h3 This recapture looks unnatural, but
i.f5 14 4^fg5 i.xe4 Gall^her is keen to slow down any
This is the main alternative to black counterplay on the kingside.
14...h6 of Game 43. It eliminates the Thus he avoids 21 Sxf3, when
immediate tactical threats but 21.. .^h4! 22 Sffl g5 activates Black’s
amounts to positional submission. kingside pawn majority, as 23 g3 fxg3
After all, Black has the worse pawn 24 ..i.xg5 ^g6 is unclear. As we shall
structure and he should therefore be see, in the game White succeeds in
aiming for dynamic play with his breaking through in the centre and
pieces rather than exchanging off his queenside before Black’s kingside on-
most active minor piece for a knight. slai^ht becomes really dangerous.
15 ?ixe4 21.. .afe8 22 4?f2
It was more accurate to play 22
Sfel first, as now Black had the
chance to generate counterplay with
22.. ...1.d8!?, e.g. 23 ..i.xf4 ^xf4 24 ^xf4
M4+ 25 <4>gl l.g3!
22.. .a5?
Black misses his chance and is
gradually ground down.
23 Sfel f5 24 SxeB-i- SxeB 25 b4
axb4 26 4lxb4 i.d8 27 4ld5 i.h4+
28 ^1 .ig3 29 c4
The white pawns begin to roll and
15...»h4 they are three gainst one. As usual, in
No better is 15...Se8, e.g. 16 ®f3 a simple endgame the black clump of
^h4 17 ®d3 ^e5 18 ®b5 a6 (if Black kingside pawns proves no match for
goes into passive mode with 18...^eg6 the white majority on the queenside.
19 ^g5 Sf8 then the tactical 20 ‘53xf7! 29.. .41.4 30 Ba3 Bd8
Sxf7 21 M.xi7+ ‘^xf7 22 g3 seems to Perhaps 30...b5 was the last chance.
win for White) 19 ®d5! (forcing Black 31 4le7+ 4?f7 32 4lc6 Be8 33 d5 g5
into a bad endgame) 19...®xd5 20 34 Sd3 g4
l.xd5 ^d3 21 Sdl ^xcl 22 Saxcl At last the black pawns crash
Se7 23 Sel Sae8 24 <4>f2 <4>f8 25 ^c5. through, but it is too late. The white
White has killed off Black’s counter¬ d-pawn will carry the day.
play and he later exploited the black 35 hxg4 fxg4 36 fxg4 Ba8 37 4le5+
weaknesses on the queenside to win *f6 38 i.c3 4?g5 39 d6 f3 40 i.d2-l-
material in Gallagher-Almada, Chiasso White parries the threat of mate and
1991. the d-pawn now decides the game.

122
 Nimzowitsch Counter-Gambit (2...d5 3 exdS c6)

40...i.f4 41 d7 i.xd2 42 Sxd2 *f4

43 4?g3 44 d8» 1-0

Game 45
Gallagher-Ong Chong Ghee
Kuala Lumpur 1992

1 e4 e5 2 f4 d5 3 exd5 c6 4
exf4 5 ?if3 i.d6 6 d4 1 i.c4
0-0!

However, as Gallagher points out

this makes a vital difference in that 12
h3 (the equivalent of 13 h3 in the pre¬
vious two games), is no longer play¬
able: White loses a pawn after
12.. ...1.xf3 13 Wxf3 (or 13 Sxf3 ^^iigeS!)
13.. .^xd4! etc. Therefore White is
deprived of the plan which proved so
effective in the games above.
So how should White continue? If
Black avoids capturing with the 12 ®d3 then 12.....i.f5 is irritating.
pawn on d5 and now White has to White could instead carry out the
play d5xc6 himself, when we trans¬ ‘Hebden’ plan outlined in Game 43 at
pose to Games 43 and 44 above, but White’s 13th move. However, Black
with the white bishop on c4 rather would be a tempo up on his line after
than on b3. Which side does this slight 12 ^f2 l.f5 13 ^d3 ^a5 14 l.b3
difference favour? Generally speaking, ^xb3 15 axb3 Se8 16 ^fel ®h4 etc.,
the bishop is safer on b3 than on c4, which must be important in such a
though less flexible (lacking the option sharp position.
of retreating to the kingside). How¬ It therefore seems reasonable to
ever, there is an important tactical nu¬ conclude that 7...0-0 is more accurate
ance which has a considerable impact than 7...cxd5.
on the assessment of the variation, as One other alternative should be
we shall see. considered here. In probably the most
8 0-0 i.g4 9 dxc6 ?ibxc6 10 ?ie4 well-known game in the Nimzowitsch
i.c7 11 c3 Sc8 Counter-Gambit, Nunn tried
Here ll...^g6! is almost identical to 11.. .‘53d5?! against Illescas at the Dubai
Games 43 and 44, except that we are Olympiad 1986. It seems very logical
one move earlier and the bishop is on to centralise the knight and blockade
c4 not b3. the d-pawn, but in fact the knight

123
The King's Gambit

proves vulnerable on this square. The Sxg4 i.xcl 19 Wxcl f5 20 Wg5! etc.
game continued 12 4lc5! Sb8 13 Wei! In the present game Black makes a
(White hurries to force the exchange radical attempt to exploit the exposed
of queens as a preliminary to exploit¬ position of the bishop on c4. How¬
ing the weaknesses in Black’s pawn ever, ll...‘Sig6! remains the most chal¬
structure) lenging move.

13...Se8 (for 13...g5 see below) 14 12i.b3 13 h3 i.h5

Wh4 Wxh4 15 ^xh4 ^e3 16 i.xe3 Here 13....^f5 is similar to Game 43
Sxe3 17 Sael Sxel 18 Sxel and except that Black has gained the extra
White had a clear edge, as his queen- move ...Sc8 since he avoided 7...cxd5
side pawns far outweigh Black’s and White played .^b3 voluntarily.
stunted majority on the kingside. Illes- However, this difference doesn’t seem
cas has suggested 13...g5 as an im¬ to have any significant effect upon the
provement. This is certainly more in combinative line 13....^f5 14 4^fg5 h6
the spirit of the opening, as Black 15 Wh5! etc. It must be better for
seeks to prove that his kingside pawns Black to have his rook on c8 rather
have dynamic potential. However, than a8, but White still has a strong
Black seems to be busted after 14 attack.
.^xd5 Wxd5 15 <Sie4 (threatening g5
and a fork on f6, so Black’s reply is
forced) 15....^d8 16 ‘Sifxg5! .kxg5.
Mikhalchishin now suggested that 17
.^xf4 was unclear in New in Chess, but
the Fritz program took only seconds
to discover the killer move 17 Sxf4!!,
when Black is defenceless, e.g. 17...f5
18 ^xg5; 17...i.xf4 18 ^f6-i-; 17...M5
18 Wg3 f6 19 ^xf6-t Sxf6 20 Sxf6;
17...h5 18 Sxg4 hxg4 19 .^xg5 with a
decisive attack; or finally 17...‘^h8 18

124
 Nimzowitsch Counter-Gambit (2...d5 3 exdS c6)

14 Wd3 i.b6?! 21.. .Wd8 22 a4 Sxel 23 Sxel Se8

This relocation of the bishop weak¬ 24 b5 Sxel-H 25 .^xel axb5 26 axb5
ens the f4-pawn. ^cel 27 c4 Wdl 28 i.b4 i.d6 29
15*h1 i.xd6 Wxd6 30 c5 Wd8 31 i.b3 Wc8
Black has no way to undermine the 32 d5!
white centre and his bishop is poorly A good example of tactics justifying
placed on h5. Gallagher now plans to strategy. Black cannot capture on c5,
increase his advantage by exploiting since the d-pawn runs through to
his 4-2 pawn advantage on the queen- queen.
side. Meanwhile, the black majority 32.. .4^f8 33 d6 34 Wd5 #d7 35
on the kingside is inert. ^e5 36 *h2 Wxh5 37 #xf7-l-
15...a6 16 4^eg5 i.xf3 17 ^xf3 Wf6 1-0
18 i.d2 Sce8 19 Sael h6 20 i.c2 A model endgame for White in this
Jlc7 21 b4! variation, which should be compared
The beginning of a rapid advance with Gallagher’s similar effort against
on the queenside. Keller (Game 44).

125
The King’s Gambit

Summary
After 1 e4 e5 2 f4 d5 3 exdS c6 4 <Sic3 cxd5?! 5 fxe5 d4 White appears to have
good chances with both 6 ^e4 and 6 ^b5+. However, the variations are tricky,
so the reader is urged to carefully examine the analysis in Games 41 and 42. In¬
stead, in the main line 4...exf4 5 ^f3 .^d6 6 d4 ^e7 7 .^c4 Black should play
7.. .0-0! (Game 45) rather than 7...cxd5 8 i.xd5 0-0 (Game 43 and 44). The white
bishop is then on c4 rather than b3 in the critical variations, which is clearly to
Black’s advant^e.

1 e4 e5 2 f4 d5 3 exd5 c6 4 4lc3

4.. .exf4
4.. .cxd5 5 fxe5 d4 (D)
6 <Sle4 - Game 41
6 .^b5-l- - Game 42
5 4lf3 i.d6 6 d4 4le7 7 i.c4 (D) 0-0
7.. .cxd5 8 i.xd5 ^bc6 9 0-0 0-0 10 i.b3 i.g4 11 ^e4 kc7 12 c3 ^g6
13 h3 i.f5 14 ^fg5 (D)
14.. .h6 - Game 43
14.. ..^xe4 - Game 44
8 0-0 - Game 45

mMmmm \Em w
m4r&i Si*M*

M M M W,
H iti B H

m mmmmn m ""mn
5...d4 7 kc4 14 ^fg5
 CHAPTER EIGHT \ #111'

Falkbeer Counter-Gambit 5 WS M
(2...d5 3exd5 e4)

1 e4 eS 2 f4 d5 3 exdS e4 (if 6...Wxg2 7 i.e4 Wg4 8 Wxg4 i.xg4

In the Falkbeer Black sacrifices a 9 .^xb7 wins for White. A safer-
pawn to seize space in the centre and looking alternative is 6...Wd8, but
deprive White of the important devel¬ White still builds up a dangerous ini¬
oping move <2^13. However, the e4- tiative with 7 <2^13 ‘^f6 8 We2-i-! - this
pawn, the keystone of Black’s strat¬ is better than 8 0-0 .^c5-i- - 8.. ^e7 9
egy, can be eliminated with 4 d3! And .^e3 0-0 10 0-0-0, threatening to take
although Black then achieves free de¬ on h7) 7 ^ge2 ^h6 8 f5! -2ixf5 9 0-0
velopment for his pieces, the modem ^e3 10 i.xe3 Wxe3-i- 11 ^hl i.d6 12
verdict is that White has good winning <2if4 0-0 13 Wh5 g6 14 <2ixg6! gave
chances. Hence the Falkbeer has be¬ White a winning attack in Murey-
come something of a museum piece at Nikitin, USSR 1970. Also good is 5
the highest levels of chess and we can Wxd3, holding on to the extra pawn.
only give two illustrative games in this Black’s other possibility is 4...'ifxd5
chapter. Nevertheless, perhaps it is 5 We2 <2if6 and now:
time for a rehabilitation of this
counter-gambit, since Onischuk’s play
in Game 46 challenges the theoretical
assessment of the main line.

Game 46
Jonkman-Onischuk
Hamburg 1992

1 e4 e5 2 f4 d5 3 exd5 e4 4 d3! 2^f6

This is certainly better than 4...exd3
5 .^xd3 Wxd5 6 ‘2ic3, when 6...We6-i-

127
The King's Gambit

a) Gallagher’s preference is for 6 The alternative 6 .^e3 is examined

<Sid2!? However, in the variation in Game 47.
6.. .1.g4 7 ^gf3 i.xf3 8 gxf3 e3 9 ^e4 6.. .1.c5 7 #e2 i.f5
±e7 10 i.xe3 0-0!? 11 i.g2 ^c6 12 0-0 Black’s two other sharp ideas have
(as recommended by Keres) it is not been refuted;
clear how much the extra pawn is a) 7...i.f2+? 8 ^dl Wxd5+ 9 ^fd2!
worth after say H.-.'SihS 13 Wd2 f5!? f5 10 ‘Sic3 Wd4 11 ‘Sicxe4 fxe4 12 c3
14 ^g3 <2^16. The white bishop on g2 We3 13 Wh5+ ^f8 14 i.c4 Wxf4 15
looks very miserable. Wd5, when the double threat of 16
b) Perhaps 6 ‘Sic3 is better. Play Wd8 mate and 16 <Sixe4 was decisive in
could go 6....^b4 (forced) 7 .^d2 .^xc3 Reti-Breyer, 1920.
8 .^xc3 .^g4 (again there is little b) 7...0-0? 8 Wxe4 Se8 9 ^e5 f6 10
choice, as White planned 9 .^xf6) 9 .^d3 g6 11 Wc4! leaves White a pawn
dxe4 Wxe4 10 Wxe4-i- <Sixe4 11 .^xg7 up after ll....^d6 (or else 12 d6-i- will
Sg8 12 i.d3! ^c5. If now 13 ±c3 then be strong) 12 0-0 fxe5 13 ‘Sic3 etc.
13.. .‘Sixd3-i- 14 cxd3 ^c6, intending Other moves can be met by normal
18.. .0.0-0, is unclear or perhaps better development, e.g. 7..Me? 8 .ke3 <Sid7
for Black. So White should play 13 9 <Sibd2 etc.
.^f6! to stop Black castling. Then after 8 4^c3!
13.. .‘Sixd3-i- 14 cxd3 15 h3! .^f5 16 Simple development frustrates
g4 (returning the extra pawn to speed Black’s plans. The greedy 8 g4? al¬
his development) 16...Axd3 17 0-0-0 lowed Black a devastating attack after
White will have a virtually decisive 8.. .0.0 9 gxf5 Se8 in Spielmann-
initiative against the black king, which Tarrasch, Ostrau 1923.
is trapped in the centre. For example, 8.. .#e7 9i.e3 2^xc3
17.. .5.6 18 g5; or 17...i.e4 18 Sel; or Here 9...i.b4 10 i.d4 0-0 11 0-0-0
finally 17...^b4 18 a3 i.e4 19 Sh2 favours White. However, a key posi¬
<Sid3-i- 20 ‘^d2 and Black finds himself tion is reached after 9....^xe3 10 Wxe3
in a tangle. <Sixc3 11 Wxe7-i- <^xe7 12 bxc3.
5 dxe4 4^xe4 6

White’s slight lead in development

128
 Falkbeer Counter-Gambit (2...d5 3 exd5 e4)

and space advantage set his opponent A highly interesting moment. The
problems. Black can capture the pawn famous game Bronstein-Tal, Riga
on c2 or go after the d5-pawn: 1968, went 12 i.a3 ^d7? 13 0-0-0 i.e4
a) 12...i.e4?! 13 ^g5! i.xd5 14 0-0-0 14 ‘Sig5 (Keres believes that 14 Sel f5
(the attack on the bishop is very awk¬ 15 ‘Sig5 may be even stronger)
ward to meet) 14....^xa2 (Gallagher 14.. ..^xd5 15 g3!! and Black was wiped
refutes 14....^e6 with 15 ‘Sixe6 fxe6 16 out by some Bronstein magic. Keres
i.c4 Sf8 17 Shel Sf6 18 f5!) 15 c4 b5 recommends 12...<Sixd5 13 0-0-0 .^e6!
16 cxb5 a6 17 .^d3 axb5 18 Shel-i- as the best defence. Black does seem to
.^e6 19 f5 ‘^f6 20 fxe6 ‘^xg5 21 exf7 have enough defensive resources here,
and the passed pawn won the day in e.g. 14 ^g5 ^d7 15 Sel 0-0-0 16 ^xe6
Foune-Mahieu, Correspondence 1985. Sde8! (keeping the extra pawn) 17
b) 12....^xc2 13 ‘^d2 and now: .^c4 fxe6 18 Shfl <Si7f6. White has the
bl) If 13...i.g6? 14 Sel-i- ^d6 15 two bishops and pressure, but a pawn
^d4 ^xd5? (but 15...h5 16 f5 i.h7 is a lot of consolation. A similar possi¬
leaves his bishop shut out of the game) bility is 14 .^b5-i- c6 15 Shel 16
16 f5 i.h5 17 g4! i.xg4 18 ±^2+ and <Sig5 0-0-0 17 ^xe6 fxe6 and, since 18
19 .^xb7 wins (Gallagher). Sxe6? loses to 18...‘Sic7, again Black
b) However, Black has a superior holds on to his e-pawn.
defence in 13....^a4!, e.g. 14 .^d3 Sd8 White’s 12th move in the main
or 14 Sbl Sd8! followed by ...‘^f8 game is also supposed to be strong, but
without shutting in the rook on h8. Onischuk shows that here too Black
White can try 14 Sel-i-, but 14...‘^d6 is has adequate chances.
none too clear, e.g. 15 ‘Sie5 ‘^xd5!? 16 12.. .4.xd5 13 0-0-0 i.e6! 14 i.c4 c6
<Sixf7 Se8. Black is therefore probably 15i.xd5?!
defending satisfactorily in this varia¬ White gives up his bishop to force a
tion. However, he has to grovel and passed pawn. At first glance, this
has very few winning chances. seems an excellent idea, but the end¬
10 i.xc5 ^xe2 11 i.xe7 4^xf4 12 game that results is by no means
worse for Black. The alternative was
15 Shell?, with similar play to varia¬
tions after 12 .^a3 examined in the
previous note.
15.. .cxd5 16 c4 4^a6! 17 cxd5 ScS-i-
18 *b1 i-fB-i- 19 *a1 f6 20 i.f4
i.g4! 21 ad2 i.xf3 22 gxf3 *d7
The dust has settled and Black has
the better endgame: the white passed
pawn is vulnerable and well blockaded
by the black king. Meanwhile, the
white king is a long way from the cen¬
tre, which is usually a bad sign in the

129
The King’s Gambit

endgame. The black rook on c8 is well bishop on e5 and block White’s king-
placed and has a jumping off point on side attack.
c4 from which to attack White’s king- 32.. .4^f2! 33 ad2 4^g4 34 h5
side laterally. White’s only trump is This loses, but 34 .^d4 Se4 would
his better minor piece. He should at¬ be dreadful.
tempt to activate his rooks and accen¬ 34.. .g5! 35 Sfl ^xe5 36 fxe5 Sxe5
tuate the superiority of his bishop 37 Sgl h6 38 Sd4 ^d6 39 Sfl Sc5
over the knight by striving to open 40 b4 Scxd5 41 Sxd5+ Sxd5 0-1
lines on the kingside, so an aggressive
pawn action on the kingside with 23 Game 47
h4 and 24 h5 was required. Instead Wells-Lengyel
White plays only with his pieces, and Budapest 1993
soon drifts into a lost position.
1 e4 e5 2 f4 d5 3 exd5 e4 4 d3! 4^f6
5 dxe4 4^xe4 6 i.e3!? #h4+
Black cannot resist the check. An
important alternative was 6....^d6,
when play usually goes 7 ‘Sif3 0-0 8
i.d3 Se8 9 0-0 ^f6.

23 i.e3? b6 24 Sgl g6 25 i.d4 ShfS

26 *b1 ihcl 27 b3 f5 28 f4
A horrible move which gives away
the e4-square to the black knight. The
best chance was still 28 h4, planning
29 h5 to break things up and create
counterplay. Now Glaskov and Estrin claim that
28.. .4^b5 29 i.e5 ^c3+ 30 *b2 10 ‘SieS!? is good for White, giving the
The rook endgame after 30 .^xc3 continuation 10...‘Sibd7 11 <Sic4 <2^18
Sxc3 is lost in view of the weaknesses 12 ‘^hl. However, we should look
on d5, f4 and h2. If White’s rooks more closely at this. The really critical
were active he would have some sur¬ variation is 10....^xe5 11 fxe5 ^xd5
vival chances, but this is not the case (but not ll...Sxe5? 12 .^d4 Sxd5 13
here. .^xf6 gxf6 14 ^c3 with good attacking
30.. .4^64 31 Sd3 Sfe8 32 h4 chances) and now:
This is too late as Black can carry a) 12 i.xh7-i- '^xh7 13 #h5-i- <^g8 14
out a manoeuvre to undermine the Wxf7-i- ‘^h7 15 Wh5-i-, when White has

130
 Falkbeer Counter-Gambit (2...d5 3 exd5 e4)

perpetual check but nothing more.

b) 12 Wh5 g6 13 Wh6!? (13 Wf3
<Sixe3! 14 Wxe3 - the check on f7 is
nothing - 14...‘Sic6 is better for Black)
13.. .^xe3 14 Sxf7! ^xf7 15 Wxh7+
^e6 16 Wxg6+ ^xe5 17 looks
very dangerous for the exposed black
king.
Black could also try the immediate
8.. .‘Sif6 rather than 8...Se8. Then the
f7-pawn is nicely defended by the
rook on f8. However, White seems to
keep the advantage after 9 0-0 ‘Sig4 10 9 hxg3 #xe3+ 10 #e2 i.c5
Wd2! ^xe3 11 Wxe3 Se8 12 Wd4 ^d7 The immediate exchange of queens
13 “^hl Wf6 14 g3! Nevertheless, the 10.. .Wxe2-i- 11 .^xe2 proved good for
reader interested in playing the Falk¬ White in Spassky-Matanovic, Belgrade
beer should investigate this further. 1964.
7 g3 4^xg3 8 ^f3! 11 4^c3i.f5 12^h4 i.g4
In Tal-Trifunovic, Havana 1963, Black cannot win the c-pawn, as
White sacrificed the exchange with 8 12.. .Wxe2-i- 13 i.xe2 i.xc2?14 Id
hxg3?! However, after 8...Wxhl 9 We2 .^g6 15 f5 wins a piece.
i.b4-i- 10 c3 i.d6 11 i.g2 Wh6 12 13 lfxe3-i- i.xe3 14 i.e2 i.xe2 15
.^d4-i- ‘^d8 13 <2^13 .^g4 Black had a *xe2 i.c5 16 *f3 4^a6?
clear advantage. It is White who will This is a serious mistake, after
find himself under attack after which Black’s rook on h8 never plays
14.. .5.8. an active role in the game. It was
8.. .#e7 much better to play 16...0-0 17 ‘Sie4
If 8...Wh5? the exchange sacrifice is <Sid7, when White’s space advantage
much stronger due to the extra tempo: gives him a slight edge.
9 hxg3 Wxhl 10 We2 i.g4 (10..i.b4-i-? 17Sae1+<*f8 18 ^e4
now loses a piece to 11 .^d2-i- ^e7 12 White’s extra centre pawn, more ac¬
.^b4 because the knight is on f3. In the tive king and lead in development add
Tal game with the knight still on gl, up to a big positional advantage.
11 .^d2-i- ‘^d8! 12 .^xb4? Se8 would 18.. .ad8 19 c4 h5
win White’s queen - there is no block Hoping to get the rook into play
with ^e5) 11 ^bd2 12 i.d4+ via h6, but Black soon changes his
<^d8 13 0-0-0 and White was better in mind.
Socagin-Alatortsev, USSR 1971. Black 20 ^g2 i.b4 21 Se2 ^g8 22 a3 ±f8
will find it very difficult to bring his 23 b4
queen’s rook into the game, so White Black now has an idle bishop, a
can gradually prepare his assault on rook shut in the corner, and a knight
the black king. stranded on the edge of the board.

■ ~ 13?
The King’s Gambit

Meanwhile, White has a very strong Even so, there was no reason to give
pawn majority on the queenside, up hope. The one good thing about
which threatens to advance power¬ Black’s position was the solidity of his
fully. Black’s own majority on the pawn structure. Perhaps 23...c5!? 24 b5
kingside is inert. ^c7 should have been tried. Even
though White would then have a pro¬
tected passed pawn, at least the knight
re-enters the game. The ghastly game
move lets the knight into g6, when the
fight is soon over.
24 f5 ^7 25 She 26 i.d6
27 4^xd6-i- cxd6 28 Shel Sd7 29
Sxg6
White was planning 30 Se7-i- with a
mate to follow on f8 or a massacre on
the queenside.
30 fxg6-l- *xg6 31 ae7 ac7 32 Sxc7
^xc7 33 ae7 1-0

132
 Falkbeer Counter-Gambit (2...d5 3 exdS e4)

Summary
After 4 cl3 ^f6 5 dxe4 <Sixe4 the theoretical verdict on 6 is disputed by the
analysis in Game 46. However, the alternative 6 .^e3 in Game 47 still looks
promising for White. Both 6....^d6 7 ^f3 0-0 8 .^d3 Se8 9 0-0 ^f6 10 ^e5! and
6...Wh4-i- 7 g3 <Sixg3 8 <Sif3 We7 9 hxg3 Wxg3-i- 10 We2 favour White. In the first
line Black is facing a dangerous attack on his king; in the second he has to endure
a worse endgame.

1 e4 e5 2 f4 d5 3 exd5 e4 4 d3 5 dxe4 4^xe4 (D)

6 «if3 (D)
6 Ae3 - Game 47
6....^c5 (D) - Game 46

5...^xe4 6^f3 6...i.c5

133
 CHAPTER NINE
M I
Classical Variation (2...±c5)
±m^ 'Bill

1 e4 e5 2 f4 i.c5 making a retreat for his bishop with

In the Classical variation Black rea¬ ...a7-a6, but White finds a way to in¬
sons that White has weakened the di¬ crease the pressure by foregoing cas¬
agonal a7-gl with his rash second tling, while in Game 53 White carries
move, and so immediately places his out the ^a4 idea in taaical fashion.
bishop on c5, preventing White from
castling. White’s subsequent play usu¬ Game 48
ally therefore centres on his attempts Zoister-Costa
to drive away or exchange this annoy¬ Suhr 1992
ing bishop. After 3 ^f3 d6, White has
a choice of strategies. The first is to 1 e4 e5 2 f4 i.c5 3 ^f3
play 4 c3, aiming to snuff out the In Reinderman-Volzhin, Oakham
bishop with either the advance d2-d4, 1992, White uncorked 3 Ifh5!?, a kin¬
or as is more likely under the tactical dergarten move which actually wins
circumstances, b2-b4, so that after the e-pawn. The game continued in
...J.b6, the bishop can be eliminated surreal style with 3...^f6!? 4 Wxe5+
with ^a3, ^c4 and ^xb6. This strat¬ J.e7 5 Wc3 (normal moves leave Black
egy is seen at its best in Game 51. with good play after 5...^c6 or 5...0-0)
However, Black doesn’t have to give 5.. .^xe4 6 'txg7 i.f6 7 Ifhe. Five of
up his bishop in such a meek fashion, White’s first seven moves have been
and can play 4...f5!.^ with sharp play with his queen! Nevertheless, White is
(Games 48 and 49). Alternatively, he allowed a few eccentricities in the
can counterattack against e4 with opening. Although Black achieves an
4...^f6, as in Game 50. White’s second impressive build-up, Steinitz would
possible strategy is similar in spirit: he say that a pawn is worth a little trou¬
plays 4 ^c3, aiming for ^a4! to get rid ble. The finish of the game raises just
of the bishop in a different manner. In as many questions as the opening:
Game 52 Black counters this idea by 7.. .'te7 8 i.e2 d5 9 ^c3 ^xc3 10 bxc3

134
 Classical Variation (2...Lc5)

i.f5 11 ^13 ^c6 12 i.a3! Ifee 13 0-0-0 centre pawns for counterplay.
(why does Grandmaster Schussler rec¬ Nevertheless, Gallagher recom¬
ommend 13 0-0 here, allowing mends the more modest 8 ^xd4! as
13.. .J.d4-)-, winning White’s queen?) the way to maintain White’s initiative.
13.. .0.0-0 14 Shel d4 (why does Black After the reply 8...^f6 he analyses
give away a piece?) 15 J.d3 'iifxa2 16 several variations in his book, for ex¬
J.xf5-(- ‘i'bS 17 Ifxf6?? (this simply ample 9 J.g5 J.xd4 10 cxd4 ^c6 11
allows mate in one) 17... «al mate. ^c3! and White has dangerous attack¬
3.. .d6 ing chances. Black can also try the
Black can also try the aggressive immediate 8...Wh4-)- to disrupt
3.. .d5. However, Gallagher practically White’s smooth build-up. Here are
refutes this idea with his analysis: 4 some sample variations after 9 g3 Wh3
^xe5 ^f6 (not 4...dxe4 5 Ifh5 6 10 Wb3! (not 10 We2? ^f6 11 i.g5
J.c4) 5 d4 J.b6 6 exd5 IfxdS 7 J.e3 i.g4 12 ma ^bd7 etc.) 10...^f6:
^c6 8 ^c3 J.a5 (hoping to embarrass
White with 9...^e4, but...) 9 J.e2! and,
since 9...^e4 fails after 10 0-0 J.xc3 11
J.c4!, Black has no real compensation
for his pawn.
4c3
A logical move, preparing 5 d4 to
seize space in the centre.
4.. .f5!?
This is the life or death variation of
the Classical. Black launches an im¬
mediate attack on e4. It makes posi¬
tional sense in that White’s fourth a) 11 Wh5+} ^bd7 12 ^e6 c6 13
move has deprived him of the natural 'iT53 (or 13 ^xg7+} ‘4’f8 14 ^e6-i-
response 5 ^c3, bolstering his centre. Wxe6l and Black wins) 13...J.d6 14
The alternative 4...^f6 is the subject ^xg7+ ‘4>e7 15 i.e6 (15 Sfl ^e5 16
of Game 50, while 4..; J.g4 and 4....i.b6 J.g5 ^f3-(- 17 Sxf3 exf3 is good for
are considered in Game 51. Black) 15...'tg2 16 ^f5+ (or 16 Sfl
5 fxe5 dxe5 6 d4 ^c5) 16...‘4>d8 17 Sfl ^c5 and Black
An important alternative is 6 exf5, wins.
for which see Game 49. b) 11 i.g5 af8 12 ^d2 (also possible
6.. .exd4 7 i.c4! fxe4 8 ?lg5 are overtly aggressive continuations
A very natural move which threat¬ such as 12 Ifb5-)- [or 12 J.e6 Wg2 13
ens an unstoppable fork on f7, since Sfl ^bd7! etc.] 12...^bd7 13 ^e6 c6
8.. .^h6 9 'iifh5-)- would be very bad for 14 'iT33 Wgl 15 Sfl Ifxh2 with un¬
Black. Black is therefore compelled to clear play) 12...lfg4 (stopping 13 0-0-0
sacrifice the rook on h8 and has to and attacking the bishop) 13 'il35-(-
trust in his lead in development and ^bd7 14 ^e6 c6 15 Ifb3 ^b6 with a

135
The King’s Gambit

mess. Black cannot now castle. He could

c) 11 i.e3!? ^2 12 Sfl lfxh2 13 therefore try 17 ^xc3?, aiming to at¬
^cl2 'txg3+ 14 ±12 We5 15 0-CW) and tack the black king in the centre. Alas,
White has a dangerous attack for his White gets mated first: 17..Mf2+ 18
three pawns. ^dl i.g4+ 19 ^cl (19 i.e2 ad8+ is
crushing) 19...^d3-(- 20 ±xd3 lfe3-(-!
(the key move) 21 <^c2 Ifxd3+ 22 <^b3
i.e6+ 23 <^a4 Wc4+ 24 <^a5 b6 mate.
b) 16 Sfl! cxb2!! 17 Sxf6 and now
we have:
bl) 17...<^xf6 18 'td8+ <^g7 19
'txc7+ <^f6 (not 19...<53d7? 20 lfg3!
bxallf 21 'txg6+ <^xh8 22 lfg8 mate)
20 lfd8-(- with a perpetual, as 20... J.e7?
fails to 21 Ifd4.
b2) 17...bxal't 18 lfg7+ (not 18
^xg6+ <^xf6, but 18 af7+ ^xf7 19
8.. .-af6 9 ^f7 «e7 10 -SlxhS -ac6! 'txf7+ <^d8 20 ^8+ <^e7 21 1^17+ is
11 i.g5?le5 12i.xf6 another perpetual check) 18...‘^d8 19
In Gallagher-Costa, Biel 1990, 'iifg8-)- *^^7 and since 20 .^e6-i- ‘^c6 21
White tried 12 cxd4, but his opponent WleS+ ‘^b6 22 ^xc8-i- c6 seems to lose
soon had an overwhelming attack: for White, perpetual check with 20
12.. .1.g4! 13 lfa4+ i.d7 14 lfb3 i.xd4 'iife6-i- ‘^d8 21 'iifg8-i- is apparently
15 43c3 ^d3+ 16 ^xd3 (16 ‘^d2 was White’s best line.
the only chance according to Gal¬ So it seems that a draw is the out¬
lagher) 16...exd3+ 17 <^fl 0-0-0 18 ^f7 come of 12...lfxf6. Costa’s choice in
SfS! 19 'tc4 l.b6 20 ^e4 2x17 21 the main game can therefore be seen as
^d6+ 'txd6 22 lfxf7 Wc5 23 i.h4 a winning attempt.
'tf5+ 24 ^el 'te4+ 25 ^d2 J.a5+ and 13 »h5-l- <4^8 14 »h6-l-?!
White resigned. This is too timid. White’s chances
12.. .gxf6! depend on exploiting the precarious
A controversial moment. Accord¬ situation of the black king. An ex¬
ing to Gallagher 12...lfxf6 13 Ifh5-)- g6 change of queens should therefore be
14 'iifxh7 dxc3 is winning for Black the last thing on his mind. According
after both 15 ^xc3 Wf2+ 16 “^dl to Keres, 14 ^g6-)- ^xg6 15 'iifd5 gives
i.g4+ 17 <^cl 0-0-0 and 15 Sfl ^f3+ White a good game, an opinion which
16 gxf3 cxb2, when Black will soon is supported by Gallagher. However,
have a second queen. However, White after 15...‘^g7
has a much stronger move in 15
see following diagram
IfgS-)-!, when after 15...‘^e7:
a) 16 Wh7+ “^eS White has repeated Black seems to have tremendous play,
the position, the difference being that e.g.

136
 Classical Variation (2...^c5)

J.b3) Black plays 16...d3 and the con¬

nected passed pawns win easily.
16...e3
Black will pick up the knight on h8
whenever he pleases. His centre pawns
and attacking chances against the ex¬
posed white king soon win material.
17 cxd4 i.xd4 18 ^c3 ±xc3+ 19
bxc3 -ad3-h
White loses his bishop as 20 ‘^fl
e2+l is similar to the game.
20 '4>e2 ?lf4-(- 21 '4>xe3 ?lxd5-(- 22
a) 16 lfg8+? ^h6 17 ^d2 d3 and the '4>d4 ^e7 23 5ae1 -ac6+ 24 4>d5
connected passed pawns are decisive. i.d7 25 5hf1 5xh8
b) 16 ^d2 dxc3! (not 16...^14? 17 The knight, which has been hang¬
lfxe4 ^xg2+ 18 ^e2! i.g4+ [or ing since move 10, is finally captured.
18.. .dxc3 19 'txe7+ i.xe7 20 fiagl] 19 26 5f2 5d8 27 5e3 i.e8+ 28 ^c5
‘^£2! dxc3+ 20 ‘^xg2 and Black is lost) -ae5 29 5g3-h 0-1
17 bxc3 ^14 18 ^8+ (18 Wxe4
^xg2+ 19 ^dl ^e3+ 20 ^e2 i.g4+ 21 Game 49
“^el ^c2+ wins) 18...‘^h6 and Black Day-Costa
has a winning attack after both 19 Sfl Manila 1992
^d3+ 20 <^dl ^f2+ 21 ‘4>el e3 and 19
0-0-0 e3 20 ^b3 i.a3+ 21 ^bl i.f5+. 1 e4 e5 2 f4 i.c5 3 -af3 d6 4 c3 f5 5
c) 16 cxd4 i.b6 17 <53c3 i.g4. This fxe5 dxe5 6 exf5
critical position seems very good for An attempted improvement on 6
Black, e.g. 18 Wxb7 (or 18 0-0 ad8 19 d4. Note that 6...e4? now fails to 7
Ifxb7 axd4) 18...ad8 19 ^d5 m 20 lfa4+ ^c6 8 'txe4+.
0-0? (but 20 ^xb6 axb6 21 J.b3 'iiff4! is 6...We7 7d4 exd4+8 i.e2
terrifying for White) 20...J.c8! and
wins the white queen.
From this analysis we can conclude
that Black has excellent chances after
12.. .gxf6. This implies that the whole
variation with 8 ^g5 should be
scrapped as far as White is concerned.
Instead, Gallagher’s 8 ‘53xd4! seems to
be the best try.
14.. .»g7 15»xg7-l- '4>xg7 16i.d5
This will prove to be a fatal square
for the bishop, but White must attack
the e4-pawn. Otherwise (e.g. after 16

137
The King’s Gambit

Black should try 8...^c6!?, taking h6 35 ^d5 36 ^e3 0-1

advantage of a tactical trick rather
than developing White’s knight on c3 Game 50
for him. After 9 cxd4 ^xd4! 10 ^xd4 Bronstein-Royset
Wh4+ 11 g3 lfxd4 12 i.h5+ <^f8 13 Gausdal 1994
Wxd4 J.xd4 Black seems to be at least
equal. This may refute 6 exfS. 1 e4 e5 2 f4 i.c5 3 ^f3 d6 4 c3 -SlfO
9 ‘?^xc3 10 i.g5 i.xf5 11 -SldS Black makes no attempt to stop
Here 11 'iifb3!? was perhaps more White’s d2-d4. Instead, he hopes to
accurate. undermine the white centre after this
advance.
Black has won a pawn but has al¬ 5fxe5
lowed his opponent a tremendous ini¬ A critical alternative is 5 d4 exd4 6
tiative. It was better to play ll...lfd6! cxd4 J.b4-i- (or 6... J.b6 7 ^c3 0-0 8 e5
with unclear play. dxe5 9 fxe5 43d5 10 J.g5 ^xc3 11 bxc3
12 i.xf6 gxf6 13 -ah4! i.g6 14 5c1 'te8 [ll...'td5!] 12 i.cl3 f6. According
i.b6 to various theorists Black is doing well
If 14... J.d6 then 15 0-0 is strong. here, but Gallagher shows that White
15 5f1 163x07?! is in fact virtually winning by force
A simpler way was 16 ^xcZ-f! J.xc7 after 13 0-0!, e.g. 13...fxg5?! 14 ^xg5
17 Sxc7 ad8 18 i.g4! Ife7+ 19 ^f2 J.e6 15 J.xh7-(- <^h8 16 Ifh5! and it is
with a decisive attack. all over) 7 J.d2 J.xd2-i- 8 ^bxd2 We7
16.. .1.xc7 17 ^xc7+ '4>e7 18 «a4l 9 J.d3 0-0 (9...^xe4 10 ^xe4 d5 is a
This is much better than 18 ^xa8 better try, though the game move sets
Sxa8 19 a clever trap) and now 10 Wei is pleas¬
18.. .Bac8 19 i.c4 -SlcS 20 »b4?? ant for White according to Gallagher.
A terrible blunder. White should Instead, in the game Gallagher-
win after 20 IfbS!, e.g. 20...Sxc7 (the Dzevan, Royan 1989, White fell for it
best chance is to give up the queen) 21 with 10 0-0?! ^d5!, when he had to
i.xf7 ^xf7 22 ^xg6 hxg6 23 Wc4+ bail out with the horrible looking 11
^g7 24 'td4 etc. exd5 Ife3-)- 12 ‘^hl Ifxd3, though af¬
20.. .<id8! ter 13 Scl White won in only another
White seems to have overlooked ten moves (just how does Joe do it.?).
this move, breaking the pin on the 5...dxe5 6 ?lxe5
knight and so answering 21 J.xf7 with In Zso.Polgar-G.Flear, Brussels
21.. .^d3-i-. Now Black succeeds in 1987, White played in speculative
consolidating and turns the tables. style, sacrificing a pawn rather than
21 »xc5 Bxc7 22 «d6-h »d7 23 snatching one: 6 d4 exd4 7 cxd4 ^b4-)-
«xf6-h »e7-l- 24 »xe7-h '4>xe7 25 b3 8 i.d2 We7 9 i.cl3!? 43xe4 10 i.xe4
a6 26 a4 Sf8 27 Sgl Sc5 28 ^f3 b5 Ifxe4-)- 11 ‘^f2 ^xd2 12 43bxd2 and
29 axb5 axb5 30 i.e2 Bcl-h 31 4>d2 now, according to Flear, Black should
Bxgl 32 -axgl b4 33 i-dO 34 play 12...'iifd5! with an unclear posi-

138
 Classical Variation (2...Ac5)

tion after 13 Sel+ Ae6 14 He5 Wd6, as 16 ±xn+ -ihB 17 Wh4!

15 d5?! ^d7! is good for Black. Now there is no good answer to the
threat of 18 Wf6+.
17.. .Wxh2+ 18 Wxh2 -SlxeB 19
i.xe8?!
Here 19 Ifh4 wins instantly, e.g.
19.. .^g4-i- (Black has to prevent mate
on f6 and 19...^xf7 20 Wf6+ ‘4>g8 21
Wg7 is also mate) 20 “^gl Se2 21 ^d2
and Black is a queen down for the ex¬
change.
19.. .?lg4-(- 20 '4>g3 ‘?^xh2 21 ia4
?lg4 22 i.f4 5d8 23 ^d2 Ud3+ 24
-af3 h5 25 5e1 h4+ 26 4>g2 h3+ 27
6.. .0-0 7 d4 -axe4? *g3 h2 28 i.c2 g5 29 i.xd3 1-0
The correct path was 7...Ad6 with
fairly equal chances after 8 ^f3 ^xe4 Game 51
9 J.d3 Se8 10 0-0 etc., as in Tartak- Spassky-Martinez
ower-Schlechter, St Petersburg 1909. Oviedo 1991
8 Wd3!
Instead 8 ^f3 J.d6 would transpose 1 e4 e5 2 f4 i.c5 3 4113 d6 4 c3
to the Tartakower game mentioned in i.g4
the last note. Bronstein prefers to win This move has a bad reputation,
material despite the temporary dis¬ though it may appear eminently logi¬
comfort. Ultimately, the black king cal to dissuade White from playing 5
will prove more exposed than White’s. d4 by pinning the knight.
8.. .»h4-h 9 g3 -axg3 10 «xg3 »e4-h An important alternative is
11 -if 2 4...J.b6!, hoping to cajole White into
A simple enough route to victory the premature 5 d4?!, when 5...exd4 6
was 11 Ifxhl 12 Ag2 Ifgl 13 ^f3 cxd4 J.g4 is good for Black. The natu¬
J.d6 14 ^xgl J.xg3 15 hxg3, but ral response is 5 ^a3 with the possible
Bronstein sees that he can win by di¬ follow-up 5...^f6 and now;
rect attack. see following diagram_
11.. .»xh1 12i.h6g6 13dxc5
Some care is required, since 13 'iifh4, a) The slow 6 d3 provoked the ag¬
with the seemingly decisive threat of gressive response 6...^g4!? 7 d4 f5 in
14 Wfe, fails to 13 Wh4 i.e7! 14 Wxe7 Arnason-LSokolov, Haninge 1989.
Ifxh2-)- etc. White has now gained the two bish¬
13.. .5e8 14 »f4 i.f5 15 i.c4 ^c6 ops, but Black’s well entrenched
Instead, 15...Se7 would have held knight on e4 frustrated all his attempts
on longer, but White had many deci¬ for an advantage after 8 h3 ^f6 9 fxe5
sive moves, e.g. 16 .^g5. ^xe4 10 ^c4 d5 11 ^xb6 axb6 12

139
 The King's Gambit

J.d3 ^g3! (to stop White from cas¬ tage with 5 fxe5 dxe5 6 Ifa4+! J.d7
tling kingside) 13 Sgl 0-0 14 Wc2 c5! (the only move not to drop e5) 7 Wc2
15 ^f4 ^e4 16 a3 c4. A draw was ^c6 8 b4 l.d6 9 i.e2 We7 10 ^a3 a5?!
soon agreed. (10...a6) 11 b5! ^d8 12 ^c4, as in the
game Larsen-Joyner, Birmingham
1951. Spassky’s move seeks to acquire
the two bishops and a queenside space
advantage without the need for any
eccentric manoeuvres with his queen.
He succeeds, but only after some help
from his opponent.
5.. .1.xf3 6 »xf3-ac6 7 b4!
White finds a way to gain space on
the queenside.
7.. .1.b6 8 -aa3
This is too routine. It was impera¬
b) The critical move is 6 fxe5 and tive to play 8...a6! in order to prevent
now: White’s next move, which disrupts his
bl) 6...43g4 7 d4 dxe5 8 h3 ^f6 9 centre. Then after 9 ^c4 ±a7 10 fxe5
^xe5 ^xe4 10 IfhS! with advantage b5!.5 Black would have had satisfaaory
to White (Gallagher). chances.
b2) 6...dxe5! 7 ^c4 ^xe4 8 ^xb6 9 b5 ?le7 10 fxe5 dxe5 11 ?lc4 ?lg6
axb6 9 We2. Now Gallagher gives 12-axb6
9.. .<53f6 10 'txe5+ 'te7 (or 10...i.e6 11 Here is the main drawback to the
^g5) 11 'txe7+ <^xe7 12 i.c4 i.e6 13 omission of 8...a6. White has two
^xe6 ‘^xe6 14 d4 with a superior pieces, a bishop and knight, both
game for White in view of his better clamouring for the c4-square. The
pawn structure, bishop against knight ‘second best’ square for either piece
and the vulnerable position of the would be miserable compared to c4.
black king. However, the game Hec- So which piece should White put on
tor-Giorgadze, La Coruna 1995, over¬ c4, and which piece is to be disap-
turned the assessment of this line. pointed.5 Well, Black has solved his
Black played 9...^f5! and emerged opponent’s dilemma by allowing him
with the advantage after 10 d3 ^c5! 11 to exchange his knight for the bishop
'txe5+ We7 12 lfxe7+ ^xe7 13 ±f4 and then to put his bishop on its best
Sc8 in view of the double threat of square with a clear conscience.
14.. .^xd3-)- and 14...^b3 (to which 14 12.. .axb6 13 i.c4 5a4 14 d3 h6 15
^d4 is the best defence according to 0-0
Giorgadze). The two bishops and the pressure
5h3 down the f-file give White a clear ad¬
According to established theory. vantage. Black finds that he cannot
White is supposed to gain the advan¬ castle (15...0-0 16 ^xh6! wins a pawn).

740
 Classical Variation (2...Ac5)

15.. .C5 16 5b1 »d7 17 5b2 5a3 18 (Speelman thinks that 10 would
lc2 lxc3 have been more accurate) 10...dxe5 11
This leads to complete ruin, but g4 ac5 12 g5 afd7 13 i.d2 a5 14 ah4
Black is already badly placed since he (this move, attacking f7, is the only
cannot complete his development. good answer to the threat of 14...a4)
19 Sxc3 »d4+ 20 »f2 »xc3 21 14.. .axb3 and a draw was agreed. In
i.b2 Wa5 22 Wf5! Game 4, 5...i.e6 6 i.xe6 fxe6 7 d3 exf4
The decisive move. White threatens 8 J.xf4 0-0 9 ‘53a4! gave White some
to check on c8, and 22...0-0 now loses advantage.
to 23 Wxg6. 6d3 a6
22.. .»a8 23 ±xe5 Black has opened up a retreat square
Black’s centre crumbles and his king for his bishop, so that 7 ^a4 is now
is fatally exposed. useless because of 7...J.a7. White
23.. .-axes 24 »xe5-h -i-fB 25 »d6-h therefore tries another plan.
1-0 7 5f1 0-0?
Now 25...-^gS 26 e5 is curtains. It is almost always wrong for Black
to castle early in this variation, as
Game 52 White can clamp down on the king-
Gallagher-Giertz side with f4-f5! and begin a direct at¬
Suhr 1992 tack. The way to test White’s seventh
move was 7....^g4 or 7...exf4, though
1 e4 e5 2 f4 i.c5 3 af3 d6 4 ac3 White should keep the advantage, e.g.
This is the main alternative to 4 c3. 7.. .exf4 8 i.xf4 ^a5 9 i.g5 ^xc4 10
By the way, I have changed the move dxc4 h6 11 .^h4 .^e6 12 Ifd3, as in
order of this game for the sake of clar¬ Bangiev-Malaniuk, Tallinn 1986. Black
ity, as Gallagher actually played 2 ^cd has the two bishops but the pin on the
etc. knight on f6 is unpleasant.
4...ate
It is inaccurate for Black to play
...ac6 before White has committed his
bishop to c4. Thus in Hebden-Lane,
London 1987, 4...ac6 allowed 5 J.b5!
J.d7 6 aa4 J.b6 7 axb6 axb6 8 d3,
when White had the two bishops and
a better centre.
5 i.c4 ac6
Two episodes from the 1991 Short-
Speelman match should be mentioned
here (both with Short playing White
and transposing from the Vienna). 8f5!
Game 2 went 5...c6!? 6 d3 b5 7 ^b3 The prescribed move. Already there
We7 8 We2 ^hd7 9 Sfl i.b4 10 fxe5 is no satisfaaory continuation for

141
 The King’s Gambit

Black, as the unpleasant pin 9 J.g5, 'te6+ <4>b8 17 i.h6 Se8 18 lfxe5 ^d7
intending 10 ^d5, is threatened. 19 ^5 ^b6 20 i.d5 a6 21 ^d2 ^xd5
8.. .h6 9 -ad5! 22 ^xd5 Sg8 23 g4 etc. with a clear
This loses by force, so 9...^xd5 had advantage to White in the old game
to be tried. Chigorin-Pillsbury, Hastings 1895.
10-axd4 i.xd4 This is not totally convincing, but it
If 10...^xd5 then 11 J.xd5 J.xd4 12 certainly looks dangerous for Black.
f6! breaks up Black’s kingside. b) 9...0-0! This looks good after 10
11 -axf6+»xf6 12»h5! fxe5 dxe5 11 J.g5 (for 11 <^dl see be¬
Now Black is defenceless against the low) and now Black has a choice:
threat of g2-g4-g5, which smashes the bl) ll...'td6 12 0-0-0 ^h5 13 Wh4
kingside and even traps the queen on ^f4 14 J.xf4 exf4 15 ^d5 (or 15 Shfl
f6 after ...h6xg5; J.xg5. Black there¬ b5!?) and now instead of the theoreti¬
fore sacrifices a pawn out of despera¬ cal 15...^e6. Black could play 15....b5
tion. 16 J.b3 a5 with a dangerous attack,
12.. .d5 13 i.xd5 i.c5 14 5f3 since the f4-pawn is immune because
An alternative winning idea, since of a fork on e2.
14 g4 J.e7 is not conclusive. b2) ll...^xc2+ 12 ^dl ^xal 13
14.. .1.e7 15Sg3 »b6 ^d5 i.e7 14 ^xe7+ IfxeZ 15 Sfl
The only way to hold on was ^xe4!? (15...‘^h8 16 Ifh4 is dangerous
15.. .“^hZ, but in any case Black is a for Black in view of the threat of 17
pawn down for nothing. Sxf6) 16 dxe4 and now:
16 i.xh6 »g1+ 17 ^e2 »xa1 18
Sxg7-h 1-0

Game 53
Rahman-Lodhi
Dhaka 1995

1 e4 e5 2 ‘2ic3 i.c5 3 i.c4 4 d3

5 f4 d6 6 -af3 i.g4
The most active move. Black pins
the knight and introduces ideas of
...^d4.
7-aa4! b21) Here ECO gives 16...lfd6-)-
The old move is 7 h3, which leads with a big advantage to White. This is
to a critical position after 7...J.xf3 8 by no means clear. Certainly, if he is
'iifxf3 ^d4 9 and now: given just a couple of free moves
a) 9...^xc2-(-?, taking the bait im¬ White will play ^f6, forcing ...g7-g6,
mediately, seems to be bad: 10 ‘^dl and then Ifg5 and Ifh6 to mate on g7.
^xal 11 IfxgZ <^d7 12 fxe5 dxe5 13 However, White’s own king is so
Sfl i.e7 14 IfxfZ <^c8 15 i.g5 218 16 open that Black can generate all sorts

142
 Classical Variation (2...^c5)

of tactical threats to distract White 15...‘S3xc4 because of 16 Ifh6 g6 17

from his mating scheme. Further¬ Sh3 and wins) 16 J.b3 with the better
more, Black has the defensive option game for White in Kuijf-Leventic, Mi-
of ...‘4’h8 if necessary. tropa Cup 1995.
b22) In any case, 16...lfd7-)- looks Another possibility is 7...exf4 with
more flexible, with ideas of ..Mz4+ if the plausible continuation 8 ^xc5
appropriate. Possible variations after dxc5 9 J.xf4 ^h5 10 J.e3.
16.. .1fd7+ are 17 <^e2 'td4 18 i.f6??
(18 i.cl3 «xb2+ 19 ^f3 'tc3)
18.. .1fxc4+ 19 ^f2 'td4+ 20 <^el (20
‘^f3 Ifd3-)- exchanges queens)
20.. Mxe4+ and Black wins or 17 ‘^cl!
Wc6 18 b3 'txe4 19 <^b2 'td4+ 20
‘^bl with unclear play.
Not surprisingly, your author had
no stomach for these variations in the
game McDonald-Mikhalevski, London
1992. Back at move 11 (by transposi¬
tion) I tried 11 “^dl, but Black was
able to force equality with some sharp Now according to Alekhine his
play: ll...b5! 12 J.h6 ^h5 13 WxeS game with Tenner, Cologne 1907,
bxc4 14 Wxc5 gxh6 (14...cxd3 fails to continued 10...^e5? and White won
15 cxcl3 43g3 16 We5) 15 Wxh5 ^xc2! brilliantly with 11 ‘53xe5! J.xdl 12
16 ^xc2 'txcl3+ 17 <^cl 'te3+ 18 ^c2 i.xf7+ <^e7 13 i.xc5+ <^f6 14 0-0+
and a draw was agreed. ‘^xe5 15 fif5 mate. This finish is given
7.. .1.xf3 in various books.
Black could also consider 7...J.b6, However, according to Tenner, this
which is not so insipid as it appears at was all a fabrication by Alekhine. The
first glance. After 8 ^xb6 axb6 White game actually continued 10...'iife7! 11
should play 9 c3! (less accurate is 9 0-0 i.b5 f5! 12 i.xc6+ bxc6 13 'td2! fxe4
0-0 10 h3, as after 10...J.xf3 11 Ifxf3 it 14 ^g5 0-0 15 43xe4 aae8 16 i.f2
may look like White has a pleasant Ifd6! 17 We3 18 0-0 i.f3 19 i.g3
game with the two bishops and better ^xg3 20 hxg3 J.xe4 21 Sxf8+ ‘^xf8 22
pawn structure, but Black can remove Wxc5+ ‘^g8 23 dxe4 Ifxe4 and a draw
both apparent pluses with ll...^d4 12 was agreed. Also, the game was played
Wdl b5! 13 J.b3 ^xb3 14 cxb3, as oc¬ in Cologne in 1911, not 1907. This
curred in Regan-Darby, Dublin 1991) seems to be an instance of Alekhine
when White rules out ...^d4 ideas and ‘misremembering’ his games.
keeps the advantage. For example, 8 »xf3 ^d4 9 »d1
9.. .0-0 10 0-0 exf4 11 i.xf4 ^h5 12 An important alternative here is 9
Wd2 ^xf4 13 1^x14 i.xf3 14 Sxf3 ^e5 Ifg3. Once again Black has the option
15 Sg3! ‘^h8 (Black has no time for of capturing on c2. However, this

143
The King’s Gambit

seems to be bad, e.g. 9...^xc2+?! 10 10.. .bxc4 11 fxe5 dxc5 12 exf6 Wxf6
■^dl ^xal 11 Wxg7 Sf8 12 ^xc5 dxc5 White has to tread carefully just for
13 fxe5 ^xe4 14 Sfl We7 15 Ah6! is equality. False trails are 13 dxc4?
given by Keres. White stands to win as Wh4-l-, when the e4-pawn drops, and
the natural 15...0-0-0 allows 16 Wg4-i- 13 c3?! ^c6 14 dxc4 ^e5 15 i.e3 Sd8
^b8 17 'txe4 (but not 17 i.xf8 'txe5!) 16 Wh5 0-0, when Black’s control of
with a crushing position, e.g. 17...Sfe8 the light squares and the stranded
18 Sxf7 Wxe5 19 Wxe5 Sxe5 20 .i.f4. white king gave him a strong initiative
Instead, Black could try 15...Sd8 but in Tischbierek-Mikhalevski, Bad End-
this also loses, e.g. 16 Wxf8-i- Wxf8 17 bach 1995. The most sensible idea for
i.xf8 ^xf8 18 Sxf7-i- ^e8 19 Sxc7 White is 13 i.e3, e.g. 13...0-0 14 i.f2
^£2-1- 20 ^e2 ^g4 21 i.b5-i- ^f8 22 e6 cxd3 15 cxd3, preparing 0-0, with
etc. rough equality
Much better is 9...^h5!, as given by 10.. .4’xf7 11 5ixc5
Ernst. He analyses 10 Wg4 g6 11 ^xc5
dxc5 12 0-0 b5 (not 12...^xc2 13 fxe5!)
13 .i.d5 c6 14 .£b3 ^xf4 15 .i.xf4 exf4
and now suggests the piece sacrifice 16
.i.xf7-i-, leading to equality after
16.. .'^f7 17 'txf4-i- (here 17...<4>g7 18
We5+ '^h6, playing for a win, looks
dangerous after 19 SfT) 17...'i?e6 18
W{7+ ^e5 19 'tf4-H ^e6 20 Wi7+ with
a draw. Instead of the piece sacrifice,
16 Sxf4 is worthy of investigation, e.g.
16.. .<?^xb3 17 axb3 'td4-i- 18 ^hl
'txb2 19 Safi 0-0 20 'td7! 11.. .dxc5
The sacrifice has to be accepted, as
11.. .exf4? leaves Black disastrously
placed on the f-file after 12 4^b3, e.g.
12.. .^e6 13 0-0 g5 14 g3! fxg3? (things
were bad anyway) 15 .i.xg5!
(unfortunately this sacrifice can’t be
refusec^ 15...gxh2-i- 16 '^hl ^xg5 17
^5-^ ^e7 18 'txg5 Sf8 19 ^d4! WeS
20 e5 dxe5 21 'txe5+ ^d7 22 ^£5-^
'^d6 23 Sael 1-0 Lane-S.Jackson, Brit¬
ish Championship 1989.
12 fxe5 ihdl 13 c3 ^e6
9...b5 10 Axf?-!-!? Here Glaskov suggests 13...^xe5!.'
This piece sacrifice is much more 14 ^5-^ ^g8 15 Wxe5 WhU 16 g3
promising than 10 ^xc5, when after ^f3-i- 17 '^e2 ^xe5 18 gxh4 c4 19 d4

144
 Classical Variation (2...^c5)

^cl3, but White has an extra passed Korchnoi’s suggestion in the last note.
pawn in the centre, which must give Black returns the piece to force an
him a substantial advantage. endgame. However, if this was Black’s
14 0-0+ intention it would have been better to
do it after 14...'^e8 etc., as then the
king would be in the centre. For this
reason 14...'^g8 seems to be inferior to
14.. .<4>e8.
The alternative was 16...h6, but
then 17 WbO We8 18 looks very
impressive for White, though Korch¬
noi, a renowned defender, describes it
only as ‘adequate compensation for
the piece’!
17 dxe5 Wxdl 18 Sxdl 4’f7 19 ±e3
Shd8 20 Sd5 a6 21 Scl
Bangiev suggests 21 a4!? as more
After the alternative 14...'^eS 15 d4 consequent, e.g. 21...bxa4 22 flxa4
cxd4 16 cxd4 Korchnoi recommends flab8 23 b4.
16...^xe5! 17 dxe5 Wxdl 18 Sxdl '^e7 21.. .^e7 22 Bc6 Bd7 23 ^f2 h6 24
‘and Black should hold the endgame’. ^e2 h5 25 h4 5ad8 26 5xd7+ ^xd7
Instead, 16...'te7? 17 i.e3 Sf8 18 d5 27 5xa6 5f8 28 g3 5b8 29 b4 5g8
Sxfl+ 19 Wxfl ^d8 20 e6 was ghastly 30 5a5 5b8 31 5a3 5f8 32 Sd3+
for Black in Balashov-Matanovic, ^c6 33 5a3 5g8 34 5a5 5b8 35
Skopje 1970. Ba6+ 4’d7 36 ±c5 Bd8 37 4’e3
15 d4 cxd4 16 cxd4 ^xe5!? White still has all the chances, but
This counter-sacrifice is similar to after a long struggle Black won.

145
The King’s Gambit

Summary
After 2...±c5 3 ^f3 d6 the strategical plans for both sides are complicated by
some very sharp and murky tactical variations. However, some general conclu¬
sions can be reached.
In the 4 c3 line, 4...f5 seems dubious after 5 fxe5 dxe5 6 d4 exd4 7 ±c4 fxe4 8
^xd4! etc. (see the notes to Game 48) and the alternatives 4...4^f6 (Game 50) and
4.. .±g4 (Game 51) seem poor for Black. However, 4....^b6!.> seems to be ade¬
quate, judging from the variation 5 ^a3 ^f6 6 fxe5 dxe5 7 ^c4 ^xe4 8 ^xb6
axb6 9 We2 .i.f5! (see the notes to Game 51).
In the 4 ^c3 line, 4...^f6 5 .i.c4 ^c6 6 d3 is standard play. Now 6...a6 7 Sfl is
interesting, when 7...exf4 or 7...±g4 should be played, but not 7...0-0 because of
8 f5 (see Game 52). A critical alternative is 6....i.g4 (Game 53).

1 e4 e5 2 f4 ±c5 3 d6

4c3
4 ^c3 ^f6 5 .i.c4 ^c6 6d3 (D)
6.. .a6 - Game 52
6.. .±g4 - Game 53
4.. .f5 (D)
4.. .^f6 - Game 50
4.. .±g4 - Game 51
5 fxe5 dxe5 (D) 6 d4
6 exf5 - Game 49
6.. .exd4 - Game 48

mjmrnrn i
mm mxmx
"ilii..Ji,.. *

6d3

146
 mpm TEN I
Second and Third Move
Alternatives for Black

1 e4 e5 2 f4 wants to play ...g7-g5, defending the

In this chapter we shall round off f4-pawn and transposing to favourable
our examination of the King’s Gambit Hanstein or Philidor Gambit varia¬
by looking at divergences by Black tions, without allowing the Kiese-
from the main lines at move two or ritzky 3...g5 4 h4 g4 5 ‘5)e5.
move three. The most important of
these is 2...exf4 3 ^f3 h6, the so-called
Becker Defence (Games 54 and 55).
After 2...exf4 3 ‘5if3 Black also has
3.. .^f6, 3...^e7, 3...^c6 and 3...f5
(Game 56). Other moves are 2...‘?ih6,
2.. .^f6 and 2...^c6 (Game 57), 2...'tf6
(Game 58), 2...Wh4-)- (Game 59). In
general, the sidelines given here are
favoured by players who want to
avoid having to learn, all the main line
theory. Whether or not they are good
enough for equality is a moot point, as A^c2
we shall see. After 4 d4 g5 5 ^c3, 5...d6 would be
an immediate transposition to the
Game 54 game. Alternatively, Black could try
Gallagher-Juergens 5.. ..1.g7 6 g3 fxg3, as in Gallagher-
Bad Worishofen 1994 Nunn, Islington 1990, when 7 hxg3 d6
also transposes to the game.
1 e4 e5 2 f4 exf4 3 h6 A completely different idea is 4 b3,
The Becker Defence, which is simi¬ for which see the next game.
lar in spirit to Fischer’s 3...d6. Black 4.. .d6 5d4g5

147
The King’s Gambit

Here 6 h4 ^g7! would justify since the knight on gl is temporarily

Black’s opening. He isn’t forced into paralysed, it would be strange if White
6.. .g4, but can instead solidify his stood better here. After ll...‘5ic6, a
kingside pawn structure with the aim possible continuation is 12 ^d5
of reaching the Philidor Gambit posi¬ (only now) 13 hxg5, intending
tions examined in Chapter 3. White ...^f6 etc., when Black is better, or
therefore adopts a different strategy. alternatively 12 ±e3 ±g5! 13 .IxgS
6g3! Wxg5 and White cannot play 14 Wd2
White makes his pawn sacrifice as it drops the d-pawn.
permanent. On the other hand, he 7.. .1.g7 8 i.c4
gains attacking chances along the f-file Gallagher considers that 8 ^xg5 is
and opens up the position in order to interesting in his book, but in this
exploit his lead in development. So far game he prefers not to speculate. In
Black has failed to develop a single fact, this seems very dubious for
piece. White, e.g. 8 ^xg5 hxg5 9 Sxh8 .i.xh8
6.. .fxg3 10 Whs i.xd4 11 i.xg5 i.f6! Now 12
After 6....i.g7 7 gxf4 g4 8 ^gl Wh4-i- Wh7?? is not one of Joe’s better sug¬
9 ^e2 g3 10 ^f3 i.g4 11 i.e3 White is gestions, since 12....i.xc3-i- 13 bxc3
better (Bhend). Wxg5 leaves Black two pieces up for
7hxg3 nothing. But in any case I don’t think
that White has enough for the piece,
e.g. 12 i.c4 We7 13 i.xf6 Wxf6 14
^d5 (14 0-0-0 m-i-) 14...Wxb2 15
fldl 'i’f8! and if the knight moves
from d5 White has to reckon with
...Wc3-h.
8.. .Ag4
This looks better than Gallagher’s
suggestion of 8...^f6, when he analy¬
ses 9 'td3 ^c6 10 i.e3 ±g4 11 Sfl 0-0
12 0-0-0 'td7 13 Sf2 etc., as being
good for White. In the game Black
A critical alternative is 7 h4, when profitably delays developing his king’s
Gallagher gives 7...g4 8 ^gl g2! 9 knight.
i.xg2 i.e7 10 h5 i.h4-i- 11 ^e2. Now 9 Sfl lfd7
he claims that White can reach a good The careless 9...^c6.^ would be
endgame, despite the pawn minus, heavily pimished after 10 .i.xf7-i-!, e.g.
after ll...i.g5 12 i.xg5 WxgS 13 Wd2 10.. .'^f7 11 ^xg5+ ^e8 (ll...<^g6 12
Wxd2-i- 14 ^xd2 ^e7 15 ^ge2. This Wxg4 hxg5 13 Wf5-i- mates) 12 Wxg4
seems correct, but at the beginning the hxg5 13 We6+ ^ge7 (13...'te7 14
simple ll...^c6! would be a consider¬ W§b+ ^d7 15 4^d5! wins) 14 'tf7+
able improvement for Black. Indeed, '4’d7 15 Wxg7 with a clear advantage

148
 Second and Third Move Alternatives for Black

to White. 23 ±b4 Sa8 24 ^b1 dxe4 25 <^65

10«d3 ±h5! ±xe5 26 dxe5 Sxc2!!
This overprotects f7, thereby pre¬ A spectacular move. Of course, cap¬
paring to develop the king’s knight to turing the rook either way loses to
e7. He avoids the natural to 27...e3.
take the sting out of an e4-e5 advance 27 #e3 #d3! 28 Sf6 Se2-l- 29 l'xd3
by White. exd3 30 Sdl d2-i- 31 Sxg6-i- fxg6 32
11 ±d2 a6! ^c2 Sd8 33 i.xd2 Sxe5 34 b3 Se2
This excellent move rules out 35 a4 b4 36 a5 c5 37 a6 Sd7 38 g4
11.. .^c6 12 .^b5! with the threat of 13 0-1
d5 and therefore prepares to develop
the knight. Game 55
12 0-0-0 13 5id5 Fedorov-Svidler
White has completed the mobilisa¬ European Team Ch., Pula 1997
tion of his pieces but is struggling to
find a breakthrough. 1 e4 e5 2 f4 exf4 3 ^f3 h6 4 b3
13.. .51.e7 14Sde1 This little move disrupts Black’s
This loses time in a critical situa¬ plans, as now 4...g5 can be answered
tion. 14 Sf2, preparing to double by 5 .i.b2. Black therefore has to
rooks on the f-file, looks better. If change track and seek counterplay
Black plays to win the d-pawn then with ...d7-d5. Nevertheless, although
there are obscure complications, e.g. ...h7-h6 may appear to be a wasted
14 flf2 g4 15 ^f4 .^g6 16 ^xg6 fxg6 tempo in most of the variations which
17 i.f7-i- ^d8 18 d5. follow, it should be remembered that
14.. .Age 15i.c3 0-0 it is precisely this move which has
Castling queenside allows a .i.xa6! provoked White into the ‘unnatural’
sacrifice. Black therefore castles king- fianchetto of his queen’s bishop. 4 b3
side and prepares an attack on White’s cannot therefore be claimed as the
king. refutation of 3...h6.
16#d2 4...d5
White should still consider the idea
of 16 Sf2, planning either Sefl with
play on the f-file or Sh2 and Sehl
with pressure on the h-file.
16.. .b5 17 ±b3 a5 18 5ixe7-l- Wxe7
19 i.d5 «d7 20 «f2 Sa6 21 i.xc6
Sxc6 22 Axa5 d5!
White has broken the phalanx of
advancing pawns, but now he finds
that Black has fatal pressure against c2.
Black now wrenches open the light-
squared diagonal for his bishop.

149
The King’s Gambit

Black can delay this for a move, e.g. 10«f2 'SlgA 11 «d4 ±f6
4.. .‘5if6 5 We2 d5. However, he proba¬ Here ll...i.c5!? 12 1^x14 ^f2 13
bly didn’t want to give White the op¬ .i.b5 i^ixdl 14 Sxdl is unclear. White
tion of 5 e5!? has a pawn and some attacking
5 exd5 6 ±b2 chances for the exchange. This varia¬
6 c4, defending the d5-pawn, is well tion and the comments that follow are
answered by 6...c6 7 dxc6 ‘Sixc6 fol¬ based on Fedorov’s analysis in Infer-
lowed by 8...^c5, and White’s dark mator 69.
squares are looking sick. 12lfg1 i.e7 13 g3
6.. .1.e7 The most enterprising move. White
Black could capture the d-pawn, but avoids the tacit offer of a draw with 13
after 6...‘5ixd5 7 ±c4 his king’s bishop Wd4 .i.f6 etc., and instead opens lines
is pinned down to the defence of g7, against Black’s king.
and besides he is unlikely to be able to 13.. .1.c5 14«g2
hold on to the f4-pawn in the long White has to give up the exchange,
term. as 14 d4? is positional surrender: after
7 M 0-0 8 We2 14.. .±d6 the bishop on b2 is shut in
White plans to castle queenside and and 15...^e3 is on the cards.
then start a direct attack on the black 14.. .'2112 15gxf4 2lxd1?
king. However, Fedorov himself criti¬ It was better to play 15...2^xhl,
cises this move and recommends 8 with complications after 16 2^e4 f6 17
.^c4 with unclear play. 2^xc5 2^xc5 18 i.c4.
8.. .'abd7?! I6 2lxd1 2lf6 17i.c4i.f5
The game Hebden-Pein, London
1987 (which incidentally featured the
move order 4...‘Sif6 5 We2 d5 6 exd5-)-
^e7 7 .i.b2 0-0 8 ^c3) continued
8.. .fle8! 9 0-0-0 ^xd5 10 WeS ^xc3 11
dxc3 .i.d6 12 Wh5 and now Gallagher
gives 12...^c6! 13 i.c4 Wf6 or 13 c4
We7 as good for Black.
Fedorov must surely have known
about this game and Gallagher’s analy¬
sis of it before playing Svidler. Why
did he voluntarily play the ‘bad’ 8
We2 therefore? Perhaps he has a little 18 2lf2?
trick up his sleeve and intends to en¬ And here it is White who misses his
tice some future opponent into this chance. 18 ‘2le3! was the way to con¬
line! tinue the attack. Then if 18...i.xe3 (or
9 0-0-0 Se8 18...Sxe3 19 dxe3 i.xe3-i- 20 ^bl
If 9...^b6 10 We5\ with advantage i.xf4 21 flgl) 19 dxe3 Sxe3 20 Sgl g6
to White - Fedorov. 21 Wf2 White has a decisive attack in

150
 Second and Third Move Alternatives for Black

view of the weaknesses of Black’s dark (8...Wxe4 loses the queen after 9 ±b5-i-
squares on the kingside. c6 [or 9...<^d8 10 Sel Wf5 11 Se8
18.. .51.5 19 Sgl g6 20 ^5 mate] 10 Sel) 9 h5 ^h4 10 i.xf4 ^c6
21 #f3 #h4 22 d4 i.xd4 (Black is in serious trouble as he can¬
Black forces a draw, though he not develop his kingside; he therefore
could have fought on with 22...Ad6. elects for queenside castling) 11 .i.b5
23 Axd4 ^e2+ 24 Axe2 0-0-0 12 .^xc6 bxc6 13 Wd3 and White
Not 24 Wxe2 Wxd4 25 Sel f6! and quickly built up a decisive attack.
Black wins. Hence 3...‘^e7 seemed dead and
24.. .#xd4 25 #xf5 Sxe5 26 Sxg6+ buried, but then Ivan Sokolov discov¬
This leads to perpetual check, as ered 6...^d5 (rather than 6...^g6).
26.. .'^fS 27 Wf6! would be bad for After 6...^d5 7 i.c4 (7 c4 ^e3)
Black. 7.. Ae7 8 0-0 0-0 9 ^e5 i.e6 Black had
26.. .fxg6 y2-y2 a satisfactory game and even won in
After 27 Wxg6+ the black king can¬ the game Riemersma-I.Sokolov, Am¬
not escape the checks, e.g. 26...'^fS 27 sterdam 1995. Perhaps it is time to
Wf6-i- ^e8 28 i.h5-i-! ^d7 29 i.g4-i- rehabilitate 3...^e7.
etc. b) Second, we should mention
3.. .41c6. After 4 g5 5 h4 g4 6 ^g5
Game 56 this transposes to variations consid¬
Reinderman-Huzman ered in Chapter 3 (the Hamppe-
Wijkaan lee 1993 Allgaier Gambit).
c) Finally, 3...f5 seems inferior after
1 e4 e5 2 f4 exf4 3 5if3 4 e5, e.g. 4...g5 5 d4 g4 6 .i.xf4! gxf3 7
The Schallop Defence. Black coun¬ Wxf3, as in Schlecter-Teichmann, Vi¬
terattacks against e4 and thereby gains enna 1903, when White has a very
time to defend his f4-pawn with good version of the Muzio-style sacri¬
...^h5. A sharp struggle ensues, as fices considered in Chapter 3.
White is practically forced to offer a 4 e5 5 d4
piece sacrifice in the main line.
Here we shall take the opportunity
to look at some lesser played moves;
a) First, 3...^e7 is the Bonsch-
Osmolovsky variation. This has been
under a cloud since the game Spassky-
Seirawan, Montpellier 1985, which
went 4 d4 d5 5 4lc3 dxe4 6 ^xe4
^g6?! 7 h4! We?]} (this turns out hor¬
ribly after White’s imexpected reply,
but 7...±e7 8 h5 ^h4 9 Axf4 was also
good for White in Kuznetsov-Bonsch-
Osmolovsky, USSR 1962) 8 ^f2! i.g4 The quiet 5 ±e2 contains a lot of

151
The King’s Gambit

poison, e.g. 5...g5 6 0-0 Sg8 7 d4 g4? (a 7.. .g4

blunder, though after 7...d5 8 c4 c6 9 After this a highly interesting battle
^c3 White has a good game) 8 ^c3 begins. The white knight cannot re¬
gxf3 9 Axf3 Wg5 10 ^e4 Wf5 11 ^xh5 treat from f3 without causing disarray
Wxe4 12 i.xf7-H ^xf7 13 Sxf4-H Wxf4 in White’s position. The question is,
14 Wh5-)- '^g7 15 Axf4 and White had can Black find a way to capture on f3
a winning attack in Glaskov-Shapoval, without running into a big attack?
Correspondence 1985-86. 8 0-0 Sg8
5.. .d5 White’s powerful centre would
If 5...d6 then 6 We2! looks best, outweigh the piece after 8...gxf3 9
when if White is lucky Black will play .i.xf3 ^g7 10 cxd5. Black would then
losing a piece after 7 exd6 find it impossible to co-ordinate his
and 8 Wb5-i-. Gallagher says that he pieces in the face of White’s attack.
has caught two players in this trap, 9 cxd5
including Huzman, the hero of our This is forced, as 9 ^c3? allows
illustrative game! Instead Black should 9.. .dxc4! 10 ^e4 (10 ^xc4 gxf3)
answer 6...d5, when 7 c4 should be 10.. .^c6, when ll...gxf3 is really a
good for White. threat (analysis by Huzman in Infor-
Note en passant that 5...g5 is well mator 36).
answered by 6 ^fd2! 9.. .#xd5!
6±e2 Here 9...gxf3? would still be bad af¬
Instead of this, 6 c4 is given an ex¬ ter 10 i.xf3 'tg5 11 ^c3 i.f5 12 'te2.
clamation mark by Bangiev. After 10 ^c3 #d8 11 #d3! Sg6
6.. .g5 (6...c6 is safer, though a little Huzman analyses ll...gxf3 12 .i.xf3
passive) 7 g4! White has an excellent Wg5 13 ^e4 Wg6 14 Sf2 with the
game, e.g. 7...±xg4 8 Sgl ±xf3 (else makings of a strong assault by White.
the g5-pawn drops) 9 Wxf3 ^g7 10
cxd5 or 7...^g7 (best) 8 ±b4 9
flgl etc., as in Bangiev-Podrezrov,
Correspondence 1986-87.
6.. .g5 7 c4
It is curious that Huzman allows
and Reinderman avoids 7 4lxg5, as
this worked out well in R.Byrne-
Guimard, Wettheim Mem 1951, after
7.. .Wxg5 8 i.xh5 ^4+ (8...'txg2 9
^3) 9 ^fl i.e6 10 i.f3 ^c6 11 ^c3
0-0-0 12 ^e2 .i.h6 13 g3 etc. Evidently
Huzman has foimd a way to 12 e6
strengthen Black’s play in this varia¬ A visually impressive move, but
tion (or else he was bluffing!). In any Bangiev thinks that 12 We4 is better,
case, the game move is not bad. with the possible continuation

152
 Second and Third Move Alternatives for Black

12.. .gxf3 13 i.xf3 ^g7 14 ^dS i.e7 15

±xf4 and White has a dangerous ini¬ Game 57

tiative for the piece. Gallagher-Bliumberg

12.. .Axe6 13 5le5! Eupen Open 1993
Clearly better for Black is 13 WbS-i-
^c6! 14 Wxh5 gxf3 15 ±xf3 Wxd4-i-! 1 e4 e5 2 f4 5lh6?
16 '^hl 0-0-0 (Huzman).
13.. .51c6 14 5lxc6?
The only good continuation was to
capture the rook. After 14 ^xg6! hxg6
15 i.xf4 ^xf4 16 Sxf4 ^5!?, plan¬
ning ...0-0-0, the position would have
been unclear according to Huzman.
14.. .bxc6 15 ±xf4 5lxf4 16 Sxf4
Ad6
Now Black has the initiative. His
dark-squared bishop stares menacingly
at White’s kingside.
17Se4^8! This is certainly one way to get
Black safeguards his king before go¬ your opponent out of theory! How¬
ing over to the attack. ever, White simply develops his
18 fifi ^g8 19 Adi pieces, after which the knight begins
The only chance for activity is to to look ridiculous on h6.
challenge the bishop on e6, but the A more sensible knight develop¬
price of this is a second pawn. ment is 2...‘5if6, with the plausible
19.. .5b8 20 Ab3 Axb3 21 axb3 continuation 3 fxe5 (it is curious that
#d7 22 Se2 Sxb3 23 ^h1 g3 24 this is possibly the only time in the
Wc2 Sb4 25 ^e4 Sxd4 King’s Gambit that 3 fxe5, when legal,
Now White wins the exchange but isn’t a ghastly blunder: the knight on
runs into a decisive attack. The pawn f6 prevents a killing 3...Wh4-i- in reply)
on g3 will prove a monster. 3...^xe4 4 ^f3 ‘Sig5 5 d4 ‘Sixf3-i- 6
26 5lf6-l- Sxf6 27 Sxf6 Ae5! Wxf3 WhU 7 Wf2 'txf2+ 8 ^xf2 ^c6
This introduces the idea of back- 9 c3 d6 10 exd6 Axd6 11 ‘5id2, plan¬
rank mate. ning 12 ‘5ie4, as in the game Fischer-
28 Sfl fih4 29 Se3 Wade, Vinkovci 1968. White evi¬
Huzman gives the gruesome varia¬ dently has some advantage in the end¬
tion 29 h3 axh3+ 30 gxh3 'txh3-H 31 game, but when I asked Bob Wade
^gl Ad4+32Sef2g2! himself about the game, he told me:
29.. .1^64! 30 #b3 Sxh2-i- 31 ^g1 ‘The only advantage that White had is
ShH-! 0-1 that Fischer had kept me waiting the
It is mate in two moves after 32 whole day, deciding whether or not to
■^xhl. play on the Sabbath. When the game

153
The King’s Gambit

finally began, I was in no mood for a 7.. .1.g4 8 Wd3 c6 9 g3!

hard struggle.’ Informator has yet to This is still strong.
invent a symbol for the advantage of a 9.. .d5 10 exd5 b5 11 i.b3 b4 12
disgruntled opponent! ^e2 i.f5 13 Wc4 i.e4 14 dxc6!
A more ambitious alternative for The only way, as after any defensive
Black is 3 ‘?3f3 f5!.^ seeking to move 14...Wxd5 would be okay for
seize the initiative. Black.

Gallagher-Wohl, Kuala Lumpur 14.. .1.xf3 15 c7 «c8 16 hxg5! *d7

1992, went 4 exf5 e4 5 i^ieS 6 If 16...J.xhl then 17 gxh6 and there
fxe5 We7! 7 Wh5+ “^dS 8 d4 (here 8 is no good answer to the threat of 18
J.C41.5 is interesting, e.g. 8...Wxe5 9 Wf7 mate, as the black queen is pinned
±xg8 2xg8? 10 Wxh7 Wd5 11 ^c3 down by the passed pawn.
Wf7 12 ‘?3xe4 with a clear advantage to 17 l.xf4 l.d6 18 cxb8« «xb8 19
White, but 9...g6! is annoying, e.g. 10 «e6+ *c6 20 ±84+ *b7 21 l.xd6
Wh3 2xg8 11 Wxh7 2h8 12 Wxg6 dSl? Wc8 22 i.d7 Wg8 23 We7 Wd8 24
with complications) 8...exd3 9 J.xd3 «g7
Wxe5+ 10 <^dl ^f6 11 Wf3 ±c5 12 Even stronger is 24 J.g4+! Wxe7 25
‘?3c3 d6 13 J.f4 Wd4. White’s early J.xf3+ ‘^a6 26 J.xe7 etc.
aggression has come to nought and 24.. .Wa5
now 14...J.xf5 is threatened. Quite It was better to play 24...2g8, but
possibly a strong improvement will be then 25 Wxh7 l.xhl 26 gxh6 looks
found for White somewhere in this pretty hopeless for Black.
line, but at the moment 3...f5 looks 25 i.b5+ 1-0
promising. Joe must have enjoyed that a lot!
3 ^c3 d6 4 ^f3 exf4 5 d4 g5 6 h4!
f6 7 ±04 Game 58
Or 7 hxg5 fxg5 8 g3!, which looks Spassky-David
crushing after 8...g4 (if 8...1.g4 then 9 France 1993
gxf4 gxf4 10 J.xf4 Wf6? 11 43d5 etc.) 9
“Sigl J.g7 10 J.xf4. 1 e4 e5 2 f4 Wf6

154
 Second and Third Move Alternatives for Black

pieces and pawns save one on their

starting squares! It is true that some of
White’s pawn advances look very
ugly, but the fact that these pawns are
now out of the way of his pieces
means that he can develop a dangerous
initiative.
6.. .d6 7 l.f4 dxe5 8 «xd8+ *xd8 9
0-0-0+?id7
At last Black develops a piece be¬
sides his queen.
10 ±xe5 c6
Here Black’s idea is to accept the More time has to be wasted in view
gambit without disrupting his pawn of the threat of 11 “S^bS.
structure with 2...exf4. Hence he plans 11 ?if3*e8 12 ±07
a quick raid with his queen, which This exploits the hole which has
will then retreat. The advantage of this appeared in the black queenside struc¬
line is that Black avoids any weakness; ture.
the drawback is the enormous loss of 12.. .1.e7 13?id4 g6
time. As this game proves, White can ...and now Black finds that he has to
maintain the advantage even after the compromise his kingside to rule out
exchange of queens. Nevertheless, this 14 ‘?^f5. Clearly his strategy has failed.
is a plucky idea and a good practical 14 ±04 h5
decision against someone who knows This allows 15 Shfl to be answered
everything about the main line King’s by 15...Sh7. However, Black is gradu¬
Gambit! ally falling into a bind as Spassky
3 ^c3 «xf4 4 d4 «h4+ 5 g3 «d8 6 demonstrates his manoeuvring skill.
15 e5 ?ic5 16 l.d6 ?ih6 17 b4 lhe6
18 ?if3 19 Shel ?if8 20 ?le4
±66 21 ?if6+ l.xf6 22 exf6 ^xd6
23 Sxd6 Sd8 24 Sxd8-i- '^xdO 25
l.xe6 ?ixe6 26 ?ie5 '^c7 27 ?ixg6
White wins back his pawn, but I
have the feeling that he has rather let
Black off the hook. As we know, all
rook and pawn endgames are drawn!
27.. .fxg6 28 2x66 ah7 29 f7 axf7
30 Sxg6 h4 31 Sg4 hxg3 32 hxg3
af3 33 a4?!
After 33 ‘^b2 White has good win¬
After six moves only one piece has ning chances, as he can push his pawns
been developed and Black has all his quickly.

155
 The King’s Gambit

33.. .<^b6 34 Sg5 a5! 35 b5 a) 6...^f6.>! 7 e5 d6 8 We2! dxe5 9

Here 35 bxa5+ would have main¬ dxe5 43g4 10 ^d5! Wc5.5! (10...Wd8 11
tained winning chances. Now Black 0-0-0 looks very good for White) 11
escapes with a draw. Wb5+! J.d7 12 Wxc5 J.xc5 13 ‘?3xc7+
35.. .5a3 36 bxc6 '^xc6 37 Sg4 <^05 '^d8 14 ^xa8 ±c6 15 0-0-0+ '^e7 16
38 *b2 af3 39 c3 b6 40 ^b3 fifi J.g5+ and White is winning.
41 af4 Sgl 42 g4 fib1+ 43 ’i^c2 b) 6...d5!? 7 <53xd5 (for 7 hxg3! see
Sgl 44 '^d3 Sg3+ 45 '^d2 ^d5 46 end of this note) 7...Wxe4+ 8 We2
<^c2 Sgl 47 <^b2 <^05 48 <^b3 Sb1 + Wxe2+ 9 ‘?3xe2 ‘?3a6 10 43ec3 (also 10
49 *a2 Sgl 50 *b2 *d5 51 fif5+ 43xg3 J.e6 11 43xc7+ 43xc7 12 J.xc7
<^04 52 Sf4-l- <^d5 53 <^02 <^05 54 Sc8 followed by capturing on c2 looks
'^d3 Sg3-i- 55 <^64 Sxc3 56 g5 Sg3 better for Black) 10...g2! (to rule out
57 af5-l- *b4 58 *f4 Sgl 59 ab5-(- ±xa6) 11 i.xg2 c6 12 43e3 ±b4 and
*xa4 60 axb6 <4’a3 ’>4-’/a Black is probably better. Therefore,
White does best to answer 6...d5 with
Game 59 7 hxg3, transposing to our illustrative
Gailagher-Berezovsky game.
Beme 1993 6.. .d5 7
In 1992 Gallagher wrote of this po¬
1 e4 e5 2 f4 «h4+ sition: Tm looking forward to practi¬
cal testing.’ Well, he didn’t have long
to wait.
7.. .C6
A solid move, ruling out ‘?3b5 ideas.
Of course this does nothing to attend
to Black’s large arrears in develop¬
ment.
8 «e2 ±66 9 0-0-0 «^f6 10 l.g5

This is motivated by similar ideas to

2.. .Wf6 in the previous game.
3 g3 «e7 4 ?^c3
This is the most aggressive attempt
to refute Black’s opening play. White
is prepared to sacrifice a pawn to gain
attacking chances.
4.. .exf4 5 d4! fxg3 6 hxg3
The alternative was 6 l.f4, when
two variations are possible: It is a pity that this theoretically

156
 Second and Third Move Alternatives for Black

important game is marred by an im¬ White by advancing his queenside

mediate blunder. Black doesn’t see pawns. These pawns are now weak
that he is losing material after the se¬ and scattered and the hole at c5 is an
ries of exchanges which now ensure. ideal outpost for a white knight. The
The most testing move was 10...dxe4!, remaining moves were:
e.g. 11 ‘?ilxe4 (if 11 J.xf6 Wxf6 12 d5 20...Sa7 21 '^d2 Sd8 22 Sfl f6 23
cxd5 13 ‘?3xd5 J.xd5 14 Sxd5 J.e7 ^c5 Sd6 24 Sal h5 25 Se8 ^a6 26
looks slightly better for Black) <^e4 Sdd7 27 i.xc6 Se7 28 Sd8
11.. .‘?3bd7 12 ^f3 0-0-0. Black is con¬ ^b4 29 i.a8 Sxa8 30 Sxa8 Sxe4 31
stricted but can hope to unravel his c3 «3c6 32 d5 «3e5 33 «^xe5 2x65
game, whilst retaining the extra pawn. 34 c4 ^7 35 Sxa5 Sg5 36 Sa3
11 l.xf6 ±xe2 12 ±xe7 l.xd1 13 <^67 37 <^62 '^d6 38 ^3 '^c5 39 b3
i.xf8 '^xfS 14 ^xdl dxe4 15 ^xe4 Sg4 40 Sa6 f5 41 Sc6-l- '^d4 42 d6
The bishop and knight will prove ^65 43 c5 ^e6 44 Sc8 Sd4 45 c6
stronger than the rook and pawn ad3+ 46 *e2 axd6 47 268+ *f7 48
15.. .g6 16 -SlfS ^g7 17 l.c4 b5 18 c7 ^xe8 49 c8W+ <^67 50 b4 <^6
i.b3 a5 19 a4 bxa4 20 i.xa4 51 b5 ^g5 52 Wc1+ ^f6 53 Wb2+
Black has made it much easier for 1-0

157
The King’s Gambit

Summary
Black’s chances in the variations examined in this chapter, with the possible ex¬
ception of the Becker Defence, are by no means as good as those he achieves by
entering the Kieseritzky Gambit. Why should Black be content with a solid, but
slightly inferior position? It is perhaps reasonable to suppose that the King’s
Gambit would be much more popular with White players if the variations in
this chapter were to arise more often!

1 e4 e5 2 f4 (D)

2.. .exf4
2.. .<2^h6 - Game 57
2.. .Wf6 - Game 58
2.. .Wh4+ - Game 59
3 ?if3 h6 (D)
3.. .<S^f6 - Game 56

4 b3 - Game 55
4.. .d6 (D) - Game 54
 ma Of GAMES

Belotti-Loncar, Mitropa Cup 1995.83

Boudre-Flear.G, Pm 1988.115
Bronstein-Royset, Gausdal 1994.138
Chandler.C-Howard, Correspondence 1977.24
Chigorin-Davidov, St Petersburg 1874 59
Day-Costa, Manila 1992. 137
Fedorov-Adams, European Team Ch., Pula 1997..68
Fedorov-Pinter, Pula 1997.. 15
Fedorov-Svidler, European Team Ch., Pula 1997..149
Gallagher-Berezovsky, Berne 1993.156
Gallagher-Bliumberg, Eupen Open 1993.153
Gallagher-Bryson, Hastings 1994. 33
Gallagher-Flear.G, Lenk 1992. 16
Gallagher-Giertz, Suhr 1992. 141
Gallagher-Juergens, Bad Worishofen 1994 147
Gallagher-Keller, San Bernardino 1992.121
Gallagher-Klovans, Oberwart 1993. 79
Gallagher-Kuzmin, Biel 1995. 22
Gallagher-Neussner, Loosdorf1993. 80
Gallagher-Ong Chong Ghee, Kuala Lumpur 1992.123
Gallagher-Sorin, Biel 1992.119
Gallagher-Van der Sterren, San Bernardino 1992.88
Grasso-Pampa, Correspondence 1995.39
Hector-Leko, Copenhagen 1995 19
Yitctor-XiztAmaw, Antwerp 1994 91
Henris-Goossens, Charleroi 1994 45

159
The King’s Gambit

Horvath.C-HorvathJ, Budapest 1995 .112

Ivanchuk-Piket, Linares 1997. .107

Jonkman-Hansen,L.B, Wijk aan Zee 1994 .56

Jonkman-Onischuk, Hamburg 1992 .127

Kristensen.K-Sorensen, Copenhagen 1995 .42

Leisebein-Baer, Correspondence 1996 62

Lelen-Marzec, Los Angeles 1991. 66
Matsuura-Van Riemsdijk, Brazil 1995 43
McDonald-Hector, Oviedo 1992. 82
McDonald-Petr, Catford 1992. .117

Morozevich-Kasparov, Paris (rapidplay) 1995.. . 20

Neffe-Bronstein, Wrexham 1995. 51
Amsterdam 1995.... 31
Polasek-Karolyi, Prague 1988.. 71
Rahman-Lodhi, Dhaka 1995. 142
Reinderman-Huzman, Wijk aan Zee 1993. 151
Short-Akopian, Madrid 1997. 13
Short-Nikolic.P, Wijk aan Zee 199? 99
Short-Piket, Madrid 1997. 76
Short-Shirov, Madrid 1997.. .28

Spassky-David, France 1993 154

Spassky-Furman, Tallinn 1959.. 109
Spassky-Martinez, Oviedo 1991 139
Spassky-Xie Jun, Monaco 1994. .46
Wells-Lengyel, Budapest 1993. .130

Westerinen-Komeev, Zaragoza 1995 93

Westerinen-Kuzmin.A, Afosco'ze; 1989. 103
Westerinen-Pakkanen, Helsinki 1992 106
Winants-Almasi.Z, Wijk aan Zee 1995. 34
Winants-Van der Sterren, Wijk aan Zee 1995. .36
Yoos-Hjartarson, Reykjavik 1996.. 53
Yoos-Kirton, Saskatoon 1994. 63
Zoister-Costa, Suhr 1992. 134

160