How 10 pl1n in the end91m! in <h!

s,
. . . .•
' .
• .

.L I
Endgame Secrets

Christopher Lutz

B . T . Batsford Ltd, London

First published in 1 9 9 9

© Christopher Lutz 1 9 9 9

ISBN O 7 13 4 8 1 6 5 X

British Library Cataloguing-in-Publication Data.

A catalogue record for this book is

available from the British Library.

All rights reserved. No part of this book may be

reproduced, by any means, without prior permission

of the publisher.

-Printed in Great Britain by

Polestar Wheatons Ltd, Exeter

for the publishers,

B . T . Batsford Ltd,

5 8 3 Fulham Road,

London SW6 5BY

A BATSFORD CHESS BOOK

General Manager: Nigel Davies

Advisors: Mark Dvoretsky, Raymond Keene OBE,

Daniel King, Jon Speelman, Chris Ward

 Contents

Preface 5

Symbols 8

Endgames 1 - 4 5 9

Index of Players 171

Index of Material Balances 173

Bibliography 175
 Preface

Every chess player has strong as defended in a better way? Is the po­

well as weak points. Whether it's sition really a draw? Answering

tactics or strategy, attack or defence, these and similar questions helps

opening, middlegame or endgame, you to develop a better understand­

the ambitious player should strive ing of the endgame. I spent many

for real mastery in all areas. When days on the analysis of the more

looking through my own games I complex examples, probably you

realized that often there was a lack yourself won 't want to expend that

of continuity between middlegame much time. Nevertheless I recom­

and endgame. An advantage gar­ mend pondering on the games for a

nered in the earlier stage of the short while at least. lt ' s important

game was squandered later on. An that you assess every single move

equal position tumed into a lost without prejudice. Often i t ' s the ob­

endgame. Therefore, roughly two vious moves that have to be ques­

years ago, I decided to apply myself tioned. Certainly you won 't agree

more seriously to that part of the with many variations and comments

game. of mine, sorne things you may even

Apart from studying known litera­ consider to be trivial. But in view of

ture this meant analysing a multi­ the detailed annotations you should

tude of endgames, both from my at least leam something new.

own and from other players' games. The endgames are placed roughly

This work led to the present book. I in order of material balance. We be­

present 45 endgames of varying dif­ gin with pawns, go over to minor

ficulty which are arranged as exer­ pieces, rooks and finally queens.

cises. Y ou are requested, for Within the analysis you frequently

example, to find the correct move, find transitions to other material

to formulate a plan or to compare constellations. F or studying a spe­

different altematives. Often you will cific type of position you might use

only read the instruction "Analyse the index at the end of the book.

the game", Then you are invited to The examples 1 to 3 show that

play through the game without com­ pawn endings are no trivial matter,

ments and formulate your own even in grandmaster play. Adams­

thoughts: Where are the tuming Lutz is quite remarkable. In almost

points? Could the loser have any endgame there might arise a
6 Endgame Secerets

transition to a pawn ending, there­ space advantage or the "better

fore a firm knowledge in this field is bishop" (in respect to the pawn

indispensable. structure) are importan t. But watch

Endgames 4 to 1 6 <leal with minor out! An active "bad" bishop usually

pieces. There is a focus on the fight isn't worse than a passive "good"

of "knight against bishop". bishop. Besides , a "bad" bishop can

Games 1 7 to 1 9 are rook endings. protect his own pawns.

On account of the many existing Endgames with rooks and

books on this topic I have contented opposite-colo u red bishops are char­

myself with these examples. acterized by strategical fea tu res. As

Of enormous importance for the the bishops d o n ' t interfere with each

practica! player are endgames 20 to other, their co-operation with both

3 9 . You will encounter the material their own and enemy pawns is im­

combination of rook and minor portant. The stronger side may use

piece very often in your games, its own pawns to res tri ct the oppo­

even more often than pure rook end­ nent' s bishop. For the defender it's

ing. In February 1998 I held a desi ra ble to create a stronghold for

5-day-training lesson on that topic. the bishop in suc h a way that it can

This lesson was attended by the defend both wings. C ompare the

young players Holger Ellers, Holger set-up b .tg3 + L:sh4 (Nunn-Sadler)

Grund, Florian Handke, Arnd and b .t b 3 + L:sa4 (Kramnik­

Lauber, Alexander Naumann, Mik­ Kasparov). Furthermore you should

haljo Prusikhin and Christian Wil­ watch out that in the event of an

helmi. Their playing strength at the exchange of rooks a drawish

time averaged that of an Interna­ opposite-colo u red bishop ending

tional Master. The training involved might arise.

playing out given endgame posi­ There's an old rule that rook and

tions, sorne of the results are pre­ bishop form a better team than ro ok

sented here. and knight. But this i s n ' t necessarily

As there are very few known gen­ the case if all pawns are on one side

eralizations about endings with rook or if the bishop' s side suffers from

and minor piece, I have tried to for­ weak pawns. E nd g ames 37 and 3 8

mulate sorne: demonstrate the bishop' s superiority

The combination of rook and while in en d games 36 and 39 the

knight leads to a highly tactical bat­ knight is stronger.

tle. Concrete variations take prece­ Endgames 40 and 41 show one

dence over general conclusions, side bein g the e x chan g e up. In par­

look at the highly tactical encounter ticular, the analysis of Lutz-Karpov

Lutz-Hübner. In Kramnik-Lutz greatly improved my chess

Bl a c k ' s play is rather passive, there­ understanding.

fore tactical possibilities remain un­ A sort of "everyday endgame" is

derneath the surface. game 42: S eemingly there isn't

With rooks and same-coloured much going o n, but little tactical

bishops "traditional" values like a p loys keep the game a liv e.

 Endgame Secrets 7

In endgames 43 and 44 we en­ who supported me during the writ­

counter queens. Lutz-Rogers is es­ ing of this book, in particular: for­

pecially remarkable as it actually mer Batsford managers Graham

contains two further endgames (a Burgess and David Cummings for

pure queen ending and an endgame · accompanying me part of the way;

of knight against pawns). present General Manager Nigel

The bizarre example 45 concludes Davies for accepting postponed

the games. deadlines; Dr Robert Hübner for

Though I still don't consider my­ making available severa! literary

self to be an endgame expert I learnt references; Stefan Kindermann for

a lot through analysing. Hopefully giving me the idea to include exer­

the reader will enjoy the same cises and Anke Koglin for proof

expenence. reading and emotional support.

Before we go on to the games I Christopher Lutz

would like to thank all the people Cologne, December 1 9 9 8

8 Endgame Secrets

Symbols used in this book

+ check

# mate

X captures

!! a brilliant move

! a good move

!? an interesting move

?! a dubious move

? a b a d move

?? a blunder

1-0 White wins

0-1 Black wins

½-½ draw

[6.2] see diagram 6 . 2

 The Endgames

Endgame 1 54 . . . ctJ a l 55 @d3 ctJb3 56 �b6

H.Stefansson-C.Lutz ctJ c l + 57 @d2 ctJa2 58 �a5 and

Manila Olympiad 1992 @c2-b2xa2. 55 @d2 é2Je3 56 �xe3

Unpublished. fxe3+ 57 @xe3 @f'l [ 1 . 2 ]

Diagram 1 . 1 Diagram 1 . 2

White to 11JOVe White to move

Obviously, Black's position is Analyse the following moves.

quite hopeless, but the game takes 58 @f4 @f6 59 g4 @g6 60 h4

sorne unexpected tums. 50 �d4+ @f6 61 g5+ @g6 62 Wg4 h6 63 h5+

@g8 51 �xc3 ctJd5 52 �xb4? A @g7 64 g6 @g8 ½-½

blunder, 52 �d4 with the idea

@e2-f3, :c6-h6-h5 wins easily. Commentary

52 ... ctJe3+ Now the rook is gone.

However, things aren't over yet as After his blunder on move 52 Ste­

the knight is trapped. 53 @e2 l2Jxc2 fansson' s concentration faded, but

54 �c5 f4 The only move, e.g. the pawn ending in position 1.2 is
1 O Endgame Secrets

still winning. White has to know the

right plan: When Black's h-pawn is

still on h7, White's king must go to

h6 before the pawns can be pushed.

White has to reach the position

w@h6 + �g5 + �h5 v b@g8 + �h7

with Black to move, then he wins

after 1...@h8 2 g6 @g8 3 g7 or

2 .. . h x g 6 3 hxg6 @g8 4 g7. With

White to move this position is only

drawn.

58 @f4 @f6
Diagram 1 . 3
5 8 .. . @g6 59 @g4 leads to analo-
Black to move
gous lines.

59 g4? (60 @ e 5 ? �h5 61 g3 @g6 62 g4

Throwing away the win. @g7 63 @f5 @fl with a draw, cf.

I) 59 g3? is wrong as well. White line I) 60 .. . h6 (60 .. . @h6 6 1 h5 @g7

needs the tempo of the g-pawn, e . g . 62 @g5 or 60 .. . @f6 6 1 @h5 lead to

59 . .. @g6 (59 .. . h6 intending 60 h4 already known positions) 61 @f4

h5 or 60 g4 @g6 6 1 h4 @f6 62 @e4 @h5 (61...@f6 62 g4) 62 g3 @g6 63

@e6 63 h5 @f6 64 @f4 @e6 also @e5 (as Bl a c k ' s pawn is on h6,

-leads to a draw, but not 59 .. . h5? 60 White wins by a detour) 63 .. . @g7

h4) 60 @e5 @g7 61 @f5 @fl 62 (63 .. . @h5 64 @f6 @g4 65 @g6

@g5 @g7 63 @h5 @ fl ! (not imme­ �xg3 66 h5 or 63 .. . @fl 64 @f5

diately 63 .. . @g8 64 @h6 and White @g7 65 g4 @fl 66 h5) 64 @f5 @fl

wins) 64 @h6 @ g 8 . It tums out that 65 g4 @g7 66 @e6 @g6 67 @e7 h5

White to move can 't win: 65 h4 (67 .. . @g7 68 h5) 68 g5 @g7 69

@h8 66 h5 @g8 67 g4 @h8 68 g5 @d6 (the king has to execute a tri­

@g8 and we have reached the draw­ angular manoeuvre as 69 @e6 @g6

ing position mentioned above. 70 @e5 @fl 71 @f5 @g7 72 · g6?

II) 59 h4! is the correct move. The @h6 73 @f6 is only stalemate)

g-pawn mustn't be pushed unless it 69 .. . @fl 70 �d7 @g7 7 1 @e7 @g6

is clear whether g2-g3 or g2-g4 is 72 @e6 (or 72 @f8) 72 .. . @g7 73

the right choice. 5 9 .. . @g6 (59 .. . h5 @f5 @fl 74 g6+ @g7 75 @g5 with

60 g3 loses on the spot. After 59 .. . a win.

@e6 60 @g5 @fl 61 @h6 @g8 62 59 @g6

h5 @h8 the winning move is 63 g4 ! , · 60 h4 @f6

while after 59 .. . @fl 60 @g5 @g7 Now the white king doesn't even

6 1 h5 @fl 62 @h6 @g8 White wins get to h6. After 6 1 @e4 Black sim­

by 63 g3!. 59 .. . h6 62 g4 @e6 63 ply replies 6 1... @ e 6 .

@e4 @f6 64 @d5 leads to the main 61 g5+ @g6

line) 60 @g4! [ 1.3 ] 62 @g4 h6

 Endgame Secrets 11

62 . .. @g7 63 @h5 @g8! 64 @h6 The right choice. The other two

@h8 is possible as well, but the text possibilities lose:

is much more simple. I) 3 8 . .. @f6? (game) squanders an

63 h5+ @g7 important tempo: 39 @fl c5

64 g6 @g8 (39 .. . @e6 is already to late since af­

After e.g. 65 @f5 @g7 66 @e6 ter 40 @e2 c5 4 1 a 4 ! the white king

@f8 67 @f6 @g8 68 g7 @h7 69 reaches the square of the �a5 and

@f7 White can only reach a 41...cxd4 42 cxd4 @d6 43 h5 @e6

stalemate. 44 @e3 loses without a chance) 40

c 4 ! [2.2]

Endgame 2

M.Degraeve-C.Hansen

German league 1 9 9 8

Unpublished.

Diagram 2 . 2

Black to move

The decisive breakthrough. Black

Diagram 2 . 1 can't stop both passed pawns:

Black to move 40 . . . cxd4 (40 . . . dxc4 41 dxc5 bxc5

42 @e2) 41 cxd5 b5 42 @e2 b4

White just played 38 d4+. How (42 .. . a4 43 @d3 b4 44 @xd4) 43 a4

should Black react? (43 axb4? a4) 43 . . . @e5 44 h5 @e4

45 d6 b3 46 @d2 1 - 0 . 46 . .. d3 is an­

Commentary swered by 4 7 @ c 3 .

II) 3 8 . . . @d6 is mistaken d u e t o 39

As 3 8 . . . @e4 3 9 h5 isn't feasible, @fl c5 40 h5 @e6 4 1 c4 etc.

Black has to put his king on either 39 @fl

d6, e6 or f6 . White threatens to put 39 @g2!? is tricky. After 3 9 . . . c5

his king on f4 to exchange his h­ 40 a4 cxd4 41 cxd4 Schwarz

pawn against Black's kingside. shouldn't go for 4 1... b 5 ? 42 axb5 a4

Therefore Black must strive to mo­ 43 b6 @d6 44 h5 as he will end up

bilize his queenside majority. For in a lost queen ending. Instead, he

that purpose, he needs the move could play 4 1.. . @ e 7 . Only after

.. . c 6 - c 5 . White plays @fl will he reply

38 @e6! . . . b6-b5.
12 Endgame Secrets

39 c5 Endgame 3

40 a4!? M.Adams-C.Lutz

Alternatives aren't dangerous ei­ Wijk aan Zee 1 9 9 5

ther, e . g . : Earlier publications: Informant

I) 40 @e2 cxd4 4 1 cxd4 b 5 . The 62/endgame 3 ; SCHACH3/95.

white king has to watch the pawns

a5 and b 5 .

11) 40 dxc5 bxc5 41 a4 @f6 42

@e2 @g6 43 @e3 @f6 44 f3 gxf3

45 @xf3 @g6 46 @f4 d4 47 cxd4

cxd4 48 @f3 @h5 49 @e2 f4.

111) 40 c4 cxd4 41 cxd5+ @xd5

42 h5 @e5.

40 cxd4

41 cxd4 b5 [ 2 . 3 ]

Diagram 3 . 1

Black to move

This innocent-looking position is

the most complicated pawn ending I

have ever encountered in tourna­

ment play. It occurred in a

30-minute rapid game and both

players were left with only a few

minutes to finish the game. This ex­

Diagram 2 . 3 plains the abrupt end: 3 0 . . . a4 3 1 h5

White to move b4! and White resigned. His king

can 't stop the passed pawns on the

a- and e-file, e.g. 3 2 @d3 bxa3 33

The white king is just outside the @c3 e4.

square of the a-pawn, therefore 42 Analyse the pawn ending and try

axb5? a4 43 b6 <.td6 even loses. Af­ to answer the following questions:

ter 42 @e2 b4 the game will be l. Can White's position be saved

drawn as neither side can make after 30 . . . a 4 ?

progress. 2 . How do you assess altematives

to 30 . . . a 4 ?

Commentary

During analysis you will certainly

get lost in a jungle of va ri at i ons.

 Endgame Secrets 13

Hopefully I managed to find a clear White has reduced the number of

answer to both questions. pawns, but the black king gets to c 3 ,

First I look at 30 .. . a4. e. g. 36 @d2 (36 @e2 @c3, 36 f4 e4

1) 30 a4 31 bxa4 (The break- 37 @d2 @d4 3 8 @e2 @c3 39 @dl

through b5-b4 is threatened, there- e3) 36 .. . @d4 with a further branch:

fore 3 1 @d3? b4 ! isn 't playable. 3 1 B1 ) 3 7 @dl @e3 3 8 c4 @d4

c4+? bxc4 32 bxa4 @c5 loses, as B2) 37 c3+ @c4 38 @c2 f4 39

does 3 1 b4 @c4 3 2 @d2 e4 3 3 fxe4 @d2 @b3 40 @d3 (40 c4 @xc4 4 1

fxe4 34 @e3 @c3 35 @xe5 @xc2 @c2 @d4 42 @d2 e4 43 fxe4 @xe4

36 @d5 @b3) 3 l .. . bxa4. Now Black 44 @e2 @d4 and Black wins)

simply intends to gobble up the 40 .. . @xa3 41 @c2 (41 c4 @b4)

pawns with . . . @c4-c3xc2-b2xa3. 4 l...e4 42 fxe4 f3 43 @d2 @b3 44

Thus White must act quickly. e5 f2 45 @e2 fl=ih 46 @xfl a3,

A) 32 @d3? h5 33 c4·+ (Altema­ and the a-pawn queens with check.

tives aren't better: 33 @e3 @c4 34 B3) 37 @e2 @c3 38 @dl @b2

@d2 @d4 35 c3+ @c4 36 @c2 e4 (both sides can promote their

37 fxe4 fxe4, 33 @c3 e4 34 fxe4+ pawns, but Black can immediately

fxe4 35 @d2 @d4, 33 g3 g6) swap the queens off) 3 9 c4 @xa3 40

3 3 . .. @c5 34 @c3 e4 35 fxe4 fxe4 c5 @b3 4 1 c6 (41 @el a3 42 @bl

and Black wins. e4) 4 1... a 3 42 c7 a2 43 c8='i'

B) 32 h5 (White gains space on al='i'+ 44 @e2 'ifh2+ 45 @el (or

the kingside and prevents the 45 @e3 'i'd4+ 46 @e2 'i'c4+ 47

cramping .. . h7-h5, but Black's king 'i'xc4+ @xc4 48 @e3 @c3 49 @e2

can penetrate) 3 2 .. . @c4 33 g4 g6! f4) 45 .. . 'i'c3+ 46 'i'xc3+ @xc3 47

(Black supports the pawn duo e5/f5, @e2 @d4 48 @d2 f4, and the �f3

and doesn 't cede any space. drops off.

33 . . . fxg4? 34 fxg4 @c3 35 @e4 C) 32 g4! is the only hope. Black

@xc2 3 6 @xe5 leads to a draw only, has four moves now, neither of

see line C) 34 hxg6 hxg6 35 gxf5 which leads to a win, however.

gxf5 [ 3 . 2 ] C I ) 32 fxg4 33 fxg4 and n o w :

C l a ) 33 @c4 34 @e4! (34 @d2?

e4 is hopeless) 34 .. . @c3 35 @xe5

@xc2 ( 3 5 .. . @b2 36 c4) 3 6 @e6 @b3

3 7 ©fl

C l a l ) 37 .. . g 6 3 8 @g7 is a simple

draw: 3 8 .. . @xa3 39 @xh7 @b4 40

h5 g5 (or 40 .. . gxh5 4 1 gxh5 a3 42

h6 a2 43 @g8 al='i' 44 h7) 4 1 h6

a3 42 @g8 a2 43 h7 al='i' 44

h8='i'.

Cla2) 37 .. . @xa3 is more spec­

tacular: 3 8 @xg7 @b4 39 @xh7 a3

40 g5 a2 4 1 g6 a l ='i' [ 3 .3 ]
Diagram 3 . 2

White to move
14 Endgame Secrets

Hoping for zugzwang. White has

eight legal moves, but only one of

them guarantees the draw:

C l b l ) 34 @d2? �d4 just loses.

C l b 2 ) 34 c 3 ? �c4 is also losing,

as White is a tempo down compared

to line Cla. The route

.. . @c4xc3-b3xa3 is a move shorter

than .. . � c 4 - c 3 x c 2 - b 3 x a 3 . The white

king has to take the same way to the

kingside, it doesn't matter whether

the e-pawn is on e5 or e4, Have a

Diagram 3 . 3
look: 3 5 �xe4 �xc3 3 6 @e5 @b3
White to move
37 @e6 .@xa3 38 @f7 @b3 39

The endgame composer Grigoriev �xg7 a3 40 @xh7 a2 4 1 g5 a l = 'iV

analysed similar positions back in 42 g6 'iVhl 43 g7 'iVxh4+ with a

1 9 4 0 . White draws as the black king won position.

is too far away: 42 g7 (42 h5!? is C l b3) 34 g5? is another false try :

possible as well) 42 .. . 'iVb l + 34 .. . g6! 3 5 c3 �c4 3 6 @xe4 �xc3

(42 .. . 'iVa7 43 h5) 43 �h8 'iVb2 44 3 7 �e5 �b3 3 8 �f6 �xa3 3 9 �g7

h 5 . Black can't make any progress, �b4 40 �xh7 (40 h5 gxh5 4 1 �xh7

e.g. 44 .. . 'iVf6 45 �h7 'iVf7 46 h6 h4) 40 .. . a 3 4 1 h5 (41 �xg6 a2 42

�c5 47 �h8 'iVf6 48 �h7 with a h5 a l = 'iV 43 @h7 'iVe5) 4 1... a 2 42

draw. hxg6 al='iV 43 g7 [ 3 . 5 ]

C l b ) 3 3 .. . e4 [ 3 .4 ]

Diagram 3 .4 Diagram 3 . 5

White to move Black to move

 Endgame Secrets 15

Black wins as was shown by

Bekey in 1906: 43 . . . ¡vhl+ 44 @g6

¡vd5 45 @h7 'iVf7 46 g6 'iVe7 47

@h8 ¡vh4+ 48 @g8 'iVg5 49 @f7

'iVfS+ and Black wins the i0ig6.

C 1 b4) 34 h5 ! is the only correct

move. 34 . . . h6 (34 . . . @c4? 35 @x e4

@c 3 36 @ d5 @xc2 37 @c4 even

los e s ) 3 5 c3 (now this is p o ssible as

the white king has access to the

squares f 5 and g6) 3 5 . . . @c4 36

@xe4 @xc 3 37 @f5 @b3 38 @ g6

Diagram 3 . 6
@x a3 3 9 @xg 7 @b 4 40 @xh 6 a3 41

Black to move
g 5 a2 42 g 6 a l = 'iV 43 g7 and draw.

C l c ) 33 h6 34 h5 is d rawn. Relatively best is 43 .. . 'iV g l + 44

C l d) 3 3 g6 34 @ d3 ( 34 g5 ? e4 @h7 'iVcS 45 @g7 �xh5 46 f8='iV

loses, but 34 h5 is drawn as well) 'iVgS+ 47 @h7 h5, but after 48 'iVf3+

34 . . . e4+ ( 34 . . . hS 35 g5 , 3 4 .. . h6 35 White draws by permanently check­

g5 h5 36 c 3, 34 . . . @ c S 35 @ e4) 35 ing his opponent.

@ e3 @c4 36 @x e4 @c3 37 @ e5 an d C4) 32 . . . g6 [3. 7]

the white king pounces upon the en­

emy pawns.

C 2) 3 2 . . . f4+ only weakens e4, e.

g . 3 3 @d 3 @c 5 34 @e4 @c4 3 5

@x e5 @c3 36 @xf4 @xc2 37 @e 3

@b3 38 @ d3 @xa3 39 @c3 @a2 40

@ c2 with a draw.

C3 ) 32 .. . @c4 (by sa cr i fic ing a

p awn the king breaks through) 33

gxf5 (33 @ d2 ? fx g4 34 fx g4 @ d4

loses) 33 @ c3 34 h5 (threatening

35 h6) 34 h6 3 5 f4 ( 3 5 @ e4 leads

to the same position w ith an addi­

Diagram 3 . 7
tional i0i f3 . T his doesn ' t m ake a dif­
White to move
fe rence here) 3 5 . . . exf4+ 36 @x f4

@xc2 37 @e5 @b3 38 @e6 @xa 3 ... is tricky. Now again Black

39 @f7 @ b3 40 @xg 7 a3 41 f6 a2 threatens . . . @c4.

42 f7 al ='iV + 43 @ g8 [ 3.6] C4a) 33 gxfó? (Originally I con­

sidered this continuation as drawn,

This position is worth remember- but IM Günther Beikert corrected a

ing, Without the pair of pawns · mistake in my earlier analysis)

h5/h6 the game is drawn immedi- 3 3 . . . gxfS 34 @d3 (34 h5 @c4 35

ately, as the black king is too far @d2 h6 is completely hopeless)

away. 34 . . . h S [ 3 . 8 ]
16 Endgame Secrets

... we arrive at a theoretically

drawn position. Black can capture

the �h4, but then the white king will

be in time to take the �a3 and hurry

to f1 . This type of position "2!::. v 1t::.

with a pair of blocked edge pawns"

was explored by W'Bahr in the

1930s. A knowledge of these posi­

tions is of an enormous practica!

relevance. Strangely, it is important

for the attacker ( i. e . the side with

the extra pawn) to have his passed

Diagram 3 . 8
pawn quite far back and rather close
White to move
to the edge pawns. The edge pawn

has to be far advanced. For exam­

3 5 c3 (the only move as 3 5 @e3
ple, the position w@d3 + �h4 v
@c4, 35 @c3 e4 36 fxe4 @xe4 37
b@d5 + �c5 + �h5 is winning for
@d2 @d4 or 3 5 c4+ @c5 36 @c3 e4
Black with either side to move, pro­
37 f4 e3 3 8 @ d 3 e3 loses)
vided he plays l...@e5 2 @c4 @f4
C4al) 3 5 . . . @c5? (leads to draw)
etc. and doesn't spoil it with
36 c4. We reach the position from
1...c4+? 2 @c3. However, in a posi­
the previous remark, but with Black
tion w@c3 + l::.h4 v b@c5 + �b5 +
to move. The point is that after
�h5 Black can 't win. As advice to
36 . . . e4+ 37 fxe4 fxe4+ 38 @xe4
the reader, I recornmend a thorough
@xc4 39 @e3 @b3 40 @d3 @xa3
analysis of these positions.
41 @c3 @a2 42 @c2 a3 43 @el
Let us retum to the game. Instead
@b3 44 @bl [ 3 . 9 ]
of 3 6 . . . e4+ Black can also try sorne

tempo manoeuvre, but against care­

ful defence this is to no avail, e.g.

36 . . . @c6 37 @c3 @d6 3 8 @d2! (or

38 @c2!, but not 38 @d3? @c5)

@c5 3 9 @ d 3 .

C4a2) 3 5 . . . e4+ 3 6 fxe4+ fxe4+ 3 7

@e3 @c4 3 8 �xe4 @xc3 3 9 @e3 is

another drawn position of the Bahr

type.

C4a3) 3 5 . . . f4 36 c4+ @c5 3 7 @c3

leads nowhere, as <loes

C4a4) 3 5 . . . @e6 36 c4.

But there are two more moves

Diagram 3 . 9 with the king which might cause

Black to move White sorne problems:

 Endgame Secrets 17

C4a5) 3 5 . . . @ d 6 3 6 @c4 (The only This move is the climax of Beik­

move. After 36 c4 @c5 Black ert' s idea. Going for the a-pawn

reaches the position he would like leads to only a draw after 43 . . . @c4?

to. The same applies to 3 6 @e3 @c5 44 @e4 @b3 45 @d3 @xa3 46 @ c 3 .

3 7 @d3 @d5 3 8 c4+ @c5) 36 . . . @c6 Instead, the h-pawn is now the tar­

(36 . . . @e6 37 @c5 f4 38 @c4) 37 get of Black's king: 44 @f3 @f5 45

@b4 (37 @d3? @d5) 3 7 . . . @d5 @e3 @g4 46 @e4 Wxh4 4 7 @f4 (4 7

(37 e4 38 fxe4 fxe4 39 @c4 or @f3 @g5 48 @g3 @f5) 47 . . . @h3 4 8

37 f4 38 @c4 @d6 3 9 @d3 don't @f3 h4 49 @f2 @g4 50 @g2 @f4

yield anything) 38 c4+ @c6 (Black has reached a won Bahr po­

(38 . . . @d4 39 c5) 39 @c3 @c5 40 sition as the �a4 has already crossed

@d3 with a draw. the middle of the board) 51 �h3

C4a6) 3 5 . . . @c6! (this move of �e3 52 @xh4 @d2 53 @g3 @c2 54

Beikert leads to a win) 36 @c4 (36 @f2 @b2 55 @el @xa3 56 @dl

c4 @c5 or 36 @e3 @c5 lose imme­ @b2 and the a-pawn queens.

diately) 3 6 . . . @d6 3 7 @b4 (37 @b5

e4 3 8 fxe4 fxe4 3 9 @c4 @e5 and 3 7 Back to diagram 3 . 7 . In the light

@d3 @d5 lose) 3 7 . . . e4 (37 . . . @ d 5 ? of the previous variations it be­

38 c4+ @c6 39 Wc3 is drawn) 38 comes clear why

fxe4 fxe4 39 c4 (39 Wc4 @e5) C4b) 3 3 @d3 ! (Beikert) leads to a

39 . . . @c6 ( 3 9 . . . e3? 40 @c3 @c5 41 draw:

@d3) 40 c5 (40 @c3 @c5) 40 . . . e3 C4bl) 33 fxg4 34 fxg4 and

41 �c3 @xc5 42 @d3 @d5 43 C4b2) 3 3 e4+ 34 fxe4+ fxe4+ 3 5

@xe3 @ e 5 ! ! [3.10] @e3 lead to drawn positions we al­

ready looked at.

C4b3) 3 3 . . . @c5 34 gxf5 gxf5 35

h5 (35 c4? h5 or 3 5 c 3 ? h6 36 h5

@d5 lose) 3 5 . . . h6 (35 . . . @d5 3 6 c4+

@c5 37 h6) 36 c4 with a draw.

C4b4) 3 3 . . . h6 34 gxf5 gxf5 3 5 h5

@c6 (35 . . . @c5 36 c4) 3 6 @c4 �d6

37 @b4 e 4 · (37 . . . @d5 38 c4+) 38

fxe4 fxe4 39 @c3 and @d2. Black

can reach a Báhr type position

which is drawn only.

C4b5) 3 3 .. . h5 34 c4+ (now the

point of 3 3 @d 3 ! ·shows) 34 .. . @c5

Diagram 3 . 1 O (34 .. . @c6 3 5 gxf5 gxf5 36 @c3 is a

White to move draw, see above) 3 5 g5 ! [ 3 .1 1 ]

1 8 Endgame Secrets

... doesn 't win either. Black re­

strains the �c2 and intends to push

his e-pawn. 3 1 a4? e4 is now hope­

less for White, therefore only two

possibilities remain.

A) 3 1 axb4 axb4 3 2 c4+ (32 @d3

e4+ or 32 @d2 @d4 lose, but 32 c3

is playable as well) 32 . . . bxc3 33

@d3. lf White could succeeds in

playing 34 @xc3, his distant passed

pawn will give Black trouble, there­

fore: 3 3 . . . c 2 ! 34 @xc2 @d4 [ 3 . 13 ]

Diagram 3 . 1 1

Black to move

Thanks to this possibility (which

White lacks after 3 3 gxf5?) the bur­

den to move is on Black now.

3 5 . . . e4+? leads after 3 6 fxe4 fxe4+

3 7 @xe4 @xc4 3 8 @e5 @b3 3 9 @f6

@xa3 40 @xg6 @b4 4 1 @xh5 a3 42

g6 a2 43 g7 al=¡v 44 g8=¡v to a

-queen ending with a white extra

pawn (which is however drawn ac­

cording to Ken Thompson's end­

game data base). Thus we answered

Diagram 3 . 1 3
the first question: 3 0 . . . a4 is only a
White to move
draw. To answer the second ques­

tion I analyse the moves 3 0 . . . b4,

The black king is invading the
3 0 . . . e4 and 30 . . . h5.
white camp. Black threatens . . . @e3,
11) 3 0 . . . b4 [ 3 . 1 2 ] . . . e5-e4, . . . ©f2, . . . e 4-e 3 - e2-e l = ¡v

with transition to a favourable

queen ending:

A l ) 3 5 b4? @c4 and

A2) 3 5 @b2? e4 36 fxe4 fxe4 37

b4 (3 7 <it>c2 ©e3) 3 7 . . . @c4 are

unedifying.

A3) 35 g4? (weakens the �f3)

3 5 . . . g6! ( 3 5 . . . fxg4 36 fxg4 @e3 37

b4 @f2 3 8 b5 e4 3 9 b6 e3 40 b7 e2

41 b8=¡v el=¡v 42 ¡vf8+ @g2 43

¡vxg7 ¡vxh4 44 g5 is dead drawn.

3 5 . . . @e3 36 gxf5 @xf3 3 7 @d2 @f2

38 @d3 @f3 39 @d2 leads to a

Diagram 3 . 1 2 repetition of moves) 3 6 @d2 (36 g5

White to move @e3 37 b4 @d4 38 @b3 f4 or 36

 Endgame Secrets 19

gxf5 gxf5 37 <iÍ>d2 h5 isn't desirable. White has to take the opposition

36 h5 <it>e3 37 hxg6 hxg6 38 gxf5 as soon the Black king reaches c5 or

gxf5 39 b4 <ii>f3 40 b5 e4 4 1 b6 e3 d5) 36 .. . @e3 3 7 b4 <ii>f2 3 8 b5 e4 3 9
42 b7 e2 43 b8='iV e l = 'iV is a win
fxe4 fxe4 40 b6 e3 41 b7 e2 42
for Black according to the databas e)
b8=¡y e l = 'iV 43 'iVa7+ and White
3 6 .. . h5 37 g5 f4 3 8 <ii>c2 @e3 leads
takes on h7.
to queen ending which is lost for
B) 31 c4+ bxc3 ( 3 1... <it> c 5 32 a4)
White.
3 2 <ii>d3 c2 3 3 <ii>xc2 @d4 is analo­
A4) 3 5 <ii>d2?! (not the most accu­
gous to line A. White simply plays
rate) 3 5 .. . h5 3 6 <ii>c2 ( 3 6 g 3 ? g6 3 7
34 h5 <ii>e3 3 5 b4 axb4 3 6 axb4 e4
@c2 <it>e3 38 b4 <ii>xf3 is lost)
3 7 fxe4 fxe4 3 8 b5 and so on.
36 . .. @e3 37 b4 e4 3 8 fxe4 fxe4 39
Thus 3 0 .. . b4 is a draw.
b5 @f2 40 b6 e3 4 1 b7 e2 42 b8='iV

el=� 43 'iVf8+ <ii>xg2 44 'iVxg7+

III) But 3 0 .. . e4! wins. Black
<ii> h 3 . Black captures the h-pawn,
threatens .. . b5-b4.
though the position is still a draw.
A) 3 1 fxe4+ fxe4 [ 3 . 1 5 ]
A5) 3 5 h 5 ! (in anticipation of the

queen ending White pushes his

h-pawn)

A5a) 3 5 .. . <it>e3 36 b4 e4 (36 .. . <it>d4

37 �b3) 37 fxe4 fxe4 3 8 b5 <ii>f2 3 9

b6 e3 40 b7 e2 4 1 b8='iV e l = 'iV 42

¡yf8+ <iÍ>gl 43 'iVxg7 'iVe2+ 44 <ii>c3

'iVxh5 45 g4 with a draw.

A5b) 3 5 .. . g6 36 h6! (White blocks

the �h7. This move can follow after

3 5 .. . g5, too. Drawish as well is 36

hxg6 hxg6 3 7 <ii>d2 g5 3 8 g3 f4 39

gxf4 gxf4 40 b4 <it>c4 41 b5 <it>xb5

42 �d3 <it>c5 43 <it>c3 [ 3 . 1 4 ]. Diagram 3 . 1 5

White to move

. .. loses in a straightforward way:

3 2 @d2 (32 c3 a4 3 3 bxa4 bxa4 34

g4 <ii>c4 3 5 <ii>xe4 @xc3, 32 b4 axb4

3 3 axb4 h5 or 32 h5 b4 lose as well)

3 2 .. . <ii>d4 3 3 c3+ <it>d5 together with

.. . a5-a4 and penetration of the black

king.

B) 3 1 c3 h5 32 g3 g6 and .. . a5-a4

is similar.

C) 31 g4 is the most stubbom

answer.

Diagram 3 . 1 4 C l ) 3 1... fx g 4 32 fxe4+ <it>e5 3 3 c4

Black to move a4 (33 .. . bxc4 34 bxc4 g3? 35 c5

20 Endgame Secrets

even loses) 34 bxa4 bxc4 3 5 a5 and C 3 b 3 ) 34 h5 g5 3 5 hxg6 hxg6, 3 5

White has no problems. h6 b4 or 35 c4 bxc4 36 bxc4 h6,

C2) 3 1... g 6 3 2 gxf5 gxf5 3 3 fxe4+ with a win for Black in either case.

fxe4 34 c4+ bxc4 35 bxc4+ @xc4 C3b4) 34 @e3 34 . . . @e5 35 @d3

36 @xe4 @b3 37 @d3 @xa3 38 @d5 ( 3 5 . . . g6? 3 6 c4 bxc4+ 3 7 @xc4

@c3 h5 39 @c4 is another drawn is only drawn, but now Black wins

position of the Bahr type. as he captures White's a-pawn)

C3) 3 1... e x f3 !. wins: C 3 b4 1 ) 36 a4 bxa4 3 7 bxa4 g6 3 8

C3a) 3 2 @xf3 fxg4+ 3 3 @xg4 b4 @e3 h6 39 @d3 (39 @f4 @d4)

34 axb4 axb4 3 5 @f5 @d4 3 6 @e6 3 9 . . . @c5 40 @e4 @b4 4 1 @e5 g 5 .

@c3 3 7 @f7 @xc2 3 8 @xg7 (3 8 h5 C3b42) 36 c4+ bxc4+ 37 bxc4+

g5) 3 8 . . . h5 @c5 38 a4 ( 3 8 h5 a4 and . . . g7-g5)

C3b) 32 gxf5 3 2 .. . @e5 ( 3 2 . . . b4? 3 8 . . . g6 39 �c3 h6 40 @d3 g5 4 1 h5

3 3 a x b 4 axb4 34 @xf3 @e5 3 5 @e3 g4 42 @e4 @xc4 43 @f4 @b4 44

@xf5 36 c4 with a draw) 33 @xf3 @xg4@xa4.

@xf5 [ 3 . 1 6 ] Thus Black wins after 30 . . . e4!.

If you look at the lines above, you

will discover that White often gets

counterplay along the moves g2-g4

and h4-h5, while Black can prevent

these possibilities with . . . h7-h5.

Therefore another move comes to

mind:

IV) 3 0 . . . h S ! [ 3 . 1 7 ]

Diagram 3 . 1 6

White to move

Black wins as his king is more ac­

tive. He intends the creation of a

passed pawn with . . . g7-g6, . . . h7-h6,

. . . g6-g5. White has no defence:

C 3 b l ) 34 b4 a4 3 5 @e3 ( 3 5 h5 g6

3 6 h6 g5) 3 5 @e5 3 6 @d3 @d5 3 7 Diagram 3 . 1 7

@e3 g6 and h7-h6, . . . g6-g5. White to move

C3b2) 34 c4 bxc4 35 bxc4 @e5

3 6 @e3 (36 @g4 @d4 37 @f5 @xc4 Indeed, this move wins even more

38 @e6 @b3 39 @f7 @xa3 40 simply than 3 0 . . . e4. After 31 @d3

@xg7 h5) 36 . . . g6 37 a4 h6 and Black has the pleasant choice

. . . g6-g5. between
 Endgame Secrets 21

A) 3 l...e4+ 32 fxe4+ fxe4+ 33

@ d 2 @d4 ( 3 3 . . . a4? 34 bxa4 bxa4 3 5

@e3 @c4 36 @xe4 @c3 37 @d5

@xc2 3 8 @c4 is erroneous) 34 c3+

@d5 and . . . a5-a4 or

B) 3 l...a4, e. g. 3 2 bxa4 bxa4 33

g3 g6 34 @c3 e4 3 5 fxe4+ fxe4 3 6

@d2@d4.

Thus Black has two winning

moves, 30 . . . e4 and 30 . . . h5, while

30 b4 and the game continuation

30 a4 are only drawn. I hope the

Diagram 4 . 2
reader has gained an im_pression of
Black to move
the complexities of pawn endings.

. . . and comment: "Black is losing

the �g6 along with the game be­

Endgame 4
cause of zugzwang". Do you agree ?
V.Hort-M.Bertok (Analysis)

Zagreb 1 9 6 9
Commentary
Earlier publication:

Winning Endgame Technique

The · assessment is not correct. In
(A.Beliavsky and A.Mikhalchishin).
position 4 . 2 Black has a surprising

reply:

60 ... �f7!

N ow 61 exf7 @xfl leads to a

theoretical draw. In comparison to

the well known wrong-bishop end­

game Black has an extra g-pawn,

However, White can't exploit this

circumstance, e.g. 62 @h6 @g8 63

�a2+ @h8 64 i..b3 g5 and White

has to take the pawn or to stalemate

his opponent. Other tries aren't

crowned by success either, e. g . :

Diagram 4 . 1 61 e7 �es

White to move 62 �a2

Otherwise 62 . . . @f7.

This position originates from an 62 ... �b5

analysis of Beliavsky and Mikhal­ 63 �e6 �es

chishin. They give the line 56 gxf6+ 56 e5 fxe5+ 57 @e5 �b5 is a

@xf6 57 e5+ @g7 5 8 @g5 i..c6 59 draw as well, therefore White can 't
1

i..bl �e8 60 e6 [ 4 . 2 ] , win in position 4 . 1 .

22 Endgame Secrets

Endgame 5 win anymore. On the other hand we

J .Hickl-C.Lutz reach a theoretical draw after 44 .. . b4

Nussloch 1996 45 .txb4 .txb4 46 @g2, cf. end­

Earlier publication: game 4) 45 J.a5 .te7 46 J.c3 .tf6

Informant 66/ endgame 1 1 . 47 .tb4 .tes 48 .ta5 .tb2 49 .tb4

.tf6 (Black manoeuvres to and fro,

but in the end his king has to go to

the b-pawn) 50 .tas .td4 51 .tb4

@e5 52 g4 hxg4+ (52 . .. h4 is to no

effect as Black can 't exchange the

bishops, e.g. 53 .tas @d5 54 .tb4

@c4 55 .tas .tc3 56 .txc3 @xc3

57 g5) 5 3 @xg4 [ 5 . 2 ]

Diagram 5 . 1

Black to m o v e

Black was lucky to survive a diffi­

cult middlegame and now enjoys an

extra pawn. With 3 9 .. . h5 Black ties

down the weakness g3. 40 @c2

How should Black continue ?

Diagram 5 . 2

Commentary Black to move

Black must combine the attack N ow we have reached the afore­

against the �g3 and the advance of mentioned endgame. According to

the b-pawn. At a convenient mo­ Ken Thompson's endgame database

ment he will go for an endgame "B the position is a draw. 5 3 .. . @dS 54

+ � v B " . Unfortunately I had too @f3 @c4 55 .te7 (55 .td2 is also

little time left to put this idea into drawn) 55 . .. .tc5 (Black can push

effect correctly. away the white bishop, but for that

I) 40 . .. .td6? was the game con­ he needs time-time for the white

tinuation. Oddly, Black can now no king to hurry over) 56 .th4 (56

longer win. 41 .t e l @f7 42 @d3 .tgS? @d3 is a mistake as then nei­

@e6 43 @e4 b5 44 @f3 (White is ther bishop nor king of White can

prepared to exchange pawns with stop the b-pawn. N ow there is no

g3-g4) 44 . .. @fS (After 44 .. . @dS 45 need to fear 56 .. . @d3 57 .t e l .td4

g4 hxg4+ 46 @xg4 the white king is 5 8 .tb4) 56 . .. b 4 57 @e2 b3 [ 5 . 3 ]

too close to the b-pawn, Black can 't

 Endgame Secrets 23

�g3 and to control the square b4.

This aim he achieves from the

squares e l and d6. After the "natu­

ral" 40 . . . jl,d6? White can occupy

the safe square e1. Thus better is

40 . . . .i. f2 ! 41 .i.eS (the only move,

but now the black king can attack

the bishop) 4 1... @ f7 42 @d3 @e6.

White can react in two different

ways, but the result is the same:

A) 43 j¿,f4 b5 44 @e2 jl, c S 45

jl,d2 ( 45 @f3 b4 46 g4 b3 4 7 .iL.C 1

Diagram 5 . 3
hxg4+ 4 8 @xg4 @d5 49 @f3 @d4)
White to move
45 .. . @d5 46 @f3 @c4 47 g4 hxg4+

48 @xg4 @d3 (the white king is too

58 @dl (There is still one obsta­
far away) 49 .tas jl,d4 50 jl,b4 @c4
ele White has to negotiate: 5 8 @d2?
51 jl,e7 .iL.cS 52 ..th4 b4 [ 5 .4 ] , and
.i. e 3 + ! 59 @ d l @d3 60 .tf6 .i.d4 is
the b-pawn marches through.
winning for Black. Dehler and

Cheron analysed this position al­

ready in the 1920s) 58 . . . i.e3 59

i.f6 @d3 60 .i.b2 (The only move.

White has to counter . . . .iL.e3 with

.i.b2. However, if the black bishop

is not on e3 White must avoid the

square b2. Were it White's tum to

move in the present position, he

would lose, e.g. 1 .iL.f6 .iL.d4 or 1

i.a3 @c3 2 .i.e7 @b2 3 .iL.f6+ @a2

4 .i.eS jl,cS 5 i.f6 .ia3 6 .i.eS i.b2

7 .i.f4 .i.f6 8 i. c l i.gS) 60 . . . i.f4 6 1

jl,f6 jl,d2 62 i.g7 j¿,f4 63 jl,f6 .ig5

64 �g7 i.e3 65 .ib2 @c4 66 i.f6 Diagram 5 .4

@b4 (66 . . . jl,d4 65 jl,e7 is to no White to move

avail) 67 .ib2 @c5 68 .iL.cl @d4 69

.iL.xe3+ @xe3 70 @e 1 @d3 7 1 @b2 B) 43 @e4 b5 44 i.c7 (the square

@c4 72 @ b l @c3 73 @el b2+ 74 d6 is denied) 44 .. . b4 45 @d3 @f5

@b 1 @b3 and finally the game was (while White captures the �b4

drawn. Black eliminates the 8g3) 46 @c4

II) If you take a closer look at the @g4 47 @xb4 jl,xg3 48 �d8 .iL.h4!

position befo re Black' s 40th move (the only winning move) 49 .iL.c7

you will detect that the white bishop i.gS 50 @c3 h4 5 1 @c2 i.f4 [ 5 . 5 ] ,

has two tasks: he has to protect the and the h-pawn promotes.
24 Endgame Secrets

Endgame 6

C.Lutz-A.Zude (Analysis)

German championship 1 9 9 6

Unpublished.

Diagram 5 . 5

White to move

III) But there is another winning

move viz. 40 .. . W f7 !. After 4 1 Wd3

i.. f2 ! transposes to the variation Diagram 6 . 1

above, while after 4 1 i..e 1 We6 42 Black to move

�d3 @d5 [5.6] the white king has

to fall back d u e t o zugzwang. Black dispensed with this position

as he wasn't sure whether he (to

move) can reach a draw. What do

you think?

Commentary

The black position looks critical,

as after capturing the �g4 White has

passed pawns on the b- and f-file.

But Black's bishop blocks the

square b6, while his king blocks f6.

To overcome the blockade White

has to bring his king to d5 or to g8

Diagram 5 . 6
(via h6). Against careful defence
White to move
this is not possible as Black might

use bis a-pawn to divert the white

bishop. If the a-pawn is exchanged

against one of the f-pawns positions

arise which were analysed in Speel­

man's Analysing the Endgame.

Things could proceed like this:

50 ... a4

51 i..xg4 a3

52 i..e6 @f6
 Endgame Secrets 25

53 i.a2 [6.2] Neither 60 b6 i.xb6 61 @g5

i.d8+ nor 60 @h5 @f6 yield White

anything.

60 @f6

61 f4 i.b6

62 i.bl i.c5

63 @f3 [6.3]

Diagram 6 . 2

Black to move

The a-pawn is blocked. But White

can't go on.

53 ... i.b6

54 @f3 @f5 Diagram 6 . 3

55 i.bl+ @f6 Black to move

56 @g3
63 ... i.b6
56 @e2 i.a7 57 f3 a2 58 i.xa2
63 . . . a2? 64 i.xa2 @xf5 65 i.bl+
@f5 59 @fl @xf4 and Black estab­
and f4-f5 lead to positions analysed
lishes a blockade. He shuffles his
by Speelman. White wins, e.g.
bishop along the diagonal a7-gl and
65 . . . @f6 66 f5 @e5 67 @g4 i.e3 68
waits. As soon as the white king
@h5 @f6 69 i.c2 i.b6 70 @h6
reaches d3 or c4, his king will go to
i.e3+ 7 1 @h7 @fl 72 i.b3+ @f6
e5. If the white king goes to h3,
73 i.e6 [6.4].
there will follow . .. i.f2 or . . . @g5,

e.g. 60 @g2 i.b6 6 1 i.d5 i.c5 62

@h3 @g5 63 @g3 i.d6+ 64 @f2

i.c5+ 65 @e2 @f4. '

56 i.c5

57 f3

The backward f-pawn remains

White's biggest headache: If it stays

on f2 , the king is tied to its defence.

If it advances to f3 , it blocks the

king.

57 ... i.b6

58 @g4 i.e3

59 f5 @e5 · Diagram 6.4

60 i.a2 Black to m o v e
26 Endgame Secrets

Due to zugzwang Black must al­ Black must play move 40 to reach

low White to bring the king to e8 the time control. What would you

respective to advance of the f-pawn. play?

64 @e4 i.c5

65 @d5 i.e3[6.�] Commentary

Once again the last move before

the time control tums out to be the

decisive mistake.

40 élJ c l ?

Throwing away the game.

40 .. . élJc5 is correct: 41 f3 (4 1 b4

élJxe4 42 lt:ixa6 CDc3) 4 1... b 4 with

an equal game (Shirov). The CDc5

and the �b4 build a blockade

against the white king. Though

White can continue to attack the

queenside, Black has enough coun­

Diagram 6 . 5 terplay, e.g. 42 @f2 @g8 43 �e3

White to move @fl 44 é2Jd5 a5 45 é2Jb6 @e6 46

élJc4 a 4 . 47 bxa4 élJxa4 48 @d3

White can't get further, as he has
CDc5+ 49 @c2 f5 [7.2].
'to take care of the �f4. Therefore

the starting position is drawn.

Endgame 7

A.Shirov-J .Piket

Madrid 1 9 9 7

Earlier publication: Informant

69/338 (A.Shirov).

Diagram 7 . 2

White to move

41 �xa6 étJx�
42 @fl!

lt becomes clear that the élJa2 is

trapped. To free this piece Black has

Diagram 7 . 1 to give the b-pawn.
B l a c kto move 42 �el
 Endgame Secrets 27

42 . . . @g8 43 @el @f7 44 @d2 46 ltJc5+ @a7 47 ltJd3 @b7 48 f4

@e7 45 @c2 b4 46 @d3 (Shirov). g6 49 e5 i..e7 50 ltJc5+ @a7 5 1 g5

43 ltJc5 ct:Ja2 i..f8 52 ltJe4 @b8 53 ltJd6 @c7 54

44 ltJd3 @g8 ltJxf7 i..g7 55 ltJd6 @d7 56 @c5

45 @el @f7 i..f8 57 @b6 i..e7 58 @xa6 @c7 59

46 @d2 b4 b5 cxb5 60 @xb5 @b8 61 @c6

47 ltJc5 @e7 i..d8 62 ltJc4 1-0

Loses at once. 47 . . . ltJc3 48 @d3

@e7 (48 . . . ct:Jdl 49 f3) 49 ct:Ja6 ltJa2 Commentary

50 @c4 (Shirov) would have only

Besides his space disadvantage
deferred the end: 50 . . . ltJc3 51 f3
Black has to take care of the a6 and
ltJe2 ( 5 1... ltJ d l 52 @d3) 52 ltJxb4
c6 pawns. But he can scrape a draw.
ltJf4 53 ltJd5+.
Especially important are possible
48 ltJa4 1-0
pawn endings which arise after tak­
Black is defenceless against
ing on c 5 .
@c2-b2.
46 ltJc5+ @a7

The transition to the pawn ending

Endgame 8
now loses: 46 . . . i..xc5 47 @xc5 @c7
J.Piket-J.Lautier
48 e5 f6 (48 . . . g5 49 f3 and the white
Monaco rapid toumament 1 9 9 6
king gets to b6 or d6) 49 exf6 gxf6
Unpublished.
50 f4 @d7 5 1 g5 fxg5 52 fxg5 e5 5 3

g6 @e6 54 g7 @f7 55 -@xc6 e4 56

b5 e3 57 bxa6 e2 58 a7 el='iY 59

g8=1W+ @xg8 60 a8=1W+ with a

won position according to Kent

Thompson's endgame database.

47 'Lld3 @b7?

The losing move, as now the

kingside pawns will be fixed on the

light squares. 47 . .. e5 draws, the idea

is to place the pawns on e5, f6 and

g5. The bishop, though it is then

nominally bad, can protect the

Diagram 8 . 1 pawns, e. g. 48 ltJc5 (48 f4 exf4 49

White to move e5 i..c7 or 48 g 5 ! ? @b7 49 ltJc5+

i..xc5 50 @xc5 @c7 5 1 f3 g6 are no

Analyse the game. Can Black de­ improvement) 4 8 . . . g5 49 ltJd7 @b7

fend in a better way? 50 ltJf6 i..e7 51 ct:Jd7 f6 52 ltJc5+

28 Endgame Secrets

@a7 [ 8 . 2 ]. White can't make any

progress.

Diagram 8 . 3

Black to mo v e

IV) 48 . . . @c8 is somewhat more

complicated: 49 e5 i.e7 50 lbc5

Diagram 8 . 2
A) 50 . . . g5 51 fxg5 i.xg5 52
White to move
lbxa6 i.f4 53 4Jc5 (53 @d4 i.d2)

5 3 . . . i.xe5 54 a6 i.b8 (54 . . . @b8 55

48 f4 g6 lbd7+) 55 g5 (White wins as the

Weakens the square f6, but this black pieces can 't guard the ¿Ja6

doesn 't matter anymore: and the kingside at the same time)

I) 48 . . . f6 49 l2:Jc5+ i.xc5 50 @xc5 55 @d8 (55 . . . i.a7 56 l2:Jxe6,

@c7 51 g5 fxg5 (51...f5 52 exf5 55 e5 56 @d3 cJ;;d8 57 @e4 @e7

exf5 53 g6) 52 fxg5 e5 53 g6. 5 8 @f5 and lbd3) 56 lbd3 @c8 57

II) 48 . . . @c7 49 e5 i.e7 50 lbc5 @c5 @c7 5 8 ctJe5 [ 8 .4 ] , and the ¿Jfl

i.xc5 5 1. @xc5 f6 52 exf6 gxf6 5 3 is lost.

g5.

III) 48 . . . @a7 49 e5 i.e7 50 lbc5

g5 (50 . . . i.f8 51 g5 with the threat

52 g6, and after 5 1... g 6 52 lbe4 we

reach the game) 51 fxg5 i.xg5 52

lbe4 i.e7 (52 . . . i.f4 53 @d4 @b8 54

lbd6 f6 5 5 exf6 i.xd6 56 g5 wins)

53 lbd6 i.xd6 54 exd6 @b7 55 g5

@c8 56 @d4 �d8 57 �e5 @d7 58

@f6 �xd6 59 cJ;;xf7 e5 60 g6 [8.3]

and White promotes his pawn first.

Diagram 8 .4

Black to mo v e

B) 50 i.xc5 5 1 @xc5 @c7 52 f5

exf5 (52 g6 53 fxe6 fxe6 54 g5) 53

 Endgame Secrets 29

gxf5 @d7 (53 . . . f6 54 e6, 53 . . . g6 54 Endgame 9

e6, 5 3 . . . g5 54 @d4 @d7 55 @e4 R.Ekstroem-C.Lutz

@e7 56 @f3 f6 57 exf6+ @f7 58 Dresden zonal tournament 1 9 9 8

@g3 @xf6 59 @g4) 54 @b6 g5 55 Unpublished.

fxg6 fxg6 56 @xa6 g5 57 @b7 g4

58 a6 g3 59 a7 g2 60 a8=iY g l = iY

61 iYc8+ [8.5] with a won queen

ending.

Diagram 9 . 1

White to move

White wants to transform his extra

pawn into a win. He has two moves

Diagram 8 . 5 available: exchanging queens with

Black to move 41 iYxe6+ or centralizing the king

with 4 1 @ fl , which allows an ex­

49 e5 ile7 change on e3. Which move is

50 t:bc5+ @a7 better?

50 . . . ilxc5 5 1 @xc5 @c7 52 g 5 .

51 g5 ilf8 Commentary

52 t:be4

Getting to d6 decides the issue. Immediately after the time control

52 ... @b8 White squanders the win.

53 l'.bd6 @c7 I) 4 1 iY x e 6 +leads to a win. After

53 . . . ilxd6 54 exd6 @c8 55 @c5 4 1.. . @ x e 6 White has a plan consist­

@d7 56 @b6 @xd6 57 @xa6 @c7 ing of three steps:

58 b5. 1 . The king gets to the queenside.

54 l'.bxf7 ilg7 2. With a tempo manoeuvre he

55 tbd6 @d7 forces . . . b6-b5.

56 @c5 ilf8 3 . After a3-a4 the passed pawn on

57 @b6 ile7 the b-file enables the white king to

58 @xa6 @c7 enter the enemy camp.

59 b5 cxb5 A possible execution of this plan

60 @xb5 @b8 could look like this: 42 @fl ile7 43

61 @c6 ild8 @e2 @d7 44 @d3 @c6 45 @c3 @b5

62 t:bc4 1-0 46 @b3 (step one is done, step two

3O Endgame Secrets

is more complicated) 46 . . . j¿,d8 47 two is comple t ed) 5 5 . . . @d7 56 a4

Ci:Jg7 j¿,e7 48 Ci:Jf5 .id8 (48 . . . .if8 b x a4+ 57 Wxa4 @c6 58 b5+ @ b7

loses more quickly: 49 a4+ @c6 50 59 @b4 Wb6 ( 5 9 . . . .i b6 60 Ci:Jxf6

b5+ @d7 51 Ci:Je3 @e6 52 a5 bxa5 j¿,xd4 6 1 Ci:Jxh7 is ho p eless as well)

53 @a4 .ib4 54 b6 @d6 55 b7 @c7 60 f3 ! [ 9 . 3 ]. Now this tempo comes

5 6 Ci:Jxd5+) 49 Ci:J h 6 ! [ 9 . 2 ]. in handy. White' s king gets to c5,

with an easy win.

Diagram 9 . 2
Diagram 9 . 3
Black to move
B l a c kto move

This move is the beginning of a

II) However, in the game there
fine tempo manoeuvre to drive back
fo ll o w ed 4 1 W fl ? 1i'xe3 42 fxe3 b 5 !
the black king. After 49 a4+?! @c6
[9 .4].
50 b5+ @d7 White has difficulties

creating a passed pawn. 49 f3 ? ! im­

mediately forces the black king to

retum, but White shouldn 't waste

the reserve tempo f1-f3 that easily

as he can't gain a tempo with the

knight alone. Furthermore the

squares e3 and g3 are weakened.

49 . . . .ie7 50 Ci:Jg8 .id8 51 @c3! (the

white king carries out a triangula­

tion) 5 1... @ c 6 ( 5 1... W a 4 52 Wb2,

and after 52 . . . Wb5 53 Wb3 the king

is driven back, while 52 . . . b5? 53

@a2 leads to a selfmate) 52 @c2 Diagram 9 .4

@d7 53 Wb3 @c6 (53 . . . We6 54 @a4 White to move

and @b5) 54 @a4 b5+ 5 5 Wb3 (step

 Endgame Secrets 31

Since his knight got caught on the Endgame J O

edge White can no longer win: 43 L. Yurtaev-G .Serper

@e2 (43 @g2 @g6 44 'Llg3 i.xg3 US S R 1 98 8

45 @xg3 f5 leads to the next note) E a rl ier publica ions: t lnformant

43 . . . @g6 44 @f3 f5. Here White re­ 47/ 8 2 1 ( G .Se rp er ), Winning End­

alized his misfortune: game Technique ( A . B e ia

l v s ky and

A) 45 4Jg3 i.xg3 46 @xg3 h5 A. M i kh a lc hishin)

leads to a drawn pawn ending, in

spite of White's extra pawn. White

even has to play accurately as Black

can create a distant passed pawn on

the g-file, e.g. 47 @f3 fxg4+ 48

hxg4 hxg4+!? (48 . . . h4 is immedi­

ately drawn as neither king can be­

come active) 49 @xg4 @f6 50 a4

(50 @f3 @f5 and . . . g5-g4, @f5-e4

only troubles White) 50 bxa4 51

b5 a3 52 b6 a2 53 b7 al='i' 54

b8='i'. Though Black has a sym­

bolic edge he can 't win as his king

D ia gr am 1 0 . 1

is rather exposed, e.g. 54 . . . 'i'dl + 5 5

White to move
@g3 'i'g 1 + 56 @f3 g4+ 57 @e2

'i'g2+ 5 8 @ e l . Analyse the ga m e .

B ) 45 @e2 i.c7 46 @d3 i.d6 47

@c3 i.c7 4 8 @b3 i.d6 and White 1 étJe 4 i.b6 2 fS gxfS 3 étJf 6+

has to retum as 49 a4 bxa4+ 50 @g7 4 étJxhS+ @g6 5 ttJf6 @g5 6

@xa4 f4 loses the knight. 4J e 8 i.f2 7 4J d6 @g6 8 @c6 i.d4 9

C) 45 @g2 therefore followed in c4

étJ i.c3 10 @dS @gS 11 4J d 6

the game, with a draw after @g6 1 2 étJbS i.b2 13 'Ll d4 @g5 1 4

4 5 . . . i.c7 46 @f3 i.d6 47 @g2 i.c7 g3 i.c3 15 4Jf3 @g6 16 @d6 i.b4+

4 8 @ f3 . 17 @d7 i.c3 18 @e7 @g7 1 9 @d7

@g6 20 e6 fxe6 21 @xe6 i.f6 22

étJeS + i.xeS 23 @xe5 @g5 24 h4+

@g4 25 h5 1-0

Commentary

Both S e rp er and Belia v s ky /

M i kh a lc hishin consider position

10 .1 as wi nn ing for Wh te i . B esides

the ext a pawn White s

r ' cen tr a ized
l

king ca rr ies weight. Furt he rm ore a

knight is stronger than a bishop if

32 Endgame Secrets

all pawns are on one wing. Never­ Black draws, e.g. 1 4 t2Je2 i.d2 1 5

theless I have doubts about the cor­ t2Jxf4 @g3 or 1 4 @f3 @g5.

rect assessment of the position. 8 ... i.d4

1 t2Je4 .i.b6 9 cbc4 i.c3

1...@g7 2 @d6 .i.b6 3 @e7 i.e3 4 10 @d5 @g5

t2Jd6 with a win according to Ser­ 11 ctJd6 @g6

per. After the exchange of the f­ 12 ctJbS i.b2

pawns the passed e-pawn will de­ 13 cbd4 @g5

cide the issue. 14 g3

2 5 gx5 By putting his pawn on a dark

Black shouldn't allow f5-f6 and square, White prevents . . . f5-f4.

e5-e6. But now the h-pawn is lost. 14 ... i.c3

3 t2Jf6+ @g7 15 t'.l)f3+ @g6

4 t2Jxh5+ @g6 16 @d6 i.b4+

5 t2Jf6 @g5 17 @d7 .i.c3

6 t2Je8 i.f2 18 @e7 @g7

7 t2Jd6 @g6 19 @d7 @g6

8 @c6 20 e6 fxe6

At a suitable moment White has 21 @xe6 [ 1 0 . 3 ]

to swap off a pair of pawns with

e5-e6 to win the remaining f-pawn.

Beliavsky/Mikhalchishin give 8 e6

fxe6 9 @xe6 f4 1O @e5 i.g3 11

cbf5 i.h2 1 2 cbh4 @h5 13 cbf3 i.g3

1 4 @f5 together with cbd4-e2xf3 as

a win. But Serper recommends

1 0 . . . i.e3 1 1 t2Jf5 @g5 1 2 @e4 i. c l

1 3 t2Jd4 @h4 [ 10 . 2 ].

Diagram 10.3

Black to move

21 .tf6?

The decisive error. According to

Serper White wins after 2 1... .i. a 5 22

t2Je5+ @g5 23 h4+. But I do not

agree after 23 . . . @h5 24 @xf5 i.c7.

Black simply keeps his bishop on c7

or b 8 . 2 2 tb h 4 + @g5 23 t2Jxf5 i. e l

Diagram 1 O .2 24 g4 i.c3 25 @d5 @f4 26 @c4

White to move i.f6 (Serper) is also a draw as the

black king can't be driven back.

 Endgame Secrets 33

22 ctJe5+ i.xe5 wing". I was rather surprised when

23 @xe5 @g5 only one day ( ! ) after my studies I

24 h4+ @g4 reached the present position in a

25 h5 1-0 toumament game. Unfortunately my

freshly acquired knowledge didn 't

Endgame 1 1 prevent me from committing

C.Lutz-G.Ginsburg mistakes.

Gennany 1 9 9 8 In comparison to position 10.1,

Unpublished. White's king is less well placed. In

addition, the position o f the h-pawns

is different.

47 h5

To have any winning chances at

ali, White has to prevent .. . h7-h5.

I) I quickly dismissed 4 7 @e2 as

the �h4 is fixed on a dark square.

Contrary to endgame 1 O, White

can't force through either g2-g4 or

f4-f5, e . g. 47 . .. h5 48 étJf6+ @g7 49

ctJe4 i.a5 50 @d3 i.e 1 51 g3 @f8

52 @c4 @e7 53 @d5 @d7 and

White can't go on (54 f5 gxf5 55

Diagram 1 1 . 1
ctJf6+ @e7 56 ctJxh5 f4 ) .
White to move
II) 47 ll)f6+ @g7 48 g4 isn't suc­

cessful either, since after 48 .. . i.e7

Analyse the game.
49 g5 i.a3 (not 49 . .. h6? 50 gxh6+

47 h5 i.d2 48 g3 gxh5 49 ctJf6+ @xh6 50 ctJg8+) Black manages to

@g7 50 @e2 i.cl 51 l2Jxh5+ @g6 exchange pawns with .. . h7-h6.

52 g4 i.b2 53 @f3 i.c3 54 @e4 h6 Black no longer has any problems:

55 @d5 f5 56 exf6 @ti 57 ctJg3 50 @e2 h6 51 @d3 hxg5 52 hxg5

i.xf6 58 ctJf5 @g6 59 @e6 i.b2 60 @f8, 50 ctJg4 i.cl or 50 ll)e8+ @f8

Cjje7+ @g7 61 f5 i.c3 62 ctJd5 i.b2 5 1 ll)c7 i.cl' 52 ll)d5 �g7.

63 ctJf4 i.c3 64 ctJh5+ @g8 65 f6 47 i.d2?

i.b4 66 ti+ @h7 1-0 Black goes for a trap which, how­

ever, White avoids. As a conse­

Commentary quence, Black loses an essential

tempo.

Sorne time ago I studied end­ I) If Black stays passive, White

games of the type "ctJ v i. " . In Win­ pushes his pawn to h6, cramping

ning Endgame Technique I carne Black to the utmost, e.g. 47 . .. i.c3

across two chapters of interest. One 48 h6 i.b4 49 ll)f6+ @h8 50 g4

of them deals with the configuration i.f8 51 g5 i.e7 52 étJd5 i.d8 53

"ctJ + a� + c� v i.", the other with @e2 @g8 54 @d3 @f8 5 5 @c4 @e8

the type "ctJ + 4� v i. + 3� on one [ 1 1. 2 ]

3 4 Endgame Secrets

This position occurs in the game

as well. White has different tries:

A) 56 f5 .i.b2 doesn't help at all.

B) 56 �e4 @g6 57 f7! ? �xf7 5 8

@f5. If the white king gets to f7 or

g6, the game is won. But his black

counterpart can't be driven away.

The bishop pressurizes the �f4, e . g .

58 . . . .id2 59 'Dg3 .ie3 60 éZJe4 .i c l

61 éZJd6+ �e7 62 'Dc4 @f7 63

éZJe5+ �g7 64 éZJd3 .id2 65 @e6 (if

Diagram 1 1 . 2
White had time for f4-f5, é2Je5,
White to move
f5-f6+ he would win) 65 . . . h 5 ! 66 g5
56 'Df6+ .ixf6 57 gxf6 @d8 58 (66 gxh5 @h6 67 f5 .ig5 is imme­
@c5 @d7 59 �d5 @d8 60 @d6
diately drawn as the f-pawn won't
@e8 61 @c7 @f8 62 @d7 @g8 63
cross the square f6) 66 . . . h4 67 @f5
@e 7.
h3 68 'Df2 h2 69 @g4 .te3 70 éZJhl
11) But there is a way to reach the
.ic 1 and White is unable to win the
draw, namely 47 . . . gxh5 48 éZJf6+
black pawn without giving up his
�g7 49 éZJxh5+ @g6 50 g4 (50 éZJf6
own .
.id2 5 1 g3 h5 52 @e2 .ic 1 53 @f3
C) 56 é2J g 3 ! ? (exploiting the fact
' .i d 2 is ineffective as Black has too
that the black king can 't take on f6
much space) 50 . . . f5! (the freeing
due to éZJe4+) 56 . . . .id2 (the sim­
move) 51 exf6 (With his king still
plest, though 56 . . . .ixf6 is a draw as
on f1 White has no other choice)
we will see later) 57 @e5 (57 f5
5 1 . . . h6. Black draws in spite of two

minus pawns. He simply keeps his .ic3 58 éZJe4 .ib2 and it doesn't

bishop along the diagonal al-h8, help that the knight has reached e4)

e.g. 52 @e2 .ic3 53 @d3 .ib2 54 57 . . . .ic3+ 58 @f5 .ixf6 59 'De4

@e4 .ic3 55 @d5 @f7 [ 1 1. 3 ]. .ib2 60 é2Jd6+ �e7 6 1 l2Jc4 .icl 62

éZJe5 .id2 63 é2Jd3 @f7 and once

again there is no possible progress.

Position 1 1 . 3 is Black's main defen­

sive position, we will encounter it

time and again.

48 g3 gxhS

Now it is too late for this move.

But I don't see a defence anymore.

Interesting is 48 . . . f5 [ 1 1 .4 ]:

Diagram 1 1 . 3

White to move
 Endgame Secrets 35

49 0Jf6+ @g7

50 @e2!

This intermediate move side-steps

Black's trap. After 50 0Jxh5+? @g6

51 g4 i.xf4 52 0Jxf4 @g5 Black

can eliminate both white pawns.

50 ... i.cl

51 ét:Jxh5+ @g6

Black has to choose whether to

put his king to g6 o r t o f8 . From g6

the king can support the idea

. . . f7-f5. 5 1... @ f8 is rather passive.

Diagram 1 1 .4
White might win like this: 52 @d3
White to move
i.b2 53 @e4 i.c3 54 �d5 i.b4 5 5

ét:Jf6 [ 1 1. 5 ]
Black prevents h5-h6, but this is

still not enough.

I) 49 exf6? (spoils the position)

49 . . . @f7 50 hxg6+ (after 50 h6 i.c3

51 @e2 i.xf6 52 lt:Jxf6 @xf6 the

pawn ending is drawn, e.g. 53 @f3

@f5 54 g4+ @f6 55 @e4 @e6 56

@d4 @f6 57 @d5 @f7 5 8 @d6 @f6

59 @d7 @f7 60 g5 @f8 61 @e6

@g8 62 @e7 @h8) 50 . . . @xg6 51

@e2 ( 5 1 ét:Jf2 i.b4 52 ttJe4 @f5 53

lt:Jg5 h6 doesn't save the M6 either)

5 1... i. a 5 52 @f3 h5. Black wins the

M6, with a draw.

Diagram 1 1 . 5
II) 49 ét:Jf6+ wins: 49 . . . @g7
White to move
(49 . . . @f7 50 h6 g5 51 lt:Jxh7 gxf4

52 ctJg5+ @g6 53 h7 @g7 54 e6 Now Black only has the choice

i.b4 55 gxf4 and White wins) 50 between different evils:

h6+! (without this trick the win I) 55 . . . i. e l 56 ét:Jxh7+ @g7 57

wouldn't be easy) 50 . . . @xh6 ét:Jf6 i.xg3 58 ét:Jh5+.

( 50 . . . @h8 51 e6 followed by the II) 55 . . . @g7 56 ©c6 i.el 57

king marching to d7) 5 1 e6 i.b4 52 ét:Jh5+ @f8 (57 . . . @g6 5 8 g4 leads to

ét:Jd5 ( an important gain of time as positions we will encounter later) 58

52 e7? i.xe7 53 lt:Jg8+ @h5 54 @d7 i.c3 59 0Jf6 (59 e6? fxe6 60

lt:Jxe7 @g4 55 @g2 g5 5 6 ét:Jd5 gxf4 @xe6 is only drawn, as B l a c k ' s h­

and . . . h7-h5-h4 is only drawn) pawn is still on h7. White needs the

52 . . . i.d6 5 3 e7 i.xe7 54 0Jxe7 @h5 square g6 for his king, e . g . 60 . . . i.b2

55 @g2 @g4 56 ét:Jd5 and the extra 6 1 f5 i.c3 62 f6 i.b2 63 f7 i.a3 64

piece decides. @f6 i.b2+ 65 @e6 i.a3 66 0Jf6

3 6 Endgame Secrets

@g7 67 l'.bd7 �b4) 59 .. . @g7 60 much more dangerous) 59 .. . h5 60 f7

@e8 i..b2 (60 .. . � e l 61 l'.bh5+ @g6 �a3+ 6 1 @d7 h4 62 l'.be4+ @g6 63

62 g4 and 63 f5+) 6 1 l'.be4 and l'.bd6 e6 g3 64 e7 i..xe7 65 @xe7 g2 66

wms. f8='iV g l ='i' 67 'iVg8+, winning.

III) 5 5 .. . h6 (now g6 is weakened, III) 54 .. . �b2 55 @d5 �a3 (Black

therefore White prepares e5-e6) 56 seizes control of the d6 square.

g4 �d2 57 f5 @e7 5 8 l'.be4 i..b4 59 Therefore the white king has to go

f6+ @f8 (59 .. . @d7 60 l'.bd6) 60 e6 via c6) 56 �c6 f5 [ 1 1. 6 ]

fxe6+ 6 1 @xe6 �a3 62 f7 i..b2 63

l'.bg3 �c3 64 l'.bf5 i..d2 65 �f6 with

a win.

52 g4 i..b2

Black is doomed to passivity.

I) 52 .. . i..xf4 53 l'.bxf4+ @g5 54

@f.3 isn 't possible.

II) 52 .. . f5 53 @f.3! (thanks to

47 .. . i..d2? this move is possible.

White creates two passed pawns. In­

stead, 53 exf6? h6 leads to position

1 1.3 ) 5 3 . .. �d2 54 l'.bg3 fxg4+ 55

Wxg4 �c3 56 f5+ @f7 57 @f4

i..d2+ 58 @e4 �c3 59 lbh5 (the Diagram 1 1. 6

knight blocks the �h7 and controls White to move

f6) 59 .. . @e7 60 @d5 i..b2 61 e6

i..c3 62 f6+ and Black 1s N ow White can win in two differ­

defenceless. ent ways:

53 .@f3 i..c3 A) 57 ctJg3 fxg4 58 f5+ @g5

54 @e4 ( 5 8 .. . @f7 59 @d7 and ctJh5, e5-e6)

The white king is ready to win 59 f6 h5 (59 .. . @f4 60 e6, 59 .. . i..b2

�f7 by @d5-d6(c6)-d7-e8 and 60 f7 i..a3 61 e 6 @f6 62 @d7) 60

f4-f5+. 54 f5+ @h6 would have l'.bxh5 @f5 (60 . . . @xh5 61 e6) 61

been pointless. @d5 @g6 62 l'.bg3 (62 e6? g3)

54 h6 62 .. . @f7 (62 .. . � c l 63 @e6) 63 @e4

Now the black king has sorne @e6 64 ibhS i.. c l 65 l'.bf4+ @f7 66

more breathing space, but g6 is @f5.

weak. On the other hand, there was B) 57 exf6 (Black now lacks a

no real choice: move to reach position 1 1. 3 )

I) 54 f5+ 5 5 @f.3 and l'.bg3. 57 .. . i..b2 (57 . .. h6 58 @d5 @f7 59

II) 54 i.. a l 55 @d5 �b2 56 @d6 @e5 and @f5) 5 8 @d5 @f7 59 @e4

f5 ( otherwise @e7-e8) 57 l'.bg3 fxg4 @g6 60 f5+ @f7 6 1 g 5 ! (White uses

58 f5+ @g5 ( 5 8 . .. @f7 5 9 lbh5 fol­ the fact that the pawn is still on h7)

lowed by e5-e6 and f5-f6) 59 f6 6 1... i.. cl 62 l'.bf4 i..b2 63 @f3 h6

(though both sides have passed (otherwise White plays @g4, l'.bd5,

pawns now, the white ones are @h5, g5-g6) 64 g6+ @xf6 65 @g4
 Endgame Secrets 37

i.d4 66 lt:Jd5+ @g7 67 @h5 and 68 II) 56 lt:Jg3 ! fxg4 57 f5+ (the most

f6+. elegant, though 57 �e6 wins as

55 @d5 well) 57 . . . @f7 5 8 e6+ @e7 59 é2Jh5

White still intends playing i..b2 60 f6+ i.xf6 61 lt:Jxf6 g3

@c6-d7-e8. Since . . . h7-h6 already (61...@xf6 62 @d6 g3 63 e7 g2 64

happened, he can also choose a plan e8=¡v gl=¡v 65 ¡vf8+) 62 lt:Jg8+

based on f4-f5+: 5 5 @f3 @f8 63 . e7+ @f7 64 @d6 g2 65

I) 5 5 . . . i. d 4 56 é2Jf6 i.b2 57 f5+ lt:Jf6! [ 1 1. 7 ]

@g5 (57 . . . @g7 58 @e4 i.c3 59

lt:Jh5+ and White forces e5-e6, see

diagram 11.5, line III) 58 é2Je4+

@h4 59 e6 fxe6 60 fxe6 i.a3 61

@f4 @h3 62 @f5 i..e7 63 é2Jf6 @g3

64 @g6 i.d8 65 @f7 .@f4 66 e7

i.xe7 67 @xe7 @g5 68 @f7 and

White wins.

II) 5 5 . . . i.b4 (Black sets a trap) 5 6

lt:Jf6 i.a3 (56 . . . i..e7 57 f5+ @g5 58

lt:Je4+ @h4 59 é2Jd6 f6 60 lt:Jc8) 57

f5+ @g5 58 é2Je4+ @h4 59 lt:Jd6

(but not 59 e6? fxe6 60 fxe6 i.e7

61 @f4 h5 62 g5 i.xg5+ 63 lt:Jxg5 Diagram 1 1 . 7

stalemate!) 59 . . . f6 60 lt:Jc4, winning. Black to move

55 f5

Black forces proceedings. After 65 . . . gl=¡v 66 e8='iV+ @xf6 67

5 5 . . . i.b4 White has the pleasant ¡vf8+, winning the enemy queen.

choice between 56 @c6 f5 57 lt:Jg3 56 @ti

fxg4 57 f5+ (which is similar to dia­ By now we have reached position

gram 1 1 . 6 , line I with the h-pawn on 1 1 .3 .

h6) or 56 @e4 and 57 @ f3 . 57 lt:Jg3 i.xf6? !

56 exf6? 57 . . . i.d2 leads to a draw as was

This lousy move spoils every­ shown above. The text complicates

thing. As we already know, Black things somewhat.

can draw now. It requires only a lit­ 58 é2Jf5 @g6?

tle bit of calculation to find the win: But this is the decisive mistake as

I) 56 gxf5+? (leads to a draw the black king cedes ground. 5 8 . . . h5

only) 56 . . . @xh5 57 @e6 @g4 5 8 f6 is the only move. As 59 g5 i..xg5 60

h5 59 f7 (59 @f7? @f5) 59 . . . i.b4 fxg5 @g6 and 59 lt:Jd6+ @g6 60 f5+

60 @f6 h4 61 e6 h3 62 e7 i.xe7+ @g5 lead to an immediate draw

63 @xe7 h2 64 f8=¡v h l = ¡v and the White has to take the pawn, but af­

queen ending is a draw. ter 59 gxh5 i..b2 [ 1 1. 8 ]

38 Endgame Secrets

Diagram 1 1 . 8 Diagram 1 1 . 9

White to move White to move

. . . we reach a remarkable position. 69 . .. i.a3 (bad is 69 . . . i.f6? 70

Averbakh analysed in 1958 that @f5 and 'Dg4, @e6, f4-f5-f6 or

White wins if his king gets ahead of 69 .. . i.d8? 70 ébd5 i.a5 71 f5 @h6

the pawns ( e.g. to g5, with the 72 f6 i.d2 73 t2:Je7 i.b4 74 t2:Jf5+

knight on g4). But according to @h7 75 f7 followed by the king

Beliavsky/Mikhalchishin and Ion marching to e8) 70 t2:Jd5 @h6 7 1 f5

Tisdall (in Batsford Chess Endings) i.b2 72 f6 i.d4 with a draw as

White can't force the win with his White sticks onto the �h5.

king on d 5 . Sorne exemplary lines: 11) 62 ébg4+ @e6 63 h6 i.b2 64

60 @e4 (W e now reach the game f5+ @f7 65 @d5 (65 @f3 i. c l 66

H.Olafsson-V.Ivanchuk, Reykjavik @g3 i.g5) 65 . . . i.c3 (65 . . . i. a l ? has

1990 with colours reversed) been played by Olafsson in an

60 . . . @ f6 6 1 ctJe3 i.c 1 and now: analogous situation. The bishop

1) 62 t2:Jd5+ is analysed in detail can 't switch to the diagonal c 1-h6 in

by Tisdall: 62 . . . @ g 7 ! ? (62 . . . @e6 is time: 66 ébe5+ @f6 67 @e4 .1i.b2 68

possible as well. The point is the t2:Jg6 with a win) 66 ébe5+ @f6 67

drawing position after 63 f5+ @fl h7 @g7 68 f6+ @xh7 69 @e6 i.b4

64 ctJf4 @f6 65 h6 i.d2 66 ctJh5+ with a draw.

@f7 67 h7 i.c3 68 @f4 i.d2+ 69 59 @e6

@e4 i.c3 70 f6 i.xf6 71 t2:Jxf6 Now White is back in the driving

@g7) 63 @f5 @h6 64 @g4 i.a3 65 seat.

t2:Je3 i.c 1 66 ctJf5+ @h7 67 @g5 59 ... i.b2

( the white king has only temporary 60 ébe7+ @g7

access to this square) 67 . . . i.a3 68 61 f5 i.c3

ébe3 i.e7+ 69 @g4 [ 1 1. 9 ]

 Endgame Secrets 39

6 1... i. d 4 62 ét:Jd5 h5 63 g5 h4 64 Endgame 12

éZJf4 i.e3 65 f6+. A.Karpov-P.Leko

62 éZJdS i.b2 Dortmund 1 9 9 4

63 ét:Jf4 i.c3 Earlier publication:

64 'tJhS+ @g8 lnformant 61/426 (A.Karpov).

65 f6

65 @e7 is possible as well.

65 ... i.b4

65 . . . i.b2 66 @e7 i.a3+ 67 @e8

@h7 68 ét:Jf4 i.b4 69 f7 @g7 70

ét:Jd5 i.f8 7 1 ét:Je7.

66 ti+ @h7

[ 1 1.1 0 ]

Diagram 1 2 . 1

White to move

White enjoys a space advantage

and the black pieces are tied to the

defence of the queenside. How did

Karpov (to move) squeeze out a

win?

Diagram 1 1 . 1 0 Commentary

White to m ove

39 g4!

Black lost on time while complet­ The black pieces are busy on the

ing this move. But Black is lost any­ queenside, therefore White opens up

way as the white knight will reach another front on the kingside. The

f5: 67 @f6 i.a3 68 éZJg3 i.b4 move g2-g4 is a standard manoeu­

(68 . . . i.b2+ 69 @e7 i.a3+ 70 @e8 vre to attack the pawn chain

@g6 7 1 ét:Jf5 @f6 72 ét:Jxh6) 69 éZJf5 f7-g6-h5. Black can react in two dif­

i.f8 70 ét:Je7 i.g7+ 7 1 @ e 6 i. f8 72 ferent ways. Either he takes on g4,

@d7. enabling White to create a danger­

ous passed pawn with f3xg4 and

h4-h5, or he allows a capture on h5

and remains with a weak pawn

there. In any case, Black is in a

sorry state.

39 ffi

40 gxh5 gxh5
40 Endgame Secrets

41 @f4 White is tw
�c7 is . o pawns up b
42 @f5 't.Jd7 b en pnse whil ' ut the
43 ..te2 't.Jce5 e taken into ac 1 e . . . ..tf5 has to

should Wh 't c�unt as well H

1 e contmue ? . ow
43 f4 't.Jf3 44 @
gives Black :6 étJf8+ 45 @xf6
arpov). orne counterpl Commentary
43
('LJKxh4 ay

h !f Black succeeds .
t mgs aren 't I m taking on 7
44 � xd4 ! d4 . e ear an e
masi comes u y�ore. But Al�
Otherwise th
· p1ece
dangerous: 44 f: /tawn will be sacrifice P with a .

(Karpov). 3 . 45 ..txh5 'LJ

46 b4!
44 c4
The bisho
al-h8 · p gets to the ¿·ragon�
45 °f4 't.Jxc5

46 ..txh5 't.Jed7
46
47 ..tes 't.Jxa4
46 .. . ..t�-�7 47 ..tb2 ..tf5
48 h5 't.Jac5 47 ..tb ! + and 't.Jf6+.
2
49 h6 a4 48 @el ..txe4+

5
º @xf6 't.Jf8 . Black returns t . .llf5
his king to 8 he prece to b .
5 0 ..txa4 @d6 @ b e . After 48 nng
. . . 't.Jb7-d6 (K 51 ..tb5 @e7 d x 2 ..tf5 50 � .. . .ixb2+ 49
50... arpov). an
@f7 52 @c5 [ 13 .2t @g7 51 @d4

51 @e a3
7
52 't.Jc2 a2
..tf7 ! 't.Jce6
53
54 �xf8 't.Jd4
1-0

Endgame 1 3

V- Z.Almasi-C.Lut

ienna open tournament

z 199 6

Unpublished.

Diagram 1 3 2

Black to mo�e

· · · the white k.
lowed by the mg gets to b8, fol-

b-pawn. advance of the

49 ..txe5+ @g8

50 @b2 @ti

51 @b3 .id7

Diagram 13 1 52 @c4 @e6

White to mo�e 53 i..g3 ..tb5+

 Endgame Secrets 41

54 @c5 @d7 Endgame 1 4

55 @b6 @c8 J.U.Maiwald-K.Bischoff

Black manages to block the c- German championship 1997

pawn, but now the white extra Unpublished.

pawns decide.

56 �d6 �d3

57 a4 �e2

58 b5 axb5

59 a 5!

59 axb5? �f3 is only a draw.

59 ild3

60 a6 ile2

61 a7 1-0

White will capture the �b5, bring

his king to f6 and exchange the �a7

against the �g6. Though Almasi's

continuation is certainly the most

Diagram 1 4 . 1
elegant, there is a similar idea which
White to move
deserves attention, viz. 46 �d2 ! ?

�f5 4 7 ilc3 �xc3 (4 7 . . . i..xe4+ 48

Black enjoys an extra pawn and
@el �f5 49 �xe5+ loses in the
the pair of bishops. Things look
same way as the game). Now White
grim for White, don't they?
should choose 4 8 c8=¡v+ �xc8 49

ét:Jxc3 @g7 50 @c2 @f7 5 1 _@d3 Commentary

�e6 52 @e4 with a slow win. The

White played 88 lbe8? and re­
piece sacrifice 48 bxc3 �xe4+ 49
signed after 8 8 . . . �c5+ 89 @el @e4
�b2 �f5 is less convincing, as the
as there is no defence against
white pawns are devalued.
. . . @f3-g2. But he could have saved

the position:

88 ét:Jxd5!

White establishes a fortress.

88 �xd5

8 8 . . . �c5+ 89 ét:Je3+ �e4 90 �h6

doesn't lead anywhere.

89 �d4

The bishop occupies the diagonal

a7-f2, the king goes to g l. Things

could proceed as follows:

89 @e4

90 �a7 �d3

91 @gl �e2

92 �b6 �d2

93 �a7 �e3+
42 Endgame Secrets

What e l s e ? 13 .. JiabS 14 ltJd5 'iVxd2+ 15

94 i..xe3 @xe3 @xd2 i..xd5 1 6 cxd5 @f8 1 7 i..e3

95 @fl @f3 l:ta8 1 8 g4 ctJd7 1 9 l:txc8+ l:txc8 20

96 @gl i..c6 l:tcl thcl 2 1 @xcl a6 22 @c2 @e8

97 @fl! [ 1 4 . 2 ] 23 g5 ltJc5 24 b4 ltJa4 25 b5 ltJc5

26 bxa6 bxa6 27 h4 @d8 28 i..fl

@c7 29 i.h3 ltJd7 30 i..fl ltJc5 3 1

i..h3 t2Jd7 3 2 f4 h6 33 i..fl hxg5 34

hxg5 @b7 35 i..e2 i..al 36 @bl

i..c3 37 �c2 i..a5 38 i..g4 i.b6 39

i..xb6 4tJxb6 40 @b3 ltJa8 4 1 i.d7

ltJc7 42 f5 @b6 43 @c4 @b7 44 a4

@b6 45 i..c6 @a7 46 @c3 @b6 47

@b4 a5+ 48 @c4 @a6 49 @d4 �b6

50 i..d7 @b7 51 fxg6 fxg6 52 e5

dxe5+ 53 @xe5 @b6 54 i..e6 1-0

Commentary
Diagram 1 4 . 2

Black to move The diagram shows a well-known

Maroczy Bind position. Black's

Black can't break the white for­
main idea is the thrust . . . b7-b5 to
, tress, it's a draw!
attack on the queenside Por the mo­

ment, White confines himself to the

Endgame 1 5
prevention of this plan. Contrary to
C .Lutz-P .Schlosser
other lines of the Sicilian it is ad­
Dresden open tournament 1 9 9 5
vantageous for White to exchange
Earlier publication:
queens, as then his advantage in
Informant 64/170.
space gains in importance while

Black is lacking the usual counter­

play along the e-file.

13 :ab8!?

A novelty which put me ó f f a lit­

tle bit as I expected 1 3 . . . a6. My

knowledge then consisted only of

the classic game A.Karpov­

L.Kavalek (Nice Olympiad): 14

ltJa4 'iWxd2+ 15 @xd2 :c6 (Later

on Black tried to improve. The cur­

rent continuation is 1 5 . . . ltJd7, e.g.

1 6 g4 f6 1 7 i..e3 f5 1 8 exf5 gxf5 1 9

h3 �f8 20 f4 d5 2 1 cxd5 i..xd5 22

Diagram 1 5 . 1
l:thd 1 with a small edge for White in
Black to move
A.Beliavsky-S.Tiviakov, Groningen
Analyse the game. 1 9 9 3 ) 1 6 ltJc3 l:tac8 1 7 t2Jd5 @f8 1 8
 Endgame Secrets 43

..te3 é2Jd7 19 h4 ..txd5 20 exd5

:6c7 2 1 h5 @g8 22 f4 é2Jc5 23 ..tg4

é2Je4+ 24 @d3 f5 25 ..tf3 b5 26 g4

bxc4+ 27 l:.xc4 :xc4 2 8 bxc4 é2Jc5+

29 ..txc5 l:.xc5 3 0 h6 ..tf8 31 @c3

fxg4 32 ..txg4 @fl 33 ..te6+ @f6

34 ..tg8 l:.c7 35 ..txh7 e6 36 ..tg8

exd5 37 h7 ..tg7 3 8 ..txd5 ..th8 39

@d3 @f5 40 @e3 :e7+ 4 1 @f3 a5

42 a4 :c7 43 ..te4+ @f6 44 :h6

:g7 45 @g4 1 - 0 . Karpov comments

on this game in several sources. One

Diagram 1 5 .2
of his remarks needs closer inspec­

tion. After 14 ltJa4 he writes (my Black to move

translation): "Usually one plays 14

é2Jd5 �xd2+ 15 @xd2 ..txd5 16 15 ..txd5?

cxd5 with a space advantage for This move is a clear mistake as

White. But this position is fairly dif­ the advantage in space and the pair

ficul t to win, especially if Black of bishops guarantee a salid edge

plays cautiously and refrains from for White. 1 5 .. . 'Dxd5 leads to equal­

'active' continuations (like ity. White can recapture both ways:

1 6 .. . b 5 ? ? ) . The move 1 6 .. . b5 is bad, 1) 1 6 exd5 ..td7 1 7 ..txe7 (after 1 7

since after the exchange of the :hdl @f8 there is no white advan­

heavy pieces the ending "2..t v ..t + tage. Black has different possibili­

0J" offers good winning chances to ties then, e.g . .. . a7-a6 and .. . b7-b5 or

White on account of the black .. . a7-a5, .. . ..td4-c5 and .. . e7-e6)

weaknesses at the queenside." I 1 7 .. . ..th6+ [ 1 5 . 3 ]

would like to add: The game Lutz­

Schlosser shows that the endgame

after 14 é2J d 5 �xd2+ 15 @xd2

..txd5 16 cxd5 offers winning

chances for White even if Black re­

frains from 1 6 .. . b5?. Therefore it

isn't surprising that Black players

usually play 1 5 .. . é2Jxd5! and

1 6 .. . ..td7. Then White can't capture

on e7 d u e t o .. . ..th6+, but otherwise

Black has no problems.

14 0J d 5 ! ?

N ow 1 4 é2Ja4 might be less effec­

tive as the b6 square isn 't weak­ Diagram 1 5 .3

ened. But the text shouldn 't be White to move

better.

14 �xd2+ 1 8 @c3 ( 1 8 @c2 :es or 1 8 @dl

15 @xd2 [ 1 5 . 2 ] ..txcl 19 @xcl b5 20 ..txd6 :b6

44 Endgame Secrets

and . . . bxc4, .. . .ri.b5 is better for (according to Speelman) he loses by

Black) 1 8 .. . �g7+ ( 1 8 .. . .ri.xcl 19 a series of minor inaccuracies: 21

.: x c1 .:es 20 .ri.xd6 is better for b3 .ri.d4 22 @fl @f8 23 @e2 @e8

White) 19 @b4? (19 @d2 with a 24 .ri.d3 e5 25 dxe6 .1i.xe6 26 .ri.e3

draw is more advisable) 1 9 .. . a5+ 20 .ri.c5 27 f4 f6 28 h3 @d7 29 .ri.d2

@xa5 (20 @a3 .:es 2 1 .ri.xd6 lhe2 .ri.d4 30 �c2 �f7 3 1 @d3 .ri.b2 32

22 .ri.xb8 .ri.b2 mate ! ) 2 0 .. Jia8+ 2 1 .ri.c3 .1i.xc3 33 @xc3 a5 34 a3 @c6

@b6 .ri.d4+ 22 c5 .1i.xc5+ 23 .:xc5 35 .ri.d3 g5 36 fxg5 fxg5 37 b4 @b6

1Ixc5 and Black has the upper hand. 38 @d4 h6 39 e5 dxe5+ 40 @xe5

11) 16 cxd5 .ri.d7. Once again it .ri.e8 4 1 .ri.e4 axb4 42 axb4 @c7 43

only troubles White if he tries to .ri.f3 @d7 44 @d5 @c7 45 @c5

force matters: 1 7 .:xc8+ ( 1 7 .1i.xe7? .ri.d7 46 .ri.e2 .ri.c6 47 g3 .ri.g2 48 h4

.ri.h6+) 1 7 . .. lixc8 1 8 .1i.xe7?! .ri.h6+ gxh4 49 gxh4 .ri.c6 50 .ri.xb5 .ri.f3

19 @el li c l + 20 .ri.dl .ri.b5 21 5 1 .ri.c4 .ri.g2 52 .ri.d5 .ri.fl 53 .ri.c6

�xd6 .ri.e 3 , and White plays with­ .ri.e2 54 b5 �d8 55 @d6 .ri.c4 56 b6

out the lihl. Of course, White can .ri.a6 57 @e6 .ri.e2 58 @f6 @c8 59

exchange all rooks by playing 18 @g6 h5 60 @g5 @d8 61 .ri.e8 .1i.a6

lle 1 , b u t then we reach a drawish 62 .ri.xh5 @e7 63 .ri.f3 @f7 64 h5

position. At a suitable moment, .ri.c8 65 h6 @g8 66 .ri.e4 .1i.a6 67

Black might break up the enemy @f6 .ri.c8 68 @e7 .1i.a6 69 @d6 @f7

pawns with . . . e7-e6. A related end­ 70 @c7 @f6 7 1 .ri. d 3 1 - 0 .

game is J.Speelman-M.Trauth

(EWG tournament Berlin 1980) 16 cxd5

[ 1 5 .4 ]: 16 e x d 5! ? analogous to Karpov-

Kavalek is interesting as well.

16 @f8

17 .ri.e3 lia8? !

Only a loss of time. After 1 7 .. . a6

18 .:xc8+ :xc8 19 .s'.c 1 .s'.xc 1 20

@xcl ét:Jd7 21 @c2! Black is a

tempo up compared to the game.

However, as Black has to stay pas­

sive this tempo is of no importance.

18 g4

White strives to gain space on the

kingside. He wants to set up the

pawn structure w¿jd5 + ¿je4 + ¿jf4 +

Diagram 1 5 .4
¿jg5 v b¿jd6 + ¿je7 + ¿jfl + ¿jg6. Ac­
White to move
cording to Hans Kmoch's terminol­

Speelman analyses this game in ogy in his classic Die Kunst der

Analysing the Endgame. White has Bauernführung this set-up is called

a small edge as Black has already quartgriff ( quart grip ) . The strength

played the weakening .. . b7-b5. of a quart grip demonstrates the fol­

Though Black still should draw lowing educational example, taken

 Endgame Secrets 45

from the Encyclopaedia of Endings 22 @e8

(Pawns), endgame 5 0 8 [ 1 5 . 5 ] . 23 gS [ 1 5 . 6 ]

Diagram 1 5 . 6
Diagram 1 5 . 5
Black to move
White to move

Seemingly, White is in trouble. 23 étJ c S ! ?

How can he defend against the Since . . . .tc3-b4 has been pre­

black king gobbling up his pawns? vented, Black tries to become active

But White can force a breakthrough: with this knight sortie. As a conse­

1 b5 @g4 (1...axb5 2 cxb5 and 3 a6) quence Black has to accept a weak­

2 c5! (the point) 2 . . . dxc5 (2 .. . @f5 3 ening of his pawn structure (though

c6 or 2 .. . axb5 3 c6 bxc6 4 a6 don't it isn't clear whether this is really

help either) 3 b6 cxb6 4 d6, and the that bad). Black has several alterna­

d-pawn promotes. White can also tives. Of special importance is the

invert the first two moves: 1 c5 @g4 defensive plan outlined under IV:

2 b5!. I) 23 .. . e6? (misplaced activity) 24

18 é2Jd7 dxe6 fxe6 25 f4. White prepares

D u e t o .. . .t b 2 White swaps off ali f4-f5, Black is in deep trouble.

rooks. II) 23 . . . b 5 ? 24 a4 bxa4 25 b4 is

19 �xc8+ lhc8 wrong, the same applies for

20 �el lhcl III) 23 .. . a5? 24 .td2 b6 25 .te3

21 @xcl a6 and a2-a3, b3-b4, @b3-a4.

After the exchange of the heavy IV) Black can go for a passive de­

pieces the white advantage is obvi­ fence with 23 .. . @d8, waiting for

ous. The only question is: Can White to show his intentions. Possi­

White win or can Black draw? bly Black just holds a draw: 24 b4

22 @c2! @c7 (24 . . . i..e5 ! ? 25 f4 �g7 intend­

lt is important to watch closely for ing . . . h7-h6 deserves attention) 25

your opponent's intentions. The text a4 @d8 26 a5 @c7 27 h4 @d8 28 f4

not only centralizes the king, but (White closes the diagonal cl-h6,

also prevents 22 .. . .t c 3 . Black can thus enabling Black to play . . . h7-h6.

then play . .. .tb4-c5. But White can't do without f3-f4 as

46 Endgame Secrets

a preparation for h4-h5, e.g. 28 @d3 (30 @d3!? is possible, as after

ét:Je5+ or 28 h5 gxh5 29 f4 h4 and 29 . . . ét:J f6 30 �f3 White can go for

while White has to <leal with the h­ e4-e5) 3 0 . . . gxh5 31 �xh5. After

pawn, Black will manage to play 3 1... f6 ? 32 �g4 and �f5 the white

. . . h7-h6 anyway) 28 . . . h6 (28 . . . @c7 king can penetra te with

is hopeless. After 29 h5 @d8 30 h6 @d3-e2-f3-g4-h5. If Black defends

�al 31 �g4 the LSh7 is an addi­ with . . . ét:Jf8, the white bishops get to

tional weakness. White wins by c8 or d 8 . But after 3 1.. . @ e 8 White

�h3, f4-f5xg6 and �e6-g8) and has no win.

now: 24 b4

A) 29 @d3 (the natural move) Altematives:

29 . . . hxg5 30 hxg5 �b2 (Black I) 24 h4 ! ? is feasible. After

avoids the exchange of the bishops) � fl - h 3 Black has to retreat his

31 �d4 (or 31 @c2 �g7 32 �g4 knight when White can proceed

@c7 33 �h3 �al 34 f5 @d8 35 along the lines above.

fxg6 fxg6 36 �e6 @e8 37 @d3 II) 24 a4? a5 allows a blockade of

�b2) 3 1... � cl 32 �e3 �b2 33 the queenside.

�d2 @c7 34 �c3 �el 35 �d2 24 'D a 4 ! ?

�b2 3 6 @c2 �d4 37 �c3 �e3 38 With 24 . . . 'Dd7 Black can trans-

�d2 �d4 and White won 't make pose to lines above.

progress. 25 b5 'Dc5

B) 29 g x h 6 ! ? [ 1 5 . 7 ] A forced retreat.

I) 25 . . . �c3 26 bxa6 bxa6 27

�xa6 'Dxa2 28 �d2 and the knight

is lost.

II) 25 . . . a5 26 b6 @d8 (26 .. . ét:Jc3

27 �a6) 27 �d2 .id4 28 �b5

'Dxb6 29 .i.xa5 and Black is com­

pletely tied down.

26 bxa6 bxa6

Let's compare the two set-ups

wLSa2 v b�a6 and WLSa5 + lSb4 v

bLSa6 + LSb7 (which could have

arisen in the event of a passive

black defence). In the first­

Diagram 1 5 . 7
mentioned case the weaknesses a6
Black to move
and f7 are far apart. In general it's

This move appears illogical as important for the defender to have

White relaxes his grip, but White his weaknesses close together. On

opens up lines on the kingside. This the other hand the exchange of the

idea has been put into effect in a b-pawns allows the black knight ac­

very similar position in the game cess to the square c 5 .

Dvoirys-Tiviakov, Podolsk 1993 2 6 . . . ét:Jxa6!? would be strategi­

(cf. endgame 16). 2 9 . .. �xh6 3 0 h5 cally better (the lSb7 is less

 Endgame Secrets 47

vulnerable than the �a6), but there VI) 34 . . . b6 35 @b 5 @b7 36 e5

is a concrete drawback: 27 i.b5+ @c7 3 7 a4 @b7 3 8 exd6 exd6 39 h4

(27 @b3 ctJc5+ and .. . @d8-c7 al­ @c7 4 0 @ a 6 .

lows Black to correct his mistake. VII) 34 .. . @b6 35 e5 @a6

27 i.xa6 bxa6 28 @b3 @d7 29 @a4 · ( 3 5 .. . @c7 3 6 @b5 @d7 3 7 @b6) 3 6

@c7 30 @a5 @b7 is of no use) e6 fxe6 3 7 dxe6 @b6 3 8 f5 gxf5 3 9

27 .. . @d8 (27 .. . @f8 28 i.xa6 bxa6 h4.

29 @b3, and the �a6 gets lost) 2 8 27 h4

i.b6+ ctJc7 (not a dream square for White intends h4 - h 5 . 27 i. x c 5 ?

the knight, but otherwise there will dxc5 28 i.xa6 i.e5, and Black cap­

follow 29 i.e8) 29 f4 (29 @d3 i.e5 tures the ¿J g 5 . White should part

and .. . i.e5-f4) 29 .. . i. d 4 ! ? ( a try to with the pair of bishops only if he

disturb White's play. After 29 . . . @c8 gets real advantages in retum.

3 0 i.a4 White can improve his posi­ 27 ... @d8

tion by @c2-d3 and h2-h4) 30 28 i.fl @c7

i.xd4 ctJxb5 3 1 i.b6+ @d7 32 @b3 29 i.h3 ctJd7

ctJc7 3 3 i.xc7 @xc7 34 @c4 ( 1 5 . 8 ]. Black prevents 30 i.g4 (as a

preparation of h4-h5) due to

30 .. . ctJ e 5 .

30 i.fl ( 1 5 . 9 ]

Diagram 1 5 . 8

Black to m o v e

Thanks to his space advantage

White wins, viz.: Diagram 1 5 . 9

I) 34 e5 3 5 fxe5 dxe5 3 6 @c5. B lack to move

11) 3 4 f5 3 5 exf5 gxf5 3 6 h4 .

III) 34 .. . e 6 35 @b5 @d7 A repetition of moves to gain time

( 3 5 .. . exd5 36 e x d 5 @d7 37 @b6 on the clock.

@c8 3 8 a4 @b8 3 9 a5 @c8 40 a6) 30 ... ctJcS

3 6 @b6 @c8 3 7 a4 @b8 3 8 a5 @c8 31 i.h3 ctJd7

3 9 h4 @b8 40 dxe6 fxe6 4 1 f5. 32 f4

IV) 34 f6 3 5 h4 @b6 36 f5. White seizes the e5 square. He

V) 3 4 @d7 3 5 @b5 @c7 3 6 e5. closes the diagonal c l -h6 and

48 Endgame Secrets

allows . . . h7-h6 as otherwise he can't

make progress.

32 h6

33 .ifl

33 gxh6!? .ixh6 34 .ig4 ttJf6 3 5

.if3 is possible, in analogy to

Dvoirys-Tiviakov. I have the im­

pression that then the pawn struc­

ture on the queenside might be

disadvantageous for White. Dvoirys

broke through in the centre with

e4-e5, using the �b7 as a target. But

Diagram 1 5 . 1 0
now White is lacking this target.
Black to move
33 hxg5

34 hxg5 @b7 40 ttJa8

35 .ie2 .ial Black redirects his knight to c7 to

36 @bl? protect the e8 square. But now both

White would like to include an­ his pieces are severely restricted and

other repetition of moves, but Black White can operate with zugzwang

can manoeuvre his bishop to the di­ motives. Black has two other

agonal e l - a 5 . 3 6 .if3 is better. options:

36 .ic3 I) 40 . . . @c7 (Black awaits White's

37 @c2 .ia5 intentions) 4 1 @b4 @b7 42 a4 @a7

After 37 . . . .ial ! ? White probably (42 . . . @c7 4 3 .ie 2 @ b 7 44 @ a 5

would have played 3 7 .i f3 . But the loses the �a6) 43 f 5 gxf5 44 .ix f5

text is better. @ b7 4 5 a5 ttJ a8 ( 45 . . . ttJ c8 4 6 .id?

38 .ig4 .ib6? and 47 .i e 8 ) 46 .i e6! and the g ­

The bishop uses its newly-gained pawn q uee n s.

freedom for a mistaken exchange. II) 40 . .. a

5 (closi n g the queensi d e )

On the one hand the white king 4 1 .i e 2! (41 @c3 @c 7 4 2 @d4 "in­

gains access to important dark tendin g f4 f xg6

- 5 and e4 e - 5" h as

squares ( d4, b4, a5), on the other been m y original comment in the

hand the knight is out of play. By Informant. This was ra ther careless

now White is winning. 3 8 . .. @c7 is as a ft er 4 2 ... @d 8 and . .. @ e 8 - f8 - g7

better, e . g . 39 .i d 4 ! ? .iel (39 . . . ttJc5 Bl ack de fen ds the kingside an d es ­

40 e5 and e5-e6) 40 @d3 .ih4 ! ? , tab li shes a · fortr ess. 4 1 .i e 2 ! intend ­

and the black bishop hinders in g 42 .i b 5 avoid s this de fe nce ) and

White's progress. now :

39 .ixb6 ttJxb6 A) 41 ... ttJ d7 42 .ib5 ttJ c5 +

40 @b3 [ 1 5 . 10 ) (42 . . . @ c7 43 .i xd7) 43 @ c4 ttJ xe4

 Endgame Secrets 49

44 i.e8 ltJg3 45 i.xt7 ltJe2 46 f5, White restricts the ltJc7 even fur­

White wins. ther. 44 i.e6? is too brilliant to be

B) 4 1... @ c 7 42 i.b5 @d8 43 @c3 true: 44 .. . fxe6 45 f6 exd5+ 46 exd5

with a further branch: e6 47 t7 exd5+ 48 @d4 (48 @d3

B l ) 43 f6 44 f5 just loses. ltJe6) 48 . . . @c6 (48 .. . ltJe6+? 49

B2) 43 ltJc8 44 f5 ltJb6 45 i.c6 @xd5 ltJf8 50 @xd6 and White

(zugzwang) 45 .. . e5 (or 45 . .. ltJc8 46 really wins ), and now White should

@c4 ltJa7 47 i.a4) 46 dxe6 fxe6 47 draw with 49 f8=éu, rather than 49

f6 e5 48 @b3. f8='iY? ltJe6+.

B3) 43 .. . e6 (Black avoids 44 ... @b6

zugzwang but he has to loosen his 45 i.c6 @a7

position) 44 dxe6 fxe6 45 @d4 @e7 Or 45 . .. @a5 46 @b3 (driving the

(45 .. . ltJd7 46 i.xd7 @xd7 47 @c4. king away) 46 .. . @b6 47 @b4 @a7

White captures the �a5 as 4 7 @c6 48 @a5 @b8 49 @b6 with a win.

48 f5 loses) 46 i.c6 ltJc8 (46 d5 4 7 46 @c3! [ 1 5 . 1 2 ]

@c5, 46 .. . @t7 47 e5 @e7 48 i.e4

dxe5+ 49 fxe5 @t7 50 i.c6 and

@ c 5) 47 e5 d5 48 i.b5 [ 1 5 . 1 1 ].

Diagram 1 5 . 1 2

Black to move

lt has become clear that White

Diagram 1 5 . 1 1 will win by means of@c4-d4, f5xg6

Black to m o v e and e4-e5. However, White strives

to implement this plan under the

The combined attack against g6 most favourable circumstances. The

and a5 wins: 48 . .. @d8 49 i.d3 ltJe7 triangulation @c4-c3-b4 forces

50 @c5 @c7 5 1 a4 @d7 52 @b5 d4 . .. a6-a5. Other moves are less

53 @c4. accurate:

41 i.d7 Ci:Jc7 I) After 46 @d4 @b6 47 fxg6

42 f5 @b6 fxg6 48 e5 @a5 the black king es­

42 .. . gxf5 43 i.c6+ and 44 exf5, capes. Why should White allow this

45 g6. if he can play 46 @c3 ?

43 @c4 @b7 11) 46 a 5 ? ! allows the Cuc7 to ac­

44 a4 cess the b5 square.

50 Endgame Secrets

46 ... <it>b6

47 <it>b4 a5+

47 . . . <it>a7 48 <it>a5.

48 <it>c4 @a6

49 <it>d4 <it>b6

50 Sl.d7 @b7

50 .. . <it>a6 51 fx g 6 fxg6 52 e5

doesn 't differ too much, e.g.

52 .. . @b7 53 exd6 exd6 54 Sl.e6

ltJxe6+ 5 5 dxe6 [ 1 5 . 1 3 ]

Diagram 1 5 . 1 4

Black to move

Black resigned as there is no de­

fence against 55 Sl.f'l and 56 i.xg6.

Endgame 1 6

S.Dvoirys-S. Tiviakov

Podolsk 1 9 9 3

Earlier publication:

ChessBaseMagazin 38 (I.Belov).
Diagram 1 5 . 1 3

Black to move

... with a won pawn ending:

55 . . . <it>c8 (55 . . . <it>c6 56 <it>e4 d5+ 57

@e5 d4 5 8 <it>f6) 56 <it>d5 @c7 5 7 e7

<it>d7 5 8 e8=¡f+ @xe8 59 <it>xd6.

51 fxg6 fxg6

52 e5 dxe5+

Or 52 . . . <it>b6 53 exd6 exd6 54

Sl.e6 ltJxe6+ 55 dxe6 with similar

lines as in position 1 5 . 1 3 .

53 @xe5 @b6

54 .ite6 [ 1 5 . 1 4 ] Diagram 1 6 . 1

White to move
 Endgame Secrets 51

White enjoys a space advantage 1) 3 8 i.xh5? is given by Belov. In

and the pair of bishops. His superi­ my opinion this exchange throws

ority is obvious. There followed: 3 5 away the win: 3 8 . . . gxh5

h4 luf6? Now White can break A) 39 @e4 dxe5 (39 . . . @c8? 40 e6

through in the centre. Better is is Belov's idea when White really

3 5 . . . @d8 36 h5 gxh5 3 7 i.xh5 @ e 8 , wins, e.g. 40 . .. fx e 6 4 1 dxe6 i.g7 42

c f. Lutz-Schlosser (endgame 1 5 ) . 3 6 i.d4 i.xd4 43 @xd4 @d8 44 @d5

e5 luh5 36 . . . lud7 37 e6 fxe6 38 @e8 45 f5 @f8 46 f6 exf6 4 7 @xd6

dxe6 luf8 39 i.d5 i.g7 40 i.d4 @e8 48 e7 f5 49 @e5 @xe7 50

with a win according Belov. White @xf5) 40 @xe5 (40 fxe5 e6 4 1 d6

creates a passed h-pawn by means @c8 and after .. . <it>c8-d7-c6 the

of f4-f5. 37 i.b6+ @d7 3 8 e 6 + @e8 black king blocks the queenside

39 exf7+ @xf7 40 i.g4 ét:Jxf4+ 4 1 while the bishop defends the king­

@e4 e6 42 i.e3 luxd5 43 i.xh6 side) 40 . . . i.g7+ 4 1 <it>f5 i.f6 42 d6

ét:Jxb4 44 i.d2 ét:Jc6 45 i.c3 d5+ 46 @c8 [ 1 6 . 3 ] and Black certainly isn't

@f4 e5+ 47 i.xe5 luxa5 48 i.c8 worse.

ét:Jc6 49 i.c7 lud4 50 i.b6 lue6+ 5 1

@g4 ét:Jf8 52 i.xb7 lud7 53 i.d4

luf6+ 54 @g5 lue4+ 55 @h6 1 - 0

The game continuation 3 7 . .. @ d 7

allowed 38 e6+, how will White

proceed after 37 . . . @b8 [ 1 6 . 2 ] ?

Diagram 1 6 . 3

White to move

B) 3 9 e x d 6 exd6 40 @e4 @c8 4 1

@f5 @d7 is no improvement as the

M4 and the ¿jb4 are weak: 42 i.f2

Diagram 1 6 . 2 ( 42 i.e3 @e7 43 i.d2 @e8 44 @f6

White to m o v e @f8 doesn't help White) 42 .. . @e7

43 i.g3 i.g7 44 @g5 i.c3, and the

Analyse the possibilities 3 8 f5 and queenside crumbles.

3 8 i.xh5. 11) 3 8 f5! is stronger. 3 8 . . . dxeS 39

fxg6 fxg6 (3 9 . . . luf4+ 40 @e4 ét:Jxg6

Commentary 41 h5 ét:Jh4 42 i.g4. The h-pawn

gets too strong and i.d8 is a further

After 3 7 . . . @b8 White has to take threat) 40 i.c5 i.f8 4 1 @e4 [ 1 6 .4 ]

care of the ¿jf4. and the king penetrates.

52 Endgame Secrets

Commentary

After 45 a5 46 llxb5 :xc6 47

llxa5 or 45 b4 46 llxa6 b3 47 llb6

and c6-c7 Black remains with a use­

less pawn on the kingside. 45 .. J�a3

46 llb8 �c3 4 7 llb6 is only a repeti­

tion. Nevertheless Black can win:

45 @g6!

In the endgame, the king is a

strong piece. Black goes for the

�h3!
Diagram 1 6 . 4
46 lixa6 @h5
Black to move
47 :b6

47 tla7 lixc6 4 8 l:.xg7 :g6+.

47 ... @h4

Endgame 1 7 48 lixb5 l:Ic2+

B.Tondivar-C.Lutz 49 �gl @xh3

Leeuwarden open toumament 1 9 9 4 Not 49 . . . l:Ixc6? 50 : b 3 . The �h3

Unpublished. is more important than the �c6, as

Black gets two connected passed

pawns.

50 :b7

50 �b3+ @h4 51 l:.b4+ @h5 52

l:.b6 g5 53 @fl g4 followed by

. .. �g5, .. . h6-h5, .. . @h4 etc.

50 g5

51 c7 g4

52 llb6 h5 [ 1 7 . 2 ]

0-1

Diagram 1 7 . 1

Black to move

Black to move is two pawns up,

but rook endings have a well-known

drawing margin How <loes he win ?

Diagram 1 7 . 2

White to move
 Endgame Secrets 53

Endgame 1 8 Commentary

C.Lutz-W.Uhlmann

Dresden zonal toumament 1 9 9 8 White intends to exchange his

Earlier publication: l2id3 for the l2ia4 to create a passed

Informant 72/293 a-pawn. To reach this goal White

could have played 45 @b 1 �d2 46

�c4 �xd3 47 �xa4, but after

47 .. . @c6 (with the idea 48 �a7 �f3)

Black's king is rather active. How­

ever, after 45 @e 1 Black is in

zugzwang and has to allow his posi­

tion to deteriorate. Either he has to

allow the enemy king to improve his

position (45 . .. �al+ 46 @b2 � d l ) or

to worsen his own king's positions

(45 .. . @d8, 45 . . . @e8 or 45 . .. @e7).

Let us take a look at both

possibilities:
Diagram 1 8 . 1
I) 45 .. . �al+? 46 @b2 �dl oc­
White to move
curred . in the game. After 47 �c4

White is a pawn up, but the black �xd3 48 �xa4 [ 1 8 . 3 ] White has put

rook is rather active. Thus a win is the pawn exchange into effect, with
questionable. The game continua­ his king on b2 instead of b 1 .
tion: 40 @d4 a4 4 1 �c3 �fl 42 @e3

42 �c4 �xf4+ 43 @c3 �f5 44 d4 f6

leads to the exchange of pawns and

a drawish position. 42 .. . �al 43 @d2

@d7 44 @c2 �a2+ 45 � e l [18.2].

Diagram 1 8 .3

Black to move

It seems to me that White is win­

ning. Black would like to resolve

, Diagram 1 8 . 2 the kingside with .. . f7-f6, but

Black to move against accurate play he won't

How should Black continue ? succeed:

54 Endgame Secrets

A) 48 f6 49 J:ra7+ and exf6.

B) 48 @e7 49 .a'.b4 .a'.f3 50 a4 f6

51 a5 fxe5 52 fxe5 followed by

.a'.b5( c4 ) - c 5 . The ¿je5 is securely

protected and the black king is cut

off. Black has no defence against

the advance of the a-pawn, e.g.

52 .. . @ d 7 53 .a'.c4 .a'.e3 54 .a'.c5 :.h3

55 a6 .a'.h8 56 @b3 .a'.a8 57 .a'.a5 .a'.a7

58 @b4 @c6 59 :.a2 @b6 60 :. a 1

@c6 6 1 @ a 5 .

C) 48 . . . @e8 is tricky. Black

Diagram 1 8 .4
avoids a check along the seventh
Black to move
rank. Now neither 49 :.b4 f6 50 a4

(50 exf6 @fl and . . . @xf6, . .. e6-e5)

There is no defence against 56
50 . . . fxe5 51 fxe5 @d7 nor 49 .a'.c4
@d4 followed by @c5-b6.
f6 50 :.es .a'.d4 is convincing. Thus:
II) A better defence is the with­
49 .a'.a8+ @e7 50 .a'.b8 .a'.f3 (50 . . . f6
drawal of the own king. 45 . . . @e8 is
51 .a'.b7+ @f8 5 2 exf6 .a'.f3 53 .a'.b4
the most accurate move. Black
@fl 54 a4 @xf6 55 a5 .a'.d3 56 a6
draws: 46 @bl .a'.d2 47 l:lc4 .a'.xd3
.a'.d7 57 .a'.a4 .a'.a7 58 .a'.a5 wins in a
48 .a'.xa4 f6 ( 1 8 . 5 ]
way similar to the game) 5 1 .a'.b4 f6

52 a4 fxe5 53 fxe5 @d7 54 .a'.c4 and

.a'. c 5 .

D) 48 @c6 49 l:la7 .a'.f3 50 .a'.xfl

@c5 (50 @b5 51 .a'.b7+ and .a'.b4)

51 a4 @b4 52 @c2 @xa4 53 .a'.f6

@b4 54 @d2 @c4 55 @e2 .a'.a3 56

.a'.xe6 @d5 57 f5 .a'.a4 (57 .. . @e4 58

f6) 5 8 @f3 and White is winning.

E) 48 . . . .a'.D has been played by

Uhlmann. 49 .a'.c4 f6 (49 . . . @e7 50

a4 f6 5 1 a5 fxe5 52 fxe5 and .a'.c5)

5 0 exf6 @e8 51 a4 @fl 52 a5 .a'.d3

53 a6 .a'.d7 54 .a'.a4 .a'.a7 55 @c3

Diagram 1 8 . 5
[ 1 8 .4 ] 1 - 0 .
White to move
 Endgame Secrets 55

49 exf6 (49 :c4 fxe5 50 fxe5 :b7+ @b8 47 a5 bxa5 48 bxa5

:d5 5 1 l:.e4 @d7 52 @c2 @c6 or 49 :a2 49 :b5 :a4+ 50 @e3 @e7 5 1

:a8+ @f7 50 :a7+ @g6 aren't bet­ d4 :al 52 @e4 @e6 53 :b6+ @b5

ter either) 49 . . . @f? 50 @b2 @xf6 54 ll.b6+ �xa5 55 1:1'.bS @a4 56 d5

and after .. . e6-e5 the black king is· a5 57 d6 :dl 58 @e5 @a3 59 :b5

rather close to the a-pawn. a4 60 ll.d5 :íel + 61 @d4 ll.dl + 62

�c4 :el+ 63 @b5 :bl+ 64 @a5

Endgame 1 9 :bs 65 :d3+ :b3 66 :d4 :bs 67

C.Lutz-T.Lutber :xa4+ @b3 68 :b4 :b7 69 :d4

German championship 1 9 9 8 �e3 70 :dl 1-0

Earlier publication:

SCHACH 12/98. Commentary

Diagram 19.1 shows a well­

known position of the French de­

fence. Black now transposed to an

endgame.

9 'Llxd4

10 i.xd4 i.xd4

11 ¡yxd4 ¡yb6

This line is unofficially called

"the vacuum cleaner line" as sud­

denly a large part of the white and

black pieces are "vacuumed" off the

board and the queens will come off,

Diagram 1 9 . 1
too. Thanks to the efforts of Viktor
Black to move
Korchnoi, Alexander Chemin, Mik­

Analyse the game. hail Gurevich and other grandmas­

ters this ending has been revived in

9 ... 'Llxd4 1O i.x d4 i.x d4 11 the mid 1980s. It tumed out that

¡yxd4 ¡yb6 1 2 ¡yxb6 lt:Jxb 6 13 'Llb5 Black not only has a very solid posi­

@e7 1 4 .0-0-0 i.d7 1 5 étJd4 'Ll a4 16 tion, but a l so that he can counter

i. d 3 'Llc5 17 :bfl g5 1 8 f5 étJxd3+ careless play by White with active

19 exd3 exf5 20 'Llxf5+ i.xf5 21 ideas. In this respect the thrust

:xf5 :bg8 22 :dfl .:íg7 23 @d2 .. . g 7- g5 is of special importance.

d4 24 :íel @e6 25 :r6+ �d5 26 b4 Nevertheless White should be able

b6 27 :en :es 28 l:txf7 :e7 29 to maintain a small but persistent

:f8 @xe5 30 :H5+ @d6 3 1 :dS+ edge due to his better bishop.

:d7 3 2 :xd7+ @xd 7 33 :[d5+ @e6 12 ¡yxb6

34 J:[xd4 :n 35 ll.g4 b6 36 a4 This is more precise than 1 2 . 0 - 0 - 0

¡¡,f2+ 37 @e3 @d5 38 b4 gxh 4 39 ¡yxd4 13 :xd4 as the black knight

ll. d 4+ @e6 40 :xb4 :xg2 41 is bad on b6. The only r easonable

:xh6+ @e7 42 :b7+ @b8 43 @e4 move to avoid an exchange of

a6 44 :b5 :e2+ 4 5 �d4 @b7 4 6 queens is 12 ¡y d 2 ! ? ( G. Ka s p a ro v -

56 Endgame Secrets

J.Timman, Horgen 1 9 9 5 ) , but there

have been hardly any followers of

this idea.

12 . .. lbxb6

13 ctJb5 g;;e7

14 0-0-0 ..td7

15 ctJd4 cba4

16 ..td3 ctJc5

[19.2]

Diagram 1 9 . 3

White to move

... and Black is winning

(J. Timman-A. Chemin, Montpellier

candidates toumament 1 9 8 5 ) .

II) 1 7 @d2 g5 ! ? 1 8 fxg5 .a'.ag8 1 9

h4 h6 20 ctJf3 hxg5 2 1 tbxg5 laxh4 !

22 .a'.xh4 �xg5 23 l:. e l (23 g4 �xe5

24 lah5 tbe4+! 25 Wc 1 cbg5 with

Diagram 1 9 .2 compensation according to

White to move M.Gurevich in Informant 49/(334))

23 . . . l:Ixg2+ [ 1 9 .4 ].
Both sides improve the position of

their knights.

17 l:ihfl

Two examples which show the

black resources:

I) 17 l:i d e l h5 18 g3 laag8 19

1:Ihfl g5 ! 20 f5 g4 (Black gets coun­

terplay on the kingside) 2 1 f6+ Wd8

22 lae3 h4 23 b4 hxg3 24 hxg3

cbxd3+ 25 cxd3 l:.h2 26 ctJe2?! (26

1:Ie2 with equal play according to

Chemin in Informant 40/(360))

26 . . . @c7 27 @d2? d4! 28 lae4 ..tc6

29 la c l @b8 3 0 .a'.xM ..tf3 31 l:. e l Diagram 1 9 .4

White to move
.a'.g5 32 l:id7 l:.xe5 33 l:Ixf7 ..txe2

34 .a'.f8+ @c7 3 5 f7 .a'.f2 ( 1 9 . 3 ]

D u e t o the weak l:le5 Black won't

lose. In L.Christiansen-M.Gurevich

(Reykjavik 1990) Black even won

later on.
 Endgame Secrets 57

17 g5?

This is the move 1 7 :hfl was de­

signed to meet. A better move is

1 7 . . . g6, e . g . 1 8 b3 �a�8 1 9 @b2 a6

20 �f3 h5 2 1 g3 l:lcg8 22 l:ldfl .i.e8

23 l:le3 g5 24 f5 g4 25 .i.e2 l:lg5?

(25 . . . .i.d7 26 c4!? with a slight

white advantage according to

M.Chiburdanidze in Informant

41/340)) 26 � c 3 ! b6 27 @ e l a5 28

a3 .i.d7 29 b4 [ 1 9 . 5 ].

Diagram 1 9 . 6

Black to move

18 fx g 5 ? ! :ag8 19 h4 h6 20 g4

(20 ctJf3 hxg5 21 hxg5 'Lle4)

20 . . . hxg5 21 h 5 leads to an unclear

situation. White can 't advance his

passed pawn and the t!:,e5 is

vulnerable.

18 'Llxd3+

Black seeks salvation in a bad

double rook ending. He is one

tempo short to create his typical

Diagram 1 9 . 5 counterplay on the kingside:

Black to move I) 1 8 . . . :ag8 19 f6+ @d8 20 h3!.

The point. The only way to stop

29 . . . axb4 30 axb4 'Lle4 31 :c7 White from blocking off the king­

:es 32 :a7 @e8 33 fxe6 fxe6 34 side with 2 1 g3 is playing 20 . . . g4 2 1

'Llxe6 :xe5 3 5 l:lf8+ @e7 3 6 l:lxc8 h4 g3, but after 22 :f4 Black only

@xe6 37 :bs and 1-0 in has a weak �g3.

M.Chiburdanidze-F.Vareille (Mont­ II) 1 8 . . . g4 · 1 9 f6+ @d8 20 l:lf4 h5

pellier 1986). M o r e .accurate, how­ 21 h4 and once again White suc­

ever, is 18 g3 h5 19 h3 as then ceeds, since after 2 1.. . g x h 3 22 gxh3

Black can no longer become active @c7 23 :g 1 the white rook gets to

on the kingside. White will counter g7.

an eventual . . . g7-g5 with f4-f5 and III) 1 8 .. . h5 1 9 f6+ @d8 20 h3 and

after . . . g5-g4 he can el ose the king­ neither 21...h4 22 .i.e2 nor

side by h3-h4. Therefore play will 20 . . . 'Llxd3+ 21 cxd3 h4 22 @d2

develop on the queenside where @c7 23 :c1 + are desirable for

White can make progress with @d2, Black.

b2-b3 and a2-a4-a5. 19 cxd3 exf5

1s rs [19.6] [19.7]
58 Endgame Secrets

ter than the text, since White doesn't

rush things.

23 d4

Forced. White seizes the file, but

the �e5 is weak and the @d2 is held

up.

24 I:.cl @e6 ! ?

This move offers White a variety

of choice, neither of which is

enough for a win, however.

24 . . . :ds is more natural: 25 :c7+

@e6 (25 .. . I:.d7? 26 I:.xd7+ @xd7 27

Diagram 1 9 . 7
:f6) 26 :f6+ @xe5 27 :fxfl :xn
White to move
28 I:.xfl b6 29 .:rxa7 (29 I:.xh7 :f8
The admission of a mistaken strat­ leads to the next note) 29 . .. :f8 30

egy. After the exchange on d3 Black @e2 :es 31 .:rxh7 :c2+ 32 @f3

can't stick to his intentions on the :xb2 and Black has counterplay.

kingside, because White can pene­ 25 :r6+ @d5

trate via the e-file. Forced as 25 .. . @xe5? 26 :cfl

20 tbxf5+ i.xf5 leads to mate.

21 :xf5 :hg8 26 b4

22 :dfl :g7 Now I invested a large part of my

Black has weak pawns on d5 and thinking time. I simply couldn't be­

f7 and on top of that his rook has to lieve that Black can survive despite

remain passively on g7. His position his obvious disadvantages ( exposed

is bad. king and passive rooks). The text

23 @d2!? move wins a pawn, but not the

White threatens @e3-d4 and al­ game. The other choices don't

most by force he wins a pawn. In promise anything better:

retrospect I'm not sure, however, 1) 26 :c7 @xe5 27 :fxfl :xn 28

whether this is enough to win. :xfl b6 (lt's this pawn that Black

Therefore 23 d4 deserves attention. must keep. 2 8 . .. h6? 29 :xb7 and

White protects the �e5 and threat­ :b5+ is hopeless) 29 :xh7 :f8 30

ens I:.f6-d6. After 23 . . . :cs+ 24 @d2 @e2 :es 31 :xa7 :c2+ 32 @f3

:c6 2 5 :f6 :e6 (otherwise :xc6 g4+ ( 3 2 .. . :xb2 33 h3!? together

and @c3-b4-c5) 26 @c3 White with 34 g3 and 35 @g4) 33 @g3

plans @b4-c5 while Black can 't be­ a'.xb2 [ 1 9 . 8 ].

come active. Today I like 23 d4 bet-

 Endgame Secrets 59

In vi e w of numerous white threats

(2 8 .:xf7, 28 .: lf s , 28 lta6+) this

m ove is forced.

28 .l:. xf7 .l:.e7

Bl ack controls the seventh rank,

b ut W hite will remain a pawn up.

29 ars @xe5

29 . . . .1:.xeS 3 0 .:as +.

30 .l:.lf5 + @d6

31 :d8+ :d7

3 1... @ e 6 3 2 lifd 5 .l:.a7 33 .: x a 4 .

32 :xd7+ @xd7
Diagram 1 9 . 8
S tronger than 3 2 . . Jha7 33 .l:. xg5
White to move
.:f7 34 @ e2 :c7 35 <it>f3 c2 3 6 a4
.l:.

.:b2 37 b 5 :b4 38 �e4 .l:. xa4 39

It isn't clear to me whether .l:.as+ @e 6 40 .l:.h5 and W hite

White's advantage is sufficient to should w in.

win. In retum for the t0ig4 White has 33 :d5+ @c6

to surrender the t0i a 3 . 34 .:xd4

II) 26 .l:.cfl .l:.e8 (in analogy to the Converting the extra p awn isn't

game this is the best defence) 27 easy as the .l:. g7 beco m es active.

.l:.xf7 (27 .l:.a6+ @xe5 28 .l:.a7 .l:.b8 P ossi b ly the king might hurry to a4.

doesn't yield anything) 27 . . . .l:.e7 28 34 . .:.f7

.l:.f8 @xe5 29 .l:.lf5+ @a6 3 0 .l:.a8+ 35 :g4

.:a? 3 1 .l:. x a 7 +.l:.xa7 ( 3 1... @ x a 7 3 2 35 @e3 @b5 .

.:as+ @c6 3 3 .l:.xa4. Contrary to the 35 h6

game the moves b2-b4 and . . . b7-b6 35 . .. .l:. f2 + ! ? 36 @c 3 .l:. xa2 37

are not inserted. This is to White's .l:.x g5 a5 should lead to a draw, too.

advantage as Black's king doesn't 36 a4 .:n+

get to a4) 3 2 .l:.xg5 .l:.f7 3 3 @e2 .l:.c7 37 @c3 @d5

34 @f3 :c2 with counterplay. 38 h4

III) 26 .l:. e l .:as 27 e6 fxe6 28 I nitiates furt her exchanges, but I

.:fxe6 . White threatens m ate, but don't s ee anything els e .

after 2 8 . . . .l:.a6 he is left with 38 gx h4

nothing. 39 .l:.d4+

26 b6 39 .l:.xh4 .:xg2 40 .l:. xh6 .l:. a2 41

27 .l:.cfl @ b3 .: a 2 .

27 .l:. el .l:.as 28 e6 fxe6 29 : fx e 6 39 @c6

.:a6 30 .l:. l e5+ @ c6 and o nce again 40 .l:. xh4 .l:. xg2

the k ing esca p es. 41 .l:.xh6+ @c7

27... .l:.e8 42 .:h7+ @b8

60 Endgame Secrets

43 @c4 [ 1 9 . 9 ]

Diagram 1 9 . 1 0

Black to move

Diagram 1 9 . 9

Black to move 53 ... @b5

After 53 . . . @ b 7 ! ? 54 l:íh7+ @b8

In spite of my extra pawn, I had we reach a position that is analysed

already accepted that the game with colour reserved in the Encyclo­

might end in a draw. White has a paedia of Endgames (Rooks 1), end­

passed pawn, but Black's king is on game 757 (Rosselli-A.Nimzowitsch,

the right side of it, namely the Semmering 1 9 2 6 ) . Black draws this

"short" one. way, too: 55 @d5 l:íxa5+ 56 @c6

43 a6 l:í a l 57 l:th8+ (the Encyclopaedia

Prevents a penetration via the b5 only gives 57 d5 l:í c l + 5 8 @b6 @c8

square. which is immediately drawn. 58

44 l:íh5 @d6 @c8 59 l:íh8+ @b7 60 @d7 a5

White pushes his a-pawn to have leads to the main line with a loss of

access to c5 with either rook or time) 57 . . . @a7 5 8 d5 l:í c l+ 59 @d7

king. Neither 44 d4 l:ía2 45 @b3 a5 60 d6 [ 1 9 . 1 1 ].

l:íd2 nor 44 @d5 l:íg5+ 45 @c6

l:íg6+ are worth it.

44 ... l:íc2+

45 @d4 @b7

46 l:íh7+ @b8

47 a5 bxa5

48 bxa5 l:ía2

49 l:íh5

To cut off the enemy king with

l:í c 5 .

49 ... l:ía4+

50 @e3 @c7

51 d4 l:íal

52 @e4 @c6 Diagram 1 9 . 1 1

53 :h6+ [ 1 9 . 1 0 ] Black to move

 Endgame Secrets 61

60 . . . @b6 (The cleanest way. 58 @a3?

60 . . . a4? loses after 61 l:Ih4 a3 62 The decisive error which spoils

:a4+. 60 . . . @a6?! is inaccurate at the previous defensive efforts. The

the least, as after 61 :h5 White white rook gets to the fifth rank
blocks the fifth rank and Black has where it serves as a shield for the
to lose another tempo for . . . @b6.
d-pawn, 5 8 . . . : e l+ is correct. Only
After e . g . 6 1... a 4 62 @d8 a3 63 :h3
after the white king hides on d8
a2 64 :a3+ @b7 65 1:Ixa2 :hl 66
does Black push his own pawn: 59
:e2 White is winning) 6 1 :b8+ ( 6 1
@d4 : ct 1 + 60 @c5 l:t c l + 61 @b6
:h5 a4 62 @e8 :e 1 + 63 @d8 a3 is
: b 1+ 62 @c7 :el+ 63 @b7 : b 1+
drawn as White won't get to the a­
64 @c8 :e 1 + 65 @d8 @a3 66 d7 a4
file with check) 61...@a6 62 @d8 a4
with a draw.
63 d7 @a5 64 @e7 :ct 1 . White can
59 :b5 a4
win his opponent' s rook, but his
N ow it is too late for checks:
king is too far away to stop the
59 . . . : e l+ 60 @d5 l:tdl+ (60 . . . a4 6 1
a-pawn.
d7 @a2 62 @d6) 6 1 @c6 a4 62 d7
54 :b6+ @xa5
:xd7 (62 . . . @a2 63 l:.d5) 63 @xd7
55 :b8
@a2 64 @c6 a3 65 @c5 @a 1 66
Now we have a "running game":
@b4 a2 67 @b3 @b 1 68 @a3+ @a 1
Black has to a) give up his rook for

the d-pawn and b) push forward his 69 :hs.

own pawn rather quickly. This 60 :d5 :et+

sounds rather simple, but the black 60 . . J�xd5+ 6 1 @xd5 @b2 62 d7

king at the edge gives rise to a hid­ a3 63 d8=� a2 is lost, as White's

den danger that Luther failed to no­ king is close: 64 �6+ @al 65

tice in time. � d 4 + @ b l 66 � d l + @b2 67 �d2+

55 @a4 @b 1 68 @c4 a 1 =¡y 69 @b3 and

56 d5 a5 wins.

57 d6 :dt 61 @d4! [ 1 9 . 1 3 ]

58 @e5 [ 1 9 .1 2 ]

Diagram 1 9 . 1 2 Diagram 1 9 . 1 3

Black to move Black to move

62 Endgame Secrets

White goes for the a-pawn. Endgame 20

61 ... !% d i + M.Suba-D.Gurevich

6 l...1%e8 62 d7 1%d8 63 @c3 @a2 Eksjo 1 9 8 2

64 @b4 a3 65 l:td2+. Earlier publication:

62 @c4 !%el+ Informan! 34/663 (M.Suba)

63 @bS 1%bl+

64 @aS 1%b8

65 1%d3+ 1%b3

66 1%d4 1%b8

67 1%xa4+ @b3

68 1%h4 1%b7

68 . . . 1%d8 69 1%h6 @c4 70 @b6

@d5 7 1 @c7.

69 1%d4 @c3

70 '.tldl [ 1 9 . 1 4 ] 1-0

Diagram 2 0 . 1 "

White to move

It ' s White to move. Formulate a

winning plan.

Commentary

Converting an extra piece into a

win might prove difficult if all the

Diagram 1 9 . 1 4 remaining pawns are on one wing.

Black to move In general, the stronger side would

like to exchange off the rooks while

the weaker side should avoid that.

As Suba shows, White can win by

force.

1 :tlb7!

The game continued less convine­

ingly 1 1%b3 1%a6 2 1%f3 1%a7, but

White won anyway.

1 1%f6

2 h 3 . 1%0

3 @e4 1%f6

4 g4 hxg4

5 hxg4 gS!?
 Endgame Secrets 63

The toughest defence. After

5 . . JHl Suba's idea unfolds: 6 .tc4

a'.f6 7 g5 a'.f5 [20.2]

Diagram 20.3

White to move

Either now or after 1 9 . . . a'.a6+ 20

Diagram 2 0 . 2 @d7 a'.a7+ 2 1 @d8 Black is forced

White to move to play . . . f7-f6. Then White re­

groups his bishop to f5, puts the

8 a'.xf7 + ! a'.xf7 9 .txf7 @xf7 1O rook onto the seventh rank and ap­

@d5 with a won pawn ending. proaches his king.

6 @es a'.f4

7 .te6 a'.f6 Endgame 2 1

8 .trs a'.a6 M . Tseitlin-A.Finkel

9 a'.bS! Beersheba 1 9 9 6

White intends to provoke . . . f7-f6. Earlier publication:

For that purpose, the rook has to be ChessBaseMagazin 5 3 (T.Enqqvist)

on the fifth rank.

9 a'.c6

White will reach his goal anyway:

9 . . . a'.a4 10 .te4 a'.c4 1 1 @f5 a'.d4 1 2

..t b l , 9 @h6 10 �c5 and .te4

a'.c6+, 9 a'.f6 1 0 a'.c5 Z:.b6 1 1 ..te4.

1O .te4 :r6

11 @d4 @h6

12 :rs a'.d6+

13 .tdS @g6

14 @es Z:.d7

15 @c6 Z:.e7

16 .tb3 l:ta7
Diagram 2 1 . 1
17 .tc4 a'.e7
White to move
18 .tds l:.a7

19 @d6 [ 2 0 . 3 ]
White to move. How do you as­

sess the position ?

64 Endgame Secrets

Commentary

Black can't win as his h-pawn has

already advanced to h5. In contrast

to endgame 20, he can't force a

pawn ending by . . . g6-g5-g4 and tak­

ing on f2. He can 't exchange rooks

either as he will end up with the

wrong bishop after a possible g3-g4.

However, in the game White

stumbled.

53 l:1e8
Diagram 2 1 . 2
White avoids exchanging rooks
White to move
after . . . 11b2. But 53 @ h 3 ! ? is feasi­

ble, e.g. 53 . . . �b3 54 @g2 l:[b2 55

65 @fl?
l:Ixb2 .lÍ.xb2 56 f3 .lÍ.c3 57 @h3 .lÍ. e l
White allows a nasty trick. After
(57 . . . .lÍ.f6 58 g4+ hxg4+ 59 fxg4+
65 @gl White draws: 65 . . . :b5 66
@f4 60 g5 .lÍ.xg5 6 1 hxg5 @xg5 62
@g2 :b2 67 11xb2 .lÍ.xb2 68 g4
@g3 with a draw) 58 g4+ hxg4+
hxg4 69 @ g 3 .lÍ. cl 70 @xg4 .lÍ.h6 7 1
( 5 8 . . . @f4 59 gxh5 gxh5 60 @g2) 59
f4@e4 72 h5.
fxg4+ @f4 60 g5 followed by
65 :es?
h4-h5.
Already now Black can win by
53 l:Ib6
means of 65 . . . .lÍ.xf2, but for the mo­
54 l:1d8 @e4
ment both players ignore this
55 l:1e8+ @d3
possibility
56 l:[f8 l:[e6
66 @g2 �bS
57 l1f4 .lÍ.cS
67 @f3 �fS+
58 11f3+ @e2
68 @g2 .lÍ.cS
59 l1f4 @el
69 @ fl ?
60 l:1a4 !í'.f6
Instead of playing 69 @ g l , W h i t e
61 l:1al+ @e2
commits the same mistake again.
62 11a2+ @d3
But this time Black is smarter.
63 @fl .lÍ.d4
69 .lÍ. x f2 !
64 @g2 �fS
70 @g2
[21.2]
After 70 �xf2 @e3 7 1 :xf5 gxf5

72 @g2 �e2 [ 2 1.3 ] (T.Enqvist)

White will lose the pawn ending.

 Endgame Secrets 65

Formulate a plan to exploit

Black's extra pawn.

Commentary

Black has an extra pawn on the

queenside which at the moment

can't advance. White can't become

active as his rook is boxed in on the

queenside and his bishop lacks a tar­

get to attack. Therefore Black qui­

etly improves his position. His plan

Diagram 2 1 . 3
consists of severa! steps:
White to move
1. Black provokes the moves

e2-e3 and h3-h4.

70 .ib6
2 He advances the king to f3 vía
71 :b2 .id4
the weakened white squares.
72 :a2 ©e4
3 . The c-pawn goes forward to c4.
73 :e2+ .ie3
4 The breakthrough . . . f5-f4 leads
74 :a2 lid5
to the capture of the �h4.
75 :a4+ ©f5
5 With two extra pawns the ex­
76 ©f3 .id4
change of rooks wins the game.
77 :a3 .tes
I don't think that Black foresaw
0-1
this plan in its entirety, but that was­

n't necessary. Tukmakov continu­

Endgame 22
ally asked himself the question:
U .Andersson-W. Tukmakov
How can I improve my own posi­
Biel 1 9 9 6
tion, how can I weaken the position
Earlier publication:
of my opponent? The realization of
ChessBaseMagazin 55 (H.-J. Hecht)
this plan took more than 40 moves,

but the win was beyond question.

Good technique is characterized by

efficiency rather than speed!

31 �d7

32 i.d2 lif7!

33 e3

White can 't allow . . . f5-f4 as then

the �g3 and the opened g-file re­

spectively are new targets of

Black's attack. After the text the

squares d3 and f3 are weakened and

the .id2 is restricted even further.

Diagram 2 2 . 1 33 :d7

Black to move
66 Endgame Secrets

An immediate 3 3 . . Jlf6 and

. . . l:.h6 is feasible as well.

34 l:.c3 l:.g7

35 �a3 .1Lb5

36 l:.c3 l:.g6 .

37 �a3 a6 [22.2]

Diagram 22 .3

White to move

Step 2 is completed.

54 :c2 .1Ld3

55 :b2 c5

56 a3

56 bxc5 allows the black rook to

Black can afford to play this move invade: 56 . . . l:.xc5

as the white bishop can no longer I) 57 @d2 @xf2 5 8 l:.xb6 l:.d5 59

'attack b6. l:íe6 f4 60 gxf4 exf4 61 SLd4 (6 1

38 l:.c3 l:.h6! exf4 e3+ 62 �e 1 l:.b5 leads to mate)

39 h4 l:.g6 6 1.. . f3 , and the f-pawn queens.

40 iLcl @d5 II) 57 SLd2 doesn 't help either:

41 .1Ld2 l:.d6 57 . . . b5 5 8 a3 h5 59 l:.a2 (or else the

42 l:.cl @e6 black rook gets to c2 or e l ) 5 9 . . . l:íc6

Step 1 is complete. 60 l:.b2 .i.c2 61 l:.a2 .i.a4 62 l:íb2

43 .1Lc3 l:.d3 l:íc5 63 l:ía2 :d5 64 l:.b2 l:.d� 65

44 .1Lb2 .1La4 l:ía2 l:.b3 66 l:í a l l:.b2 followed by

45 @e2 .1Lb5 . . . .i.b3-c4, . . . a6-a5, . . . b5 -b4. White

46 @el .1La4 can't defend as his king is caught in

47 @e2 �d5 a mating net.

Please note that Black is in no 56 c4

hurry. Now step 3 is finished. 56 . . . cxb4

48 .1Lc3 @ f7 ! 57 axb4 is less convincing.

On the way to the f3 square. 57 a4 b5

49 .1Lb2 @g6 Black permits the opening of the

50 iLal @h5 a-file as otherwise White might play

51 .1Lb2 @g4 b4-b5 and fix the �b6 on a black

52 .1Lc3 .1Lb5+ square.

53 @el @f3 58 axb5 axb5

[22.3] 59 l:ía2 h5
 Endgam e Secrets 67

60 a'.b2 f4!

61 gxf4 exf4

62 exf4 @xf4

63 :a2 @g4

64 a'.a7 @xh4

[22.4]

Diagram 22. 5

White to move

77 �f4 @l ht
There mi f
ollow: 76 ..t e l .ld3

Endgame 23
Diagram 22 4
J.Nunn-MS
H . · adler
White to mo�e
. astmgs 1 9 9 8

Earher p bl'
Jnfor u ications:
The fall of th
completion of st e �h4 means the N man/ 71/268 (M .
IC Magazine (.Sadler),
tra pawns
. Bl ac k ep
e 4 . With two ex- 211998
ChessliaseM, �.Sadler)
en6dmg ( the final s:: )go for a bishop (D T agazin 63 ,

· yomkin).
5 :g7 p .

66 .tf6 @h3

67 :g + .lc2
3
68 .tgS @h2 � �

69 @e2 : d l+ , . � � - 1/, - �

70 .tf4 :d3
■✓, - -7,,,,,�
fxg3 �xg3 ■ ■-�,
m l:t
■i-■J":i■
71
The invasion @g2
the issue.
72
of the king de 'd
c1 es
1.■•-•. "•
.tes � - • ■
73 @e3 .ld3+
� � - ¼ ■-
@'Ii. @el e3 74 .tf4 e2 75 .le5 _....r.
� /
/,
1/ �� �

73 Diagram 2 3 . 1

Threat;�ing � @fl White to move

74 .tc .. . 'é' e l - d l - c 2 .
3
75 .td2 .tbl
White t o m o ve H ·
ut the �h2 . . e is a pawn
[22.5] @g2
b
should he is under attack H up
react? · ow
0-1
68 Endgame Secrets

Commentary 60 l:ib4 (60 f6 is given by Sadler

with the idea to win the �h4 after

White can deal with the attack 60 . . . gxf6? 6 1 .:b4. But after 60 . . . g5

against h2 in three different ways. Black shouldn 't have any problems

I) 42 h 3 ? ! happened in the game. according to Sadler) 60 . . . l:ixf5 61

If ever White's kingside is under at­ l:id4+ @c5 62 .:xh4 g5 63 l:ia4 l:if4

tack, it can be fended off easily by 64 l:ia7 @d6 65 l:i a l f5 66 : d 1+

i..d5. But there is a big drawback: @c5 67 l:id5+ @b6 68 @d3 g4 69

N ow the i..d6 can no longer be hxg4 fxg4 70 i..e4 @xc6 71 g3

driven off the b8-h2 diagonal. Black l:he4 ½-½. After 72 @xe4 i..xg3

demonstrates a convincing way to White can no longer win.

the draw: 42 . . Jk8+ 43 @d5 i..g3 II) 42 .:b7 ! ? is proposed by

44 c4 @f8 45 c5 @e7 46 @e4 h5 ! Tyomkin. Straight out White attacks

(Black intends to support the i..g3 the �f7 and sacrifices the �h2:

by playing . . . h5-h4, compare this A) 42 . . . l:ic8+ 43 @d3 ! (the bishop

set-up to the construction b�a4+ ending after 43 @b5 l:ib8 is only

i..b3 in endgame 27) 47 i..d5 l:ic7 drawn) 43 . . . l:ic7 (43 . . . i..xh2 44

(Black has established a kind of for­ l:ixf7 and Black's king is in trouble)

tress) 48 c6 h4 49 @d4 l:ia7 50 i..f3 44 l:ib8+ @h7 45 h3 (45 g3 l:id7 and

i..c7 5 1 l:id5 l:.a4+ 52 @d3 i..d6 53 . . . i..xg3+) 45 . . . l:ixc2 (45 . . . l:ic3+ 46

l:ib5 i..c7 54 i..e4 l:ia3+ 55 @c4 @xc3 i..xb8 47 i..xf7 loses,

l:ia4+ 5 6 @d3 l:ia3+ 57 @c2 l:ia4 5 8 45 . . . l:id7 46 l:ia8 and the black king

i..f3 l:ia5 59 l:ib7 @d6 [23 .2] is stuck in the comer) 46 l:id8 i..e7

47 l:ie8 l:ic7 48 i..xf7 g5 49 fxg6+

@g7 and it is questionable whether

the two white extra pawns are

enou gh to win.

B) 42 . . . i..xh2 (the principal reply

as you can refute a sacrifice only by

its acceptance) 43 Wd3 (Tyomkin

shows that neither 43 @b5 :b8 nor

43 @d4 :a7! with the idea 44 l:ixa7

i.. g l+ have to be feared) 43 . . . l:id8+

(better than 43 :b8 44 i..xf7+ @f8

45 l:id7 or 43 l:if8 44 i..d5 fol-

lowed by the advance of the c­

Diagram 2 3 . 2 pawn) 44 @e2 l:ic8 45 i..xf7+ (45

White to move l:ixf7 :xc2+ 46 @fl l:ib2) 45 . . . @f8

46 i..b3 :c7 47 l:ib6 @e7 [ 2 3 . 3 ]

 Endgame Secrets 69

B) 42 .. . @f8 43 @d5 (43 �b7!?)

43 .. . �dS 44 @e6 (the white king áe­

tively joins the fight) 44 .. . �a3

(other moves are favourable for

White as well: 44 .. . @e7 45 1Ib7+,

44 h5 45 f6 gxf6 46 �xh5,

44 i.e7 45 e4 :es+ 46 @d7 l:idS+

47 @e7 i.f6 4 8 e5 @e7 49 e6) 45

i.d5 lid6+ 46 @c7 l:.f6 47 i.e4.

The c-pawn is ready to move on

while the differenee between the

two kings is obvious.

Diagram 2 3 . 3
C) 42 .. . f6 (now the Jtd6 has a
White to move
stronghold on e 5 , but the kingside is

weakened) 43 Jab7 Jte5 44 @d3+

Blaek will play .. . @dS next. I
[ 2 3 . 5 ] followed by i.e6 and the ad­
think he should be able to defend.
vanee of the e-pawn. The blaek king
III) 42 g 3 ! [ 2 3 .4 ] is the strongest
is vulnerable as well.
move as it "restriets the bishop on

d6 and leaves Blaek with huge prob­

lems" (Sadler). 42 g3 ! stieks to the

principie that in positions with

opposite-coloured bishops the

stronger side should put their pawns

onto the colour of the enemy

bishop.

Diagram 2 3 . 5

Black to move

D) 42 .. . h5 (Blaek wants to soften

up White's pawns with .. . h5-h4, but

White can attaek f7)

Dl ) 43 @d5?! JadS 44 @e6 h4 is

Diagram 23 .4 OK for Blaek: 45 Jab7 (45 gxh4

Black to move i.xh2 46 i:.b 7 Jad6+ and . . J:lf6)

45 . . . hxg3 46 hxg3 lLxg3 should

Sorne lines might eonfirm this: lead to a draw.

A) 42 .. . :es+ 43 @d5 and the D2) 43 :b7! is stronger. 43 .. . h4

i.d6 has no deeent square. 44 .:d7! (44 gxh4?! :es+ 45 @d3

70 Endgame Secrets

:c7) 44 .. . i..eS 45 gxh4 :es+ 46 White ( to move) is a healthy pawn

@d3 :c7 47 i..xf7+ @f8 48 :ds up. How would you play: 1 f3 , 1 f4

[23.6], and White remains two or maybe a different move?

pawns up.

Commentary

In addition to the extra pawn the

:rs is, for the moment, badly

placed and the black queenside

might prove weak. But the presence

of opposite-coloured bishops in­

creases Black's drawing chances.

White has three moves at his dis­

posal: The move 1 f3 was Fischer's

choice while 1 f4 is analysed by

Kasparov. 1 i..c3 ! ? deserves atten­

Diagram 2 3 . 6
tion as well. We look at the pros and
Black to m ove
cons of all three moves. '

I) 1 f3 intends a slow but steady

Endgame 24
advance on the kingside.
R.Fischer-G.Forintos
A) 1 .. . f6 (frees the :rs, but White
Monte Carlo 1 9 6 7
can take the (ia6) 2 i..d6+ @g8
Earlier publication:
(2 . . . @ f7 3 :e7+) 3 :e6 i..d5 4 :e7
Encyclopaedia of Endgames (Rooks
and :a7.
11) endgame 1 4 8 6 (G.Kasparov).
B) 1...h4 2 i..c3 followed by 3

i..e 1 and the (ih4 is weak.

C) 1.. . a 5 (the game continuation;

Black secures his queenside with

. . . a5-a4 and .. . i..d5-c4) 2 @f2 a4

(now the queenside is safe, but

Black's position on the kingside is

just too bad) 3 g4 hxg4 4 hxg4 :gs

5 i..f6! (the black king has to be

kept at the edge. 5 @g3?! followed

by . . . :e6 and .. . f7-f6 gives Black

sorne more air to breathe) 5 . . . :ds 6

f4 :d2+ 7 @g3 [24.2].

Diagram 2 4 . 1

White to move
 Endgame Secrets 71

Diagram 24.2 Diagram 24 .3

Black to move Black to move

Black has two possibilities: White wins by a direct attack

C 1) 7 . . . �d5 ! ? (the c4 square is against the enemy king, e.g.

the ideal post for the bishop, de­ 1 2 .. . l:thl (the game saw 1 2 .. . @e8 13

fending both wings) 8 l:t e 5 ! (8 f5?! l:td3 1-0) 13 lid3 .i.c6 14 l:td8+

i.c4 9 @f4 l:te2 10 l:th3 @e8 isn't i. e 8 1 5 l:tb8 l:tel 1 6 Si.d8 @g8 (or

convincing. Black's king escapes 1 6 .. . @g7 1 7 i.e7 l:txe7 18 f6+) 17

while White has problems to utilize @f6 @h7 18 i.e7 (Kasparov), and

his majority) 8 .. . i.c4 9 l:tc5 lie2 the �f7 is lost.

(9 @e8 10 l:tc8+ @d7 11 l:td8+, II) 1 f4!? is strategically risky,

9 @g8 10 lic8+ @h7 11 f5) 10 since the a8-hl diagonal is opened

@h4 @e8 11 l:tc7 @f8 1 2 @g5 líe 1 up for the i.c6. Should the l:tf5

13 l:.b7 l:te3 1 4 f5 liel ( 1 4 . .. @e8 1 5 come into play, the �g2 might prove

@h6 l:tg3 16 g5 l:tf3 17 @g7) 15 weak. The white bishop is tied to

l:tb8+ l:te8 1 6 i.g7+ @e7 1 7 lib7+ the defence of the �f4 and has lost

@d6 1 8 @ f6 and White wins. its influence on the squares g3 and

C2) 7 .. . l:tg2+ 8 @h4 lid2 9 f5 h2. But the good news is the fact

i.d5 1O @g5 lid 1 11 l:tc3 l:te 1 that White is able to win the �a6.

( 1 1.. . @ e 8 12 l:tc5 with a win ac­ Concrete variations take precedence

cording to Kasparov as 1 2 .. . �c4 13 over general assessments: 1...h4

l:tc8+ @d7 1 4 l:td8+ loses the rook) (forced in view of 2 g4) 2 l:tc3 i.e4

1 2 l:. h 3 [24.3]. (After other moves White stays two

pawns up, too: 2 . .. i.d7 3 l:tc7 @e8

4 l:ta7 f6 5 i.d6 l:td5 6 l:txa6 l:td2 7

i.b4 l:txb2 8 l:txf6 (Kasparov),

2 .. . i.e8 3 l:tc7 f6 4 �d6+ @g8 5

lia7, 2 .. . i.d5 3 l:tc8+ @e7 4 l:tc7+

@f8 5 l:ta7) 3 l:tc8+ @e7 4 l:tc7+

[24.4]. Black will lose material

wherever his king goes:

72 Endgame Secrets

Diagram 24.4 Diagram 24.5

Black to move Black to move

A) 4 . . . @d8 5 :ra7 f6 6 i.d6 :ras 7 III) 1 i. c 3 ! ? [24.6] is another

i.e7+ @c8 8 i.xf6 .:.a2 9 :rg7. promising move.

B) 4 . . . @e8 5 :ra7 f6 6 i.d6

(threatening :re7+) 6 . . . i.c6 7 :e7+

(7 :rxa6?! @d7 is less convincing)

7 . . . @d8 8 :rh7 a n d a pawn is lost.

C) 4 . . . @e6 5 !:ta 7 f6 6 l:txa6+ @fl

(The black rook intends to get back

into play with . . . :rd5-d2. White has

to be careful) 7 i.d6 l:td5 8 i. b4

(protects d2) 8 . . . :dl + (8 . . . l:t d 3 9

i.c3 f5 1O :rf6+ @e8 11 :g6) 9

@f2 :re 1 1 O i.c3 f5 ( 1 O . . . .:.c2+ 11

@e3 i.xg2 1 2 :rxf6+ @e7 13 :rh6

i.xh3 14 :rxh4) 11 :r f6 + @e8 12

Diagram 24.6
:rh6 l:tc2+ 13 @e3 .:.xg2 1 4 l:txh4
White to move
:g3+ 1 5 @d4 i.g2 1 6 @c5 [24.5]

and White keeps two extra pawns.

White releases the :fs but pre­

vents . . . a5-a4. Therefore Black is

left with the weak lJa6. The game

might continue like this: 1... :r d S ! ? 2

f3 i.d7 3 @f2 i.f5 4 g4 hxg4 5

hxg4 i.g6 6 @g3 l:td3 7 l:te5 [24. 7 ] .

White can go on with g4-g5,

@g3-g4 and f3-f4-f5.

 Endgame Secrets 73

strategical players like Kramnik or

Karpov would choose between 1 f3

or 1 i. c 3 .

Endgame 25

A. Yusupov- L. Christiansen

Germany 1993

Earlier publication: lnformant 57/72

(A.Yusupov).

Diagram 2 4 . 7

White to move

Which of the three moves should

be played in a practica} game? I

doubt that 1 f4 is the right choice in

a toumament situation. 1 f4 is

strategically risky but wins a pawn

thanks to concrete lines. If White

isn't able to calculate these lines

correctly (for instance if he is in Diagram 2 5 . 1

time trouble) he should keep away Black to move

from playing 1 f4. The risk is too

great that Black might start an at­ Assess the position. Should Black

tack against g2 with . . . �f5-d5-d2. 1 play 25 . . . i.e2 or 25 . . . i.xe3 ?

f3 is much more circumspect as

White keeps a healthy pawn struc­ Commentary

ture. He has faith that his advantage

is great enough to win. But Black is Black is a pawn down. However,

able to secure his queenside with the pair of bishops should give him

. . . a6-a5-a4. Even after 1 f3 White compensation.

has to show a certain accuracy (look 25 i.xe3?

at the moves 5 i. f6 ! and the possi­ Black cherishes false hopes that

ble line 7 Wg3 i.d5 8 -'l e 5 ! ) . The the opposite-coloured bishops and

final proposal is 1 i.c3 ! ? which is the devalued white pawn structure

slower than 1 f4 or 1 f3 but doesn't should guarantee him a draw. It will

require any calculation at all. White become apparent soon that White

keeps all the positive aspects of his actually has a clear advantage. ln­

position and prevents . . . a6-a5-a4. stead, 25 . . . i.e2 (Yusupov) would

How would top players play this po­ have been correct, with only a slight

sition? I guess that a tactician like advantage for White.

Shirov would go for 1 f4, while 26 fxe3 i.e2

74 Endgame Secrets

Black establishes a blockade on 33 @f3 ..tf5

b 5 . After 26 . . . ..td5 27 b4 h5 (to pre­ 34 e4 ..te6

vent g2-g4) White can resolve his 35 b4 IIc8

doubled pawn with 28 IIc5 and 29 36 ffü7 ..tc4

b5. 37 @f4 ii.b5

27 @ f2 ii.b5 38 IIa7 @g7

28 g4! [25.3]

Just in time before Black plays

. . . h7-h5. Now the ..tf6 has taken up

a firm post.

28 h6

29 h4 ..td7

30 g5 h5 [25.2]

Diagram 2 5 . 3

White to move

39 b3!

A remarkable move. White denies

the enemy pieces access to c4.

Diagram 2 5 . 2 Black can't attack the Lsb3 as his

White to move rook has to defend the back rank

while his bishop is tied to the Lsa6.

Without the rooks the players This move repeats a theme we have

could agree to a draw immediately. already encountered in game 23:

Black might then shuffle his bishop with opposite-coloured bishops on

along the diagonal b5-fl and White the board the stronger side should

can't do anything. With rooks on, use its pawns to restrict the enemy

things are very different. Black' s bishop.

rook has to defend against a back 39 ..tn

rank mate. White's winning idea is 40 e5 @g8

straightforward: Get the king to b6 41 @e4 ..te2

and sacrifice the exchange on a6. 42 @d4 ..tfl

Then the passed b-pawns decide. 43 ..td6!

Obviously this plan needs sorne White intends i..c7 and @c5-b6

preparation, but there is hardly any followed by the sacrifice on a6

defence for Black. (Yusupov).

31 �c7 ii.b5 43 ..tb5

32 i..e7 ..td3 44 i..c7 @f8

 Endgame Secrets 75

45 @c5 @e7 Endgame 26

46 @d5 C.H.O.D.Alexander-W .Smyslov

With the threat i..d6+ (Yusupov). England v USSR match,

46 i. d 7? ! London 1 9 5 4

Black hastens the end, but even Earlier publication:

46 . . . @ f8 47 @d6 (Yusupov) isn't Die Kunst des Endspiels

really better. Due to zugzwang (W.Smyslov).

Black has to allow the decisive sac­

rifice on a6, e.g. 4 7 . . . i. fl ( or

47 . . J:!e8 48 @c5 :es 49 @b6 and

i.d6+, :xa6) 48 @d7 i.h3+ 49

@c6 i.fl 50 @b7 :es 51 i.d6+

@g8 52 .:f.xa6 i.xa6+ 53 Wxa6 and

the b-pawn queens.

47 i..d6+ @d8

48 i.c5! �c6

49 :as+ @c7

50 i.d6+ [ 2 5 .4 ] 1-0

Diagram 2 6 . 1

White to move

Analyse the game.

21 g3 :bs 22 b3 �e8 23 i.b5

:e6 24 :dl c6 25 f5 �d6 26 i.d3

@f8 27 @g2 i.b4 28 zn @e7 29

:n :ds 30 @f3 @d6 3 1 :e2 i..c3

32 @f4 :b8 33 h4 a5 34 @g4 c5 35

@h5 a4 36 g4 i..e5 37 g5 hxg5 38

hxg5 :b4 39 :e3 :r4 40 c4 i.d4

Diagram 2 5 .4
4 1 .:f.g3 a3 42 cxd5 0-1
Black to move

There might follow: 50 . . . @b7 51

:b8+ @a7 52 .:f.d8 (Yusupov).

7 6 Endgame Secrets

Commentary i.fl is unpleasant. 25 . . . i.xf4 26

.t! c 8 + @h7 27 @f2 lets the white

It is surprising that White loses king escape) 26 b3!? (26 .:[xa7 d4

this position without much ado. It's intending . . . i.xf4 and after 27 .:[c7

true that the black pieces are being Black plays 27 . . . d3) 26 . . . d4 27 .t!c4

conducted by endgame expert .t! dl ! ? [ 2 6 . 2]

Vasily Smyslov, but the opponent

on the white side, C . H . O . D . Alexan­

der, was one of the best British

players of those days. At first sight

the position seems to be rather dead,

but on a second look you notice that

White has problems on the king­

side. Especially the squares e3, f4

and g3 are weak. Besides, Black has

a pawn majority in the centre. The

�d5 controls the squares c4 and e4,

thus restricting the i. d 3 . Note that

things would be different if this

Diagram 2 6 . 2
pawn were on d4. The black rook
White to move
might become active along the b­

file. At the moment Black enjoys White can' t move either his king

only these small advantages but it or b i shop. I don ' t see how Black can

takes him only twenty more moves make use of this circumstance, but

to win. it's understandable that White does­

21 g3 n 't go for such a line.

White prepares @g2-f3, but he 11) 2 1 c4 is another important re­

lets Black have the e-file. Neverthe­ source which we will encounter sev­

less the text is quite OK. There are eral times in the game. White

two altematives which illustrate undermines the centre, how should

sorne o f Wh i t e ' s problems. Black react? After 2 1. .. c6 22 cxd5

I) 21 l:I e l (the obvious choice) cxd5 the �d5 has lost its support.

2 1... i. f2 (the only try to stop White After 2 1 . . . d4 Black's passed pawn is

from playing 22 .t!e2 and 23 g3) 22 b loc k ed and obstructs the i. c 5 . The

lie7 (22 l:Ie2 i.g3 23 f5 @f8 is white queenside doesn't offer any

slightly favourable for Black as the attacking points as White hasn 't

move f4-f5 has only weakened the played b2 -b3 yet. However, Black

e5 square and the i.d3. On top of has another possi b i li ty : 2 1.. .dxc4 22

that the white king is badly placed) i.xc4 :d2 23 .:[b 1 (23 b3 ? .t!xa2 24

22 . . . i.g3 (22 . . . c6 23 @h2 followed b4 .:[c2 loses a pawn) 2 3 . . . i.d4

by g2-g3 leads to equality) 23 .tlxc7 (23 . . . Í!d4 24 l:I c l i.d6 25 g3 Í!d2

.tle8 24 @gl l:I e l + (after 24 . . . i.xf4 26 :b 1 , 23 . . . i.d6 24 g3 followed by

25 .:[d7 White threatens both .t!xd5 a 2- a 4 , b 2 -b3 and l:I b l - e l - e 2 ) 24 b4

and @f2) 25 i..fl Í!c 1 (Pinning the i. e3 25 g3 [ 2 6 . 3 ]

 Endgame Secrets 77

exchange by 28 . . . : g l + 29 @f3 he

has to face the threat i.b7.

III) 25 . . . c6 26 i.a6 g5 ! ? keeps the

initiative alive. White is reminded

of his kingside weaknesses, e.g. 27

:d2 (27 fxg5 hxg5 and . . . i.d6)

27 . . . gxf4 28 gxf4 i.d6 [26.4]

Diagram 2 6 . 3

B lack to move

Once again the white position

looks rather suspicious. Though the

move c2-c4 is an important idea, it

isn't yet time to play it.

21 :bS!

A characteristic manoeuvre. Be­ Diagram 26.4

fore seizing the e-file Black pro­
White to move
vokes a weakness.
22 ... :es
22 b3
23 i. b S ? !
The weak c3 square might be ex­
Over the next two moves White
ploited by Bl a c k ' s bishop or king.
retakes the e-file, but by doing so he
In addition, the L)b3 can be pressur­
plays the compromising f4-f5. 23
ized by means of . . . a7-a5-a4. 22
@g2 is a simpler solution: 23 . . . �f8
:b 1 ! ? avoids this weakening for the
24 :dl c6 25 :d2 �el 26 .i:'¡e2
moment. But the rook is rather pas­
[26.5]
sive: 22 . . . @ f8 ! ? (22 . . . i.a3 23 c4

leads nowhere) 23 c3!? (23 @g2

@e7 24 @f3 @d6. To free his rook

White has

23 . . . :e8
to

(c2-c3
play

has
b2-b3

weakened
anyway)

the
·g·g·'�g
Lfil·g·g·EJ�"
mªifmm ■�
d3 square, thus . . . :e3 is unpleasant.

23 . . . @e7 24 : e l+ @d6 25 :e2 is

less critical) 24 @g2 :e3 25 :a1 ••·,,••�m•

andnow: _

■ 8 ■ Jt ■ �8
I)

27 b4
25 . . . @e7 26 i.fl

and c3-c4) 27 :d2 and after

c6 (26 .. . @d6

:.:.1.·i■
.i:'¡e2 the rooks will come off, with a

draw.
� � �
� -
II) 2 5 . . . a5 26 i.a6 c6 27 :d2 : e l Diagram 26.5

28 ¡[e2. If Black avoids the Black to move

78 Endgame Secrets

After exchange of rooks the draw 11) 27 c4 d4 (27 . . . dxc4 28 bxc4

can be signed while after 26 . . Jígl+ leaves Black with a symbolic advan­

27 @f3 the t::.c6 is susceptible to the tage only) 28 l:lel. The black king is

manoeuvre .ila6-b7. Thus 23 @g2 cut off the centre while his white

seems to draw quite comfortably. counterpart can join the fight ac­

23 l:le6 tively. Once again White is out of

24 :dl c6 trouble.

25 f5 27 .ilb4!

Consistent but strategically risky, Now the white rook has to lose

too. precious time to get to the e-file.

25 ... l:ld6 This enables Black to centralize his

26 .i.d3 @f8 king.

[26.6] 28 :n @e7

29 :n .:.ds

30 @ f3 ? !

30 c 4 ! ? d4 3 1 @f3 @d6 32 .:.e2 is

better. Compare the note to 3 1 l:le2.

30 ... @d6?!

30 . . . �c3 prevents 3 1 c4. '

31 .:.e2?!

31 c4 is preferable, since after

3 1... d 4 32 :e2 the bishop is mis­

placed on b4 (it would be better on

b4 ) . Black has to watch ideas like

l:le4-g4.

31 .ilc3

Diagram 26.6 [26. 7 ]

White to move

27

This is
@g2?!

already a serious inaccu­

•■
!,'.,J.
.t].,.�
■- •• f,,,,[�0
racy. 27 l:l e l ? .i. t1 loses a pawn, but

:i:�i•:
White has two continuations which

should lead to an almost equal

■
� �.l, l, . t ■
�.,, .•
position:

1) 27 :d2 @e 7 ( otherwise the ■ �D�

black king will be cut off) 28 l:le2+

@d8

placed,
(now

they
king

would
and rook

rather
are mis­

change
,•,•¡•.�
their places) 29 c3 (White plans
Diagram 26. 7
b3-b4 and c3-c4 to attack the centre
White to move
pawns) 29 . . . a5 30 c4 d4 31 .:es

.i.b4 32 a3 .i.xa3 33 .:.xa5, and B lack physically prevents c2 - c4.

White has no more problems. 32 @f4?!

 Endgame Secrets 79

Over the next moves White starts

an offensive on the kingside which

doesn 't yield anything apart from a

further weakening. Black's initiative

on the queenside is much more to

the point. He intends . . Jib8 and

. . . a7-a5-a4. White has two ways to

defend:

I) 32 l:re3 (White drives away the

i.c3 and accomplishes c2-c4)

32 . . . a 5 3 3 i.fl i.f6 34 c4 d4. The

black king can penetrate the queen­

Diagram 2 6 . 8
side via c5 and b4, supported by the
White to move
a-pawn. The black bishop is ideally

placed on f6 as it protects the pawns The white problems are obvious.

on d4 and g7 as well as the squares After . . . a5-a4 Black plans either

e5 and e7. White is in a sorry state. . .. a4-a3 or .. . c5-c4. The ¿Ja2 tums

11) However, 3 2 i. a 6 ! ? is feasible. out to be the decisive weakness. ln­

White blocks the a-pawn and it isn't stead o f t h e t e x t , 34 . . . a 4 ! ? is possi­

easy for Black to get around this, ble as well.

e.g. 32 . . . l:rb8 33 l:re3 i.f6 34 l:re2 35 @h5

l:rb6 3 5 i. d 3 . The threat is l:re8 and The white kingside attack is just

after 3 5 . . . i.e5 36 c4 Black has to in vain. But things are ve:ry difficult

defend against 37 c5+. anyway:

32 :bs I) It is already to late for 3 5 a4 as

33 h4?! after 3 5 . . . l:rb4+ and . . . c5-c4 the �a4

Another weakening. Preferable is is lost.

3 3 a4, see the next note. 11) 3 5 l:rf2 ! ? allows the i..d3 to re­

33 a5 treat, but after 3 5 . .. a4 3 6 bxa4 c4 3 7

34 @g4?! i.fl l:rb2 Black wins both a-pawns.

White simply ignores Black's ad­ Nevertheless this is White's best

vanee at the queenside. He has to chance.

play 3 4 a4 to establish a stronghold 35 a4

on b5. lf after 34 l:rb4+ 35 �f3 36 g4

Black continues 35 c 5 ? ! White gets 36 bxa4? c4 demonstrates the

counterplay with 36 i. b 5 ! i.e5 37 helplessness of the i. d 3 .

Íle8 f6 38 ssi. Because of that 36 i.e5

Black should leave the c-pawn on 37 g5 hxg5

c6 and prefer 3 5 . . . i.e5 instead. He 38 hxg5 l:fü4 !

can support the bishop with . . . f7-f6 . . . i.g3 (Smyslov) is threatened.

and the king might go to c 3 . White 39 l:re3 :r4

has to reckon with . . . :g4 as well, a 40 c4

consequence ofh3-h4. Finally this move takes place, but

34 c5 [ 2 6 . 8 ] much too late.

80 Endgame Secrets

40 ... ..td4 Endgame 2 7

41 l:I g 3 V.Kramnik-G.Kasparov

Or 4 1 :e2 ..tf2 (Smyslov). Novgorod 1 9 9 5

41 a3! Earlier publications: Informan!

42 cxd5 [26.9] 63/277 (V.Kramnik), ChessBase­

Magazin 48 (S.Dolmatov)

Diagram 2 6 . 9

Black to move Diagram 27 . 1

White to move

Here the game was adjoumed and

Analyse the game.
later on White resigned. According

to Smyslov Black can play 42 .. . : t2 25 d4 hxg4 26 hxg4 f5 27 gxf5

43 ..tb 1 l:Ib2, winning the a-pawn. ..txf5 28 ..tc5 a5 29 l:Ie7 l:Ic6 30

l:Ie5 ..tbl 31 a3 @g8 32 f3 a4 33

@f2 l:If6 34 l:Ie8+ @ti 35 l:Ie7+

@g6 36 @e3 ..ta2 37 l:Ia7 ..td5 38

f4 @f5 39 l:Ixg7 :e6+ 40 @d3

@xf4 41 �ti+ @g4 42 l:Id7 ..tb3 43

..tb4 @f3 44 l:Iti+ @g4 45 l:Id7

@f3 46 l:Iti+ @g4 47 l:If8 l:Ih6 48

@e4 l:Ie6+ 49 @d3 :h6 50 ..tc5

l:Ih3+ 5 1 <it>e4 l:Ih5 52 l:If4+ @g3 53

@e3 :h6 54 xn @g4 55 @e4

l:Ie6+ 56 <it>d3 l:.e8 57 @c3 :b8 58

l:If2 ½-½

Commentary

Over the last few years Gany

Kasparov fought many spectacular

duels with his crown prince Vladi­

mir Kramnik. In the present game

Kramnik is close to victory as he

 Endgame Secrets 81

enjoys a healthy extra pawn. But the 32 . . . f6 ! ? (Black denies White's

opposite-coloured bishops impede king the squares e5 and g5, but he

the winning process. weakens the seventh rank.

25 d4 32 . . . ..t c 2 ! ? 3 3 f3 ..tdl is interesting.

Apart from being a pawn down If White's king supports the d-pawn

Black suffers from the weak tsa. the ¿jf3 might become vulnerable)

25 hxg4 3 3 lie7 ..tc2 34 :d7 :d3+ 3 5 @h4

26 hxg4 f5? ( 3 5 @f4 ..tdl 3 6 ..tf8 @g8 3 7 ..txg7

This move is criticized by Kram­ :f3+ 38 @e4 :xf2 is unclear)

nik. Black exchanges a pair of 3 5 . . . @g8 (with the threat of

pawns, but he weakens the seventh . . . g7-g5+, but 3 5 . . . :n ! ? is notewor­

rank and the e6 square. Furthermore thy) 3 6 g5? (36 f4 ..tb3 3 7 f5 is nec­

Black doesn't care about his queen­ essary, White might proceed with

side. White should be winning now. g4-g5 then) 3 6 . . . :f3 (Black has no

Better is 26 . . J�c4 27 ..tc5 a5 "with a more problems, now it's White who

clear white advantage" according to has to take care in view of his badly

Kramnik. Black refrains from weak­ placed king) 3 7 gxf6 gxf6 3 8 :d8+

ening his position and his a-pawn @f7 3 9 @g4 (39 d5? .i.f5) 39 . . . ..tdl

gets to a4. As an example how 40 :d7+ @g6 4 1 @h4 :xf2 42 d5

things might proceed, I present the :f3 43 ..tf8 @f5 44 :a 7 :d3? !

moves of the training game Holger (Stronger is 44 . . . ..tb3 45 d6 ..te6

Ellers-Christian Wilhelmi. This and White threatens to take on a 3 .

game started with position 27 . 1 , 46 lha4 and 46 d7 aren't possible

with a time limit of one hour per then due to 46 . . . �g6 and mate on

player. There followed: 25 d4 � c 4 ! ? h3) 45 d6 : x a 3 ? ! (45 . . . ..tb3 is still

(Black delays exchanging on g4, but the better choice) 46 d7 lid3 47

this isn 't very relevant) 26 ..tc5 a5 lie7. Black lost on time. After

27 @g2 a4 28 �e3 hxg4 29 hxg4 4 7 . . . ..tg4 48 :as+ �f4 49 lia4+

�c2 30 a3 �d2 31 �e8+ @h7 32 @f5 50 :xg4 :xd7 we arrive at the

@g3 [27.2] notorious endgame "l:i + ..t v :",

compare endgame 3 8 .

27 gxf5

There are sorne important

altematives:

I) 27 :e7 fxg4 28 .:.xa7 :c4 29

..tc5 ..te4 with counterplay is given

by Dolmatov. White's king feels un­

comfortable as Black intends

. . .lkl+ and . . . :c3-h3+. 30 :a3 pro­

tects the third rank, but after

30 . . . ..tdS there is the threat of

. . . :e 1 + and . . . :c2.

II) 27 ..t c 5 ! ? fxg4 28 :e7 (sug­

Diagram 27.2
gested by Kramnik) 28 . .. a 6 (or
Black to move
82 Endgame Secrets

2 8 . .. a5 29 l:ta7) 29 @g2. Black faces and the d-pawn queens. Therefore

a tough struggle in view of @g3-f4. Black has to try 27 .. . @h7, after

However, 29 .t d 6 ? ! @g8 3 0 Jle5 is which 28 gxf5 .txf5 29 .:Ia6 leads

weaker because of 30 . . . .tf7 31 a3 to positions similar to those after 27

l:tc3 32 d5 l:txa3 3 3 d6 .:ld3 34 d7? gxf5 .txf5 28 l:te5 ! , s ee next note.

@ f8 . 27 .tx5

III) 27 g5!? (avoids an exchange [ 2 7 .4 ]

of pawns) 27 . .. Ik4 28 Jl c 5 a5 29

.�,��-�-­
@g2 l:tc3 and .. . f5-f4. Black's f­

pawn restricts the white king, while

White has no use for the � g 5 . ■

,, , � m • � �� ¼,,,,7,�
IV) 27 l:t e 6 ! is strong. The point is

the exchange sac after 27 .. . sn 28 ■ ■ ■ -

gxf5!.

(28 .. . .txe6
White

29
wins,

fxe6
e.g.

leaves
28 .. . l:tc4

Black's
•m•�•a•�-
.;,,,,,,,-1/,,,,,7,.
M to� -
�- -�
rook helpless against the advance of

White's d-pawn) 29

(29 .. . l:ta4 30 d5 .:.as 3 1 d6 l:txc5 3 2

.tc5 .txe6
8 ■ ■ D ■
d7 l:td5 33 l:te8+) 30 fxe6 @g8
m m a m
(now the d-pawn can't be pushed
Diagram 2 7 .4
immediately, but White wins never-
White to move
-theless) 31 .txa7 �a4 ( 3 1... l:t c 6 3 2

e7 @f7 3 3 .tc5 .:a6 34 a3 leads to a 28 .tes?

slow but steady win for White) 32 Th is m ove is based on a misca lc u ­

.tc5 l:txa2 33 d5 .:e2 [27.3] lation ( cf. the note to 30 :es) and

(33 .. . l:ta5 34 d6 lhc5 3 5 d7) prob ab ly squanders the win. 28

Jl c5? cl o s es the e - fil e , but it has tw o

grave disadvantages : F irstly, the d­

pa wn cann ot advance as lon g as

B lack's roo k att acks the Jl c5. Sec­

ondly, Black can advance his prob­

lem ch ild , the a-pawn, to a4 . Blac k

la ter o n e s ta b l ishes the set-up b�a4

+ Jlb3 which supports his defence

eno rm ousl y . P awn and bisho p pro­

tect e a c h other and the bishop con ­

t ro l s the diagonal a2-g8, in cl udin g

the d5 square. U ntil the end of the

game the d-pawn won 't cross that

Diagram 2 7 . 3
squ are. W hite has three other moves
White to move
at his disposal :

34 @fl! (34 d6? l:txe6 35 d7 1 ) 28 .:I e7 :c4 (28 .. . a6 29 .t d6

.:! e l + 3 6 @g2 l:t d l is an unpleasant foll owed by .t es is j us t hopeless)

surprise) 34 .. . l:te5 3 5 d6 .:Ixe6 36 d7 29 .tes a5 30 l:ta7 a4 31 @ g2

 Endgame Secrets 83

(White plans advancing his king to l:th5+ @g8 3 2 d6 and White has at­

f4) 3 l... g 5 ! ? (I don't see anything tained his ideal arrangement.

else) 32 .td6 (32 a3 .te4+ 3 3 @g3 In passing, I would like to remark

.tf5 34 l:te7 klc3+ 3 5 @g2 g4 with the following: Though Kasparov

sorne kind of counterplay) 3 2 . .. @g8 and Kramnik are super-class grand­

33 .te5 @ f8 . Though Bl a c k ' s posi­ masters, they are not immune to er­

tion looks very suspicious, the game rors. The moves 26 .. . f5? and 28

isn ' t over yet. .t c 5 ? are characteristic mistakes: In

II) 2 8 d5. This move suggests it­ a worse position, Kasparov usually

self, but there is a drawback. White defends in the most active way,

wants to reach the arrangement even though this sometimes isn 't

w.i.b4 + �a3 + �d6. Then pawns appropriate. 26 . .. f5? is a pseudo­

and bishop defend each other, while active move which merely weakens

the rook is ready to enter e 7 . How­ Black's position. This shortcoming

ever, Black can interfere: 28 .. . l:tc4 of Kasparov's play is discussed by

(28 . .. a5 29 .te7 l:tc2 3 0 a3 l:td2 and the well-known Russian coach Mark

now not 3 1 d6? as then the bishop is Dvoretsky in chapter 14 of his book

wrong-placed. Better is 31 l:te5 ! , Attack and Defence. On the other

and after 3 2 .tc5 and 33 d6 White hand, 28 .t c 5 ? is typical for

arranges his pieces ideally) 29 a3 a5 Kramnik, as he sometimes lacks the

3 0 l:te8+ (30 .te7 l:te4) 30 . . . @h7 3 1 precise calculation needed to

i..e7 l:td4 32 kid8 i..e4 33 d6 a4. convert an advantageous position

White' s bishop is on a bad square, into a win.

Black has counterplay against the 28 a5!

�a3. Obviously Black grabs the chance

III) 28 l:te5 ! should decide the is­ to establish the above-mentioned

sue. Along the fifth rank the rook piece arrangement. 28 .. . a6? is much

controls the d5 square (important to weaker.

advance the passed pawn) as well as 29 l:te7

the a5 square (important for an at­ A loss of time, but other moves

tack against the �a7). Black faces aren't essentially better. 29 @g2

insurmountable difficulties. Sorne "intending @g3-f4" is winning

lines may serve as proof: according to Kramnik but I don't

A) 2 8 i..d7 29 l:te7. agree. It isn't clear what results

B) 28 i..bl 29 kla5 l:t c l + 30 from the white king advancing after

@g2 g5 (30 .. . .te4+ 3 1 @g3) 3 1 a 3 . 29 .. . l:tc6. Along the sixth rank the

C) 28 .td3 29 d 5 . rook is well-placed. Either it can

D) 28 .tc2 29 l:.a5 ( or 29 d5) offer an exchange on e6 or it can

29 . .. l:tc4 3 0 i..c5 l:ta4 3 1 l:txa7. bother the white king with checks:

E) 28 .. . l:t c l + 29 @g2 leaves Black I) 3 0 l:te7 transposes to the game.

with the same sort of problems as 11) 30 @g3 l:te6 3 1 l:td 1 (31 l:tc 1

one move before. i..e4 and .. . i..d5) 3 1... @ g 8 ! ? , with

F) 2 8 .. . l:tc4 29 a3 �d3 (29 .. . g6 30 the point that after 32 d5 l:.e2 3 3 a3

l:ta5 @g8 31 .tc5) 30 d5 l:te4 31 l:tc2 the i..c5 can't defend the �a3
84 Endgame Secrets

anymore, e.g. 34 i..e7 @f7 35 d6 II) 30 f3 i..e6 31 a3 i..d5

ik3+ 36 @f4 i..d7. (Kramnik) or

III) 30 .:es+ �h7 3 1 @g3 i..e6 32 III) 30 �g2 ! ? i..e6 3 1 a3 a4 allow

a3 a4 3 3 @f4 i..b3 (White can't go Black to defend as well.

on though he has an improved ver­ 30 i..bl

sion of the game) 34 @e5 (34 @e3 31 a3 @g8

.:e6+, 34 .:ds ©g6 35 @e3 .:e6+ 32 f3 a4

36 @d3 .:f6 gives White headaches 3 2 . . . i..a2 is another possibility,

about the �f2, e.g. 37 d5 .:fs 38 with the idea 32 @f2 .:e6 3 3 .:es+

@e 4 i..c2+ 39 ©e3 i..b3) 34 . . . @g6 @f7 34 .:e7+ @g6 3 5 .:a7 .:f6. In

3 5 f4 .:f6 [27.5] exchange for the a-pawn Black gets

the � f3 . T h e n the passed g-pawn

should offer enough counterplay for

a successful defence.

33 �f2 [27.6]

Diagram 2 7 . 5

White to move

The white king has to stop.

29 .:c6! Diagram 2 7 . 6
Black defends against the threats Black to move
3 0 .:a7 and 30 i..d6, 3 1 i.. e 5 .

30 .:es If 33 i.. b4 ! ? (idea d4-d5-d6), then

The rook remorsefully returns to 3 3 . . . .: f6 34 @f2 .:fs (Kramnik).

the fifth rank. Here Kramnik must But after the text 34 i..b4 is a real

have realized that a large part of his threat.

advantage has disappeared. 33 .:r6

I) 30 .:a7 probably has been N ecessary to stop the d-pawn.

Kramnik's intention when playing 3 3 . . . i..a2? 34 i..b4 lie6 (34 . . . .:c2+

2 8 i..c5. The �a5 is doomed to fall, 3 5 @g3 i..c4 3 6 d5 �e2 3 7 _:g5) 3 5

but 3 0 . . . .: g 6 + 3 1 @fl ( 3 1 @h2 i..e4 .:as i..b3 3 6 d5 .:es 37 .:a8+ @f7

wins the �f2) 3 l....:a6! (Kramnik) 3 8 d6 with a win for White accord­

defends. 32 .:xa6 i..d3+ leads to a ing to Kramnik.

draw. 34 .:es+
 Endgame Secrets 85

Kramnik and Dolmatov point out . . . we anive at the drawn position

that 34 d5 leads nowhere: 34 . . JHS mentioned above. This type of posi­

3 5 :es+ @t7 36 :ds ..ie4 3 7 :f8+ tion has been analysed by Horwitz

@g6 3 8 :xf5 ..ixf5 with a draw. and Kling in 1 8 5 1 and by Rauzer in

34 @f7 1928. White can't win: lf his king

35 :e7+ @g6 captures the ¿Ja4, Black's king gets

36 @e3?! to a8. Should White try to push

White permits an exchange of away the opponent, the black king

pawns. has to stay on the fifth to eighth

I) 3 6 d5 :rs 37 :d7 ..ie4 or rank. If the black king is driven to

11) 36 :a7 ..ie4 37 .: x a 4 :xf3+ the first to fourth rank, White can

3 8 @e2 @f5 aren't better, but after win. Here are sorne lines for the

III) 36 @g3 Black still has to sake of clarification: 47 ..id4 �g6

suffer. · (47 . . . @f4 4 8 @e6 @e4? 49 ..ie5 is a

36 ..ia2 win for White according to Rauzer.

37 :a7 ..id5 By subtle manoeuvring the black

38 f4 @f5! king is forced to h l )

39 :xg7 :e6+ A) 48 ..ie5 @f5 49 @d6 �g6 50

40 @d3 @xf4 @c6 @f5 (but not 50 .. . @t7? 5 1 @b5

Now Black clearly is beyond the @e6 52 @xa4 @d7 53 �b5 @c8 54

worst. It is important to notice that a @c6) 5 1 ..ih2 @e6 or

drawn position results if Black man­ B) 48 @e6 @g5 49 ..ie5 @g6 50

ages to exchange :e6 and ..id5 ..if6 @h7 5 1 @t7 @h6 52 @g8 @g6

against :g7 and ¿Jd4. 53 ..ig7 @f5 54 @t7 @g5 55 ..if8

41 :n+ @g4 @f5 56 ..ie7 @e5 57 @e8 @e6 and

42 :d7 ..ib3 White hasn't made any progress at

43 ..ib4 all.

There are two other moves: 11) 43 @c3!? "with the idea

I) After 43 d5 :es 44 @d4 :xd5+ @b4-b5" is suggested by Kramnik.

45 :xd5 ..ixd5 46 @xd5 @f5 [ 2 7 . 7 ] But after 43 .. . :e3+ 44 @b4 :d3 I

don't see a decent plan for White.

Should the d-pawn advance we

reach the drawn position above. The

only other idea, an exchange sac on

a4, remains fruitless as well, e . g . 45

:a7 @f5 46 :xa4 ..txa4 47 @xa4

@e6 with a draw.

43 @f3

44 :n+
Kramnik analyses that 44 d5 leads

to draw: 44 .. . :e3+ 45 @d2 (45 @d4

:e4+ 46 @c5 :es 47 ..tc3 :xd5+

48 :xd5 ..txd5 49 @xd5 �f4 50

Diagram 2 7 . 7
@e6 @ g 5 ! with a draw) 45 .. . :e2+
White to move
86 Endgame Secrets

46 @ e l tk2+ 47 @bl tk8 48 d6. Endgame 28

The d-pawn has crossed the critica} I.Sokolov -V.Salov

square d5, but after 48 . . . .tc2+ 49 Madrid 1994

@a 1 .tf5 50 .:n ñcl + 51 @b2 Earlier publication:

:c2+ 52 @bl .: e s + ! [ 2 7 . 8 ] lnformant 60/501 (I.Sokolov)

Diagram 2 7 . 8 Diagram 2 8 . 1 --.__

White to move White to move

. . . a perpetua} check results. White to move. How can he make

44 @g4 progress ?

45 .:d7 @f3

The remainder of the game is Commentary

without interest. Thanks to the ar-

rangement b.tb3 + ¿ja4 Black This endgame is an instructive ex­

defends. ample of the theme "The Two

46 .:n+ @g4 Weak.nesses". The black pieces have

47 .:f8 .:h6 to watch the ¿jb5. White exploits

48 @e4 ñe6+ this circumstance by attacking an­

49 @d3 ñh6 other weakness-the ¿jf7 !

so .tes ñh3+ 35 .:m Íf.b6

51 @e4 .:hs 36 @el nrs

s2 .:r4+ @g3 37 @e2 :bb8

53 @e3 ñh6 38 e4 .tes

54 .:n @g4 39 @e3 :b6

55 @e4 ñe6+ 40 ñdf2!

56 @d3 ñe8 Doubling the rooks along the

57 @e3 .:b8 closed f-file _looks strange, but only

ss .:n 1/2-1/2 at first sight.

40 .td7

41 g4! [ 2 8 . 2 ]
 Endgame Secrets 87

Diagram 2 8 . 2 Diagram 2 8 . 3

B lack to move Black to move

The decisive advance.

Endgame 29
41 hxg4
R.Fischer-J .Bolbochan
4 l...h4 42 g5 followed by :f3-f4
Mar del Plata 1959
and .l::thl (Sokolov).
Earlier publication: Encyclopaedia
42 fxg4 .1e6
of Endgames (Rooks II)/endgame
43 g5 .l::tbb8
1 5 3 4 ( editorial remarks)
44 .l::tf6! .1xd5

45 cxd5 �b7

46 @d3 �d7

47 @c4 �b7

48 .l::t l f3 �bb8

49 .l::th 3! �h8

Black has to meet the threat of

:h 7, but now the b-pawn decides

the game. 49 . . . : b 6 50 :h7 @e8 5 1

g3! , and due to zugzwang Black

loses a pawn (Sokolov).

50 :xh8 lixh8

51 b6 �hl

52 @b5 c4 Diagram 2 9 . 1

53 b7 White to m o v e

Certainly not 53 @xc4? :b 1 .

53 ... c3 There followed: 1 f4 g6 2 �f6

54 @a6 libl .1e8 3 .i.e6 .i.c6 4 g3 :g7 5 @e5

55 :n c2 .1e8 6 .i.d5 h5 7 :b6 @c8 8 .te6+

56 :c3 [28.3] 1-0 An immediate 8 @f6 wins more

88 Endgame Secrets

quickly. 8 .. . @c7 9 @f6 .:.h7 1 0 Jld5 4 . . . @c8? 5 @d6 .:.xc4 6 .:.f8+

@c8 11 .:.e6 @d8 1 2 .:.ct6+ @c7 13 loses .

.:.b6 @c8 14 Jlg8 .:.c7 15 Jle6+ 5 @d5

@b8 16 .:.ct6 1-0. In the course of Forced as 5 @d4? .:.xc4+ or 5

the game Black remains rather pas­ @d6? Jle8 6 .:.xb7 l:txc4 7 .:.b8+

sive. Therefore l.. . .:.e 1 ! ? looks like .:.es lose material.

a way to improve Black's play. The 5 .:.xc4

Encyclopaedia then gives: 2 .:.n The transition to a pawn ending

[29.2] " a n d White wins". Do you secures the draw. 5 ..tc6+ 6 @d4

agree? Jlxg2 7 .:.xg7 or 5 .:.d 1 + 6 @e5

.:.e 1+ 7 @d6 .:.d 1 + 8 Jld5 are

clearly favourable for White.

6 .:.xd7+ @xd7

7 @xc4 [ 2 9 . 3 ]

Diagram 29 .2

Black to move

Commentary

Diagram 2 9 . 3

This assessment is wrong as Black White to move

can force a draw.

1 f4 .:.el Black still has to be careful as

2 .:.f7 .:.d1+ White's king is more active.

This simple check poses 7 @c6!

problems. Other moves lose:

3 @c5 I) 7 . . . h5? 8 @d5 g6 9 @e5 @c6

3 @e5 .:.e 1 + and the king has to (9 . . . @e 7 10 g3) 10 @f6 @b5 11

retum. 3 @c4? @d6 loses material @xg6 @xb4 1 2 @xf5 @xa5 1 3 @e4

due to the threats .. . ..tb5+ respec­ b5 ( 13 . .. @b6 1 4 g4) 1 4 f5 b4 1 5 f6

tively .. . ..te6+ . b3 1 6 @ d 3 .

3 .•. .:.c1+ II) 7 . . . @d6? 8 h5 @c6 (8 . . . b6 9

4 Jlc4 axb6 @c6 1 0 @b3 @xb6 1 1 @a4) 9

White has to accept this pin. Oth­ b5+ axb5+ (9 . . . @d6 1 O bxa6 bxa6

erwise he can't go on. 1 1 @d4 @c6 1 2 @e5 @b5 1 3 �xf5

4 @d8! @xa5 1 4 @g6 results in a cheerless

 Endgame Secrets 89

queen ending) 10 @b4 @c7 11 Endgame30

@xb5 @b8 1 2 @b6 @c8 1 3 a6 bxa6 C.Lutz-R.Dautov

1 4 @xa6 @c7 1 5 @b5 @d6 1 6 @c4 German leag u e 1 9 9 7

@c6 17 @d4 @d6 18 g3 @e6 19 U n p ublished.

@c5 g6 20 @c6 (not 20 hxg6? @f6

21 @d6 @xg6 22 @e6 @h5 23

@xf5 stalemate!) 20 .. . @e7 21 @d5

@f6 22 @d6 @f7 23 @ e 5 .

8 h5

8 b5+? axb5+ 9 @b4 h5 1 O g3 g6

and this time White loses.

8 b6

8 . . . @d6? 9 b5, compare the line

7 . . . @d6.

9 axb6

9 g3 b5+ 10 @d4 @d6 11 @c3

@d5 12 @d3 @c6 draws

D iag ra m 3 0 .1
immediately.
White to move
9 @xb6

10 @c3!?
White p layed 43 nh6. How
Black has to get around the last
should Black react ?
problem. White wants to reach the

position w@a4 v b@b6 with Black

Commentary
to move. 1 O @b3 @b5 11 g3 _ @c6

12 @a4 @b6 is a draw as White I was in a p leasant m o od as the

l a ck s the r ese rv e tempo g 2 - g 3 . weak blac k pawns and the p assi v e

10 .. . @c6!? i.d7 guarantee White a d e fi nite

The sim p lest , but 10 . . . @ b5 1 1 edge.

@b3 @c6 12 @a4 @d5 (not 43 klh6

1 2 . . . @ b 6? 13 g3) 13 @a 6 @c4 14 Here I expected only 4 3 . . . i. e8 44

@ x a6 @xb4 eventually l e ads to @f3 and 45 @g4, 46 @g5 or

draw, too. 43 . . . .:. g s 44· i.d3 .:.g7 ( ot h e rwi se

11 @b3 @b5 the white rook enters via the e -fil e s )

12 g3 45 �el l:Ig8 46 @f3 _:g7 47 @ g4 !

Or 12 @a3 a5. and 48 @ g5 , even after 47 . . . e5 + . In

12 @c 6 both cases White wins the �g6 and

13 @a4 @b 6 the game. But Dautov co mes up

There is no pro g r ess p ossi b le. Al- with a su rpr ise :

tematives to 2 .:.n give Bl ack rea­ 43 �ff8!

son a ble counterplay as well, viz. 2 B lac k attains counterplay via the

i. f3 l:Ib 1 3 @ c3 :e 1 + 4 @ b3 i.e 6 + e - fil e.

and . . . l:I c4. Thus the acti v e 1.. . .:. e 1 44 lixg6

offers r easo n able chances for 44 i. x g 6? l:If6 with the threats of

de fen ce. . . . �g8 or . . . i. e 8 .

90 Endgame Secrets

44 . .. l!bc8 48 . . . l!a3 49 l!b7 l!xa4 50 l!xb6+

45 l!xc8 l!xc8 @e7 51 l!a6 is clearly to White's

[30.2] advantage.

48 @f3

48 @f2 i.e8 49 l!g7 l!a2 is less

convincing as White can't play 50

l!b7? i.h5. After 50 @e3 i.d7 51

g4 l!xa4 52 g5 l!a3+ 53 @f2 l!c3

Black stops the g-pawn while his

a-pawn creates trouble.

48 - ... i.e8

4 8 . . . l!a2 leads to a lost bishop

ending: 49 f5 l!xa4 50 f6 l!a3+ 51

@f4 l!c3 (or else Black can't pre­

vent l!g8 and f6-t7) 52 l!g8 l!c8 53

l!xc8 i.xc8 [ 3 0 . 3 ]
Diagram 3 0 . 2

White to move

Since 46 i. d l l!c4 leads to strong

counterplay White panicked and

forced a draw with 46 i.f5 (threat­

, ening 47 i.xe6 and 4 8 f5) @e7 47

l!g7+ @d6 4 8 l!g6 @ e7 . I was un­

der the influence of 43 . . . l! f f8!;

White has still a clear edge:

46 i. d 3 l!c3

Other moves d on't help:

I) 46 . . . l! c l 47 i.e2 is si m ilar to

46 . . . l! c 3 .

II) 46 . . . @e7 47 l!g7+ @d6 48 g4 Diagram 3 0 . 3

l!c3 49 i.e2 l! a3 50 g5 l!xa4 5 1 g6 , White to move

and the g-pawn nets a p i ece.

47 i.e2 l!c2 54 i.dl i.d7 55 i.a4 i.e8 56 g4

Or: i.t7 (56 . . . i.g6 57 @g5 and

I) 47 . . . l!a 3 48 f5 l!xa4 49 f6 ! (49 @h6-g7) 57 i.c2 (57 @g5? e5)

fxe6 i. e8 50 l!g 8 @ e7 a ll o w s coun­ 57 . . . i.e8 5 8 g5 i.xb5 59 g6 i.e8 60

te rp lay) 49 . . . l!a2 50 l! g8 l! xe2+ 5 1 g7 i.t7 6 1 i.h7.

@ f3 l! e l 52 @f2, and the f-pawn 49 l!g7 l!a2

p romotes. 50 :b7 i.h5+

II) 47 . . . @e7 48 l!g7+ @d6 4 9 g4 Or 50 . . . l!xa4 5 1 l!xb6+.

leads to the line 46 . . . @ e7. 51 g4 i.xg4+

III) 4 7 . . . i. e8 4 8 l!g7 ( or 48 l! g8 52 @xg4 l!xe2

@e7 49 l!g7+ @f6 50 l!b7) 53 :xb6+ [30.4]

 Endgame Secrets 91

After e.g. 43 .. Jlg8 44 í!i:h6 í!i:ff8

45 í!i:xg6 :c8 46 ilxc8 .Sxc8 47 f5

White is a whole tempo up com­

pared to the lines above and wins

easily.

Endgame 31

G .Kasparov-B. Gelfand

Horgen 1994

Earlier publication:

lnformant 61/261 (G.Kasparov)

Diagram 3 0 .4

Black to move

White has a clear edge. Most

probably Dautov didn't see all these

lines, but that isn 't necessary. lt suf­

fices to see that other moves lose

without a fight and that the compli­

cations ensuing from 43 .. Jiff8!

aren 't that easy to calculate. Indeed,

I immediately spoilt my advantage.

I could have spared myself all this

trouble if I had watched closely for Diagram 3 1 . 1

Bl a c k ' s possibilities. Then I would Black to move

have found the prophylactic move

43 i.d3 ! [ 3 0 . 5 ]. Analyse the game.

13 ... �4+ 14 �xb4 i.xb4+ 15

c3 i.f8 16 i.d3 i.e6 1 7 .Sfl 0-0-0

18 .tc2 CUe8 19 i.b3 ild7 2 0 i.g5

.tes 2 1 @e2 i.xb3 22 axb3 f6 23

.1lcl cud6 24 cuxd6+ :xd6 25 b4

i.a7 26 b5 ilhd8 27 ila2 @b8 28

bxa6 bxa6 29 b3 @b7 30 �d2 @c6

3 1 .Sxd6+ .Sxd6 32 g4 ild7 33 i.d2

.1Íb6 34 ilal @b7 35 h4 ild8 36 h5

.Sd6 37 i.el .1Íc7 38 b4 i.b6 39

ila2 .Sc6 40 .Sal l:íd6 4 1 i.d2 Jld7

42 xn @c6 43 c4 i.d4 44 h6 @b7

Diagram 3 O. 5 45 c5 a5 46 ilcl ilc7 4 7 c6+ 1-0
Black to m ove
92 Endgame Secrets

Commentary

Since he won the world champi­

onship in 1 9 8 5 Gany Kasparov has

dominated the chess world. At the

chessboard, besides his calculating

prowess, Kasparov's most outstand­

ing feature is his opening prepara­

tion. He made his mark on many

opening systems, many games he

won thanks to deadly novelties. But

in the present game the Byelorus­

Diagram 3 1 . 2
sian Boris Gelfand manages to sur­
Black to move
vive the opening against Kasparov

and to reach an equal ending. But ... with a slight white advantage

then Kasparov proves his superior­ (Kasparov). White enjoys the pair of

ity and by the accumulation of small bishops and his king is quite secure

advantages is able to achieve at e2, because Black can't open the

victory. e-file.

13 'iWb4+ 16 SLd3 ile6

Because of the pressure against f7 1 6 . . . g6!? 1 7 lue3 i.e6 1 8 c 4 ! ? is

Black forces the queens off. another move. White's knight has to

14 'iWxb4 i.xb4+ leave f5, but it can settle down at

15 c3 d5. Kasparov assesses the position

What are the most important fea­ as unclear. After 1 8 . . . 0-0-0 1 9 lud5

tures of the resulting endgame? The [ 3 1.3 ]

position of the L)e4 catches the eye:

On the one hand the pawn is iso­

lated and needs protection, on the

other hand it holds back Black's

kingside majority and supports the

luf5.

15 i.f8!

For the moment, Black lets the

ctJf5 live as White has to take care

of the L)e4. The altemative is

1 5 . . . i.xf5?! 1 6 exf5 Slc5 ( 1 6 . . Jk8

1 7 SLb3 lhc3? 1 8 bxc3 i.xc3+ 19

@e2 i.xal 20 i.a3 i.d4 21 l:t c l

Diagram 3 1.3
@d8 22 i..xf7 is clearly advanta­
Black to move
geous for White) 1 7 @e2 [ 3 1. 2 ]
Black can either erect a blockade

with 19 i.xd5 20 cxd5 lt:Je8 and

. . . f7-f6, i.e7, . . . lud6 or he can

 Endgame Secrets 93

play around the l°bd5 by 1 9 . .. l°bd7 l°bg3 (21 l'be3 l°bd6) l°bd6 doesn 't

20.0-0 i.c5+ 2 1 @hl h 6 ! ? . In both promise anything for White since

cases chances are about balanced. the �e4 is weaker than Black's dou­

11 .:n bled pawns and the tt:lg3 has to re­

To support the l°bf5. 1 7 i.g5 l°bd7 main passive.

18 i.c2 g6 1 9 l°be3 i.e7 20 i.xe7 20 i.g5

@xe7 is given by Kasparov as Not an obvious move, but its

slightly better for Black. Certainly sense becomes apparent on a second

the exchange of the black-squared look: The bishop prevents . . . l'bf6

bishop only helps Black. while the e3 square is left for the

17 0-0-0 knight. 20 .i.e3 l°bf6 doesn't pose

Though this move is not a mis­ any problems: 21 .i.xe6 (21 i.c2

take, it is rather "tame" _in the fight l'bg4 22 i.g 1 g6 and . . . .i.h6, 21

for the initiative. White realizes the l°bg3 .i.e7 and . . . h7-h5-h4) 2 1... fx e 6

exchange of the white-squared bish­ 22 l'bg3 .i.e7 followed by

ops, this way he can start to play for . . . h7-h5-h4 and possibly . . . l'bg4,

an advantage. 1 7 . . . g6 ! ? deserves at­ with strong counterplay.

tention here as well: 18 l'be3 (18 20 i.c5

i.g5 l°bh5 1 9 l'be3 h6 20 i.h4 i.e7 The black-squared bishop is ac­

looks good for Black) 1 8 . . . 0-0-0 19 tively placed. It doesn't attack any­

i.c2 (1 9 i.xa6?! l'bxe4 20 i.c4 thing, but it denies its white

i.xc4 2 1 l'bxc4 f5 is better for Black counterpart the e3 square.

owing to his central pawns. 22 20 . .. i.xb3?! 21 axb3 l°bd6 22

l'bxe5? .:es isn't playable) 1 9 .. . i.e7 l'bxd6+ i.xd6 23 .i.e3 ! [ 3 1 .4 ]

with an unclear position (Kasparov).

White should prefer 20 i.d2 fol­

lowed by 2 1 c4 and 22 l°bd5, with a

position similar to diagram 3 1 . 3 .

18 i.c2 l'be8

Black has to defend against the

threat .i.g5 and .i.xf6. The text in­

tends a later . . . l°bd6. After

1 8 . . . l'bd 7 ! ? (Kasparov) White

shouldn't play 19 i. b 3 ? i.xb3 20

axb3 l°bc5, but instead he should

hinder the development of the i.f8

with 1 9 .i.e3, e.g. 1 9 . . . g6 20 l°bh6 f6

Diagram 3 1.4
21 i.b3 i.xb3 22 axb3 .:e8 23 .:a 1
Black to move
@c7 (else l°bf7-d6+) 24 b4 and the

l°bh6 disturbs Black's play. . . . "with a clear white advantage"

19 $.b3 .:d7 according to Kasparov. Toe white

Black protects the �f7 and en­ bishop occupies the important di­

ables the knight to retum to f6 with­ agonal g l -a7. In analogy to the

out fear of i. g 5 . 20 i.xe6 fxe6 2 1 game White can play @e2, g2-g4,
94 Endgame Secrets

h2-h4-h5. Maybe it's an exaggera­

tion to say that White is clearly bet­

ter, but Black can't do anything

constructive. Even if he doubles his

rooks along the d-file White doesn't

have to react. The i..e3 controls the

d2 square and White isn 't forced to

exchange rooks.

21 @e2 i..xb3

22 axb3

After the exchange on b3 White

has the nominally better bishop. But

Diagram 3 1 . 5
for the moment the equilibrium isn't
White to move
disturbed.

22 fü 25 b4

Driving back the i.. g 5 . It turns out that the .tc5 lacks a

23 i.. c l ! ? square to retreat.

This move isn 't easy to under­ 25 i..a7

stand. After 23 ite3 i..xe3 24 ctJxe3 25 . . . i..b6? ! 26 b5 (Kasparov) with

ctJc7 White has given up the only the idea 26 .. . a5? 27 b4 axb4? 28

chance for an advantage, the better !ia8+ and the lih8 gets lost. Now

bishop. To win against a strong op­ you can understand the subtle point

ponent it is necessary to create prob­ of 23 i..c 1 : If White had cho sen 23

lems. 23 i.. c l is a test of Black's i..d2 instead, Black could play the

attention. It will be apparent soon intermediate move 26 . .. !ihd8 befare

why Kasparov didn't play 23 i..d2. taking on b 5 .

2 3 . .. étJ d 6 ? ! 26 b5

And Gelfand walks into the trap. After an exchange on a6 Black

lt ' s natural to exchange the ctJf5, but will be left with a weak �a6

White has a hidden resource. whereas White will get a passed

23 . . . ctJ c 7 ! 24 b4 i..b6 is given by pawn on the e-file. White has a

Kasparov. From e6 the knight con­ small but distinct advantage.

trols the important squares d4, f4, 26 lihd8

g5 and g7, thus the ctJf5 is neutral­ As compensation for his weak­

ized. After for instance 25 i..d2 ened pawn structure Black occupies

lihd8 26 !iadl étJe6 27 i.. c l !ixdl the d-file. The i..cl has to watch the

28 !ixdl !ixdl 29 @xdl @d7 the d2 square, therefore the rooks are

position is balanced. disconnected. How can White go on

24 ctJxd6+ .:xd6 [ 3 1 . 5 ] and make use of his structural

After 24 . .. i..xd6 25 i..e3 White advantages? ·

seizes the diagonal e3-a7, compare 27 �a2!

the note to diagram 3 1. 4 . Kasparov answers the question:

The text move prepares b2-b3 and

 Endgame Secrets 95

:d2. White has to exchange one Exchanging rooks mcreases

pair of rooks to lessen the pressure White's advantage.

along the d-file. Then White can put 30 @c6

his bishop to d2 and the remaining After 3 0 .. . h5 (to stop g2-g4)

rook might pressurize against the White can provoke a weakening at

lJa6 or the kingside. Furthermore the kingside by 3 1 l:.f3 and :g3 or

you should pay attention to the fact :h3.

that White doesn't hurry to take on 31 :xd6+ :xd6

a6 as Black hasn 't any constructive 32 g4

move. After 27 bxa6?! bxa6 2 8 :a2 Now Black's kingside majority is

Black might go 28 .. . @b7. tied down.

27 @b8 32 :d7

What e l s e ? 33 .i.d2 i..b6

28 bxa6 bxa6 34 :at @b7

29 b3 @b7? This move allows White to fix the

Ignoring the white threat. Black lJa6. Kasparov gives 34 .. . a5 35 h4

absolutely has to avoid an exchange "with a clear white advantage". As

of rooks. For that purpose 29 .. . :b6 in the game White proceeds with

30 :b2 a5 (Kasparov) is a reason­ h4-h5 and possibly h6. Black has

able choice. The :b2 is forced to the option of playing :b7, but this

fulfil defensive tasks. After 3 1 :dl is to no avail, e.g. 3 5 :b7 3 6 @d3

Black should not play 3 1 .. . :xdl 32 .i.c7 3 7 @c2 threatening 3 8 c4.

@xdl as then White can play @c2 35 h4 :d8

and :a2, attacking the lJa5. Better is 36 h5 :d6

3 1 .. . :c8, after which Black's 37 .i.el .i.c7

activity is enough to keep the 38 b4 .i.b6

balance. He should be able to ex­ 39 :a2

change the lJa5 for either the lJb3 or Here the transition to a bishop

lJ C 3 . ending has to be assessed: 39 :d1

30 :d2! [ 3 1. 6 ] :xdl 40 @xdl @c6 4 1 c4 [ 3 1. 7 ]

Diagram 3 1. 6 Diagram 3 1 . 7

B lack to move Black to move

96 Endgame Secrets

lf Black remains passive he loses: _:g7 �c8 54 .:xg6 (54 h7 is less

41....i.d4 42 @c2 .i.e3 43 @b3 .id4 convincing: 54 . . J�h8 5 5 @d3 l:íd8+

44 @a4 .ib6 (else 45 @a5 and 56 @c4 �c8+ 57 @b5 �b8+ 58

c4-c5-c6) 45 b5+ axb5 46 cxb5+ .i.b6 l:íh8 59 .ic5 .ixc5 60 @xc5

@b7 (46 . .. @c5 47 .if2+, 46 . . . @c7 [ 3 1. 8 ].

47 .i.a5) 47 .ia5 .i.c5 4 8 @b3 and

the white king enters the enemy

kingside. Therefore Black has to go

for activity as is shown by Kas­

parov: 41...@d7 42 @e2 @e6 43

.if2 .ic7 (43 . . . .ixf2? 44 @xf2 g6

45 h 6 ! f5 46 exf5+ gxf5 47 g5 and

White wins) 44 b5 axb5 45 cxb5 g6

46 s.a gxh5 47 gxh5 .id6! (other­

wise 4 8 b6) 48 b6 .i.c5 49 @d3 f5

50 @c4 .if2 5 1 @b5 fxe4 52 @a6

.ixb6 5 3 .i.xb6 h6 54 .i.e3 @f5 and

Black captures the last white pawn.

That' s why White doesn 't huny to Diagram 3 1 . 8

exchange rooks and seeks to im­ Black to move

prove his position instead. For in­

stance, if Black is forced to play Kasparov considers this position

. . . h7-h6 the transition to the bishop to be winning. But after 60 . . . g5 I

ending wins as Black then can't dis­ don't see how White makes pro­

solve the kingside. gress, e.g. 61 @c6 .:es+ 62 @b7

39 �c6 .:h8 63 @c7 l:ta8 64 l:!g8 �a7+)

40 l:tal �d6 54 . . . @f7 55 l;¡g7+ @f8. Kasparov

41 .id2 .a'.d7 ends this line with the assessment

42 l:tfl? "clear white advantage", but in my

Kasparov throws away a large opinion White even wins after 56

part of his advantage. He wants to .:h7, e . g . 56 . . Jk2+ (56 . . . .:ds 57 g5

play h5-h6, but he allows . . . a6-a5. fxg5 5 8 l:íxe7) 57 @d3 l:th2 5 8 @c4

The immediate 42 h6 is better. Kas­ (the white king enters the enemy

parov gives sorne lines which how­ camp) 5 8 . . . .ia3 (58 . . . .:h3 59 @d5

ever have to be supplemented: .:xe3 60 @e6 @g8 61 l:!xe7 �xe4

I) 42 . . . g6 43 l:tfl .id8 44 .ie3 62 l:íg7+ @h8 63 @xf6) 59 @d5

@c6 4 5 c4 ! (White can sacrifice his @g8 60 �g7+ @h8 6 1 .:a7 .ib4 62

queenside if he gets the �h7 in re­ g5 fxg5 63 .i.xg5 @g8 64 @e6 .if8

tum) 45 . . . :b7 46 S.dl ! .ie7 ( else @f6-g6) 65 :a8 and .ie7.

(46 . . . .ic7 47 .: a l �xb4 48 l:txa6+ Kasparov only considers 42 . . . g6,

@b7 49 .:xf6 .:xc4 50 S.f7) 47 .:al but

l:hb4 48 .:xa6+ @d7 49 �a7+ @d8 II) 42 . . . gxh6 deserves attention as

(49 . . . @e6? 50 �xe7+) 50 .:a8+ well. After this move Black's pawn

@d7 5 1 .:h8 @e6 52 �xh7 l:íxc4 53 structure is in ruins, but on the other
 Endgame Secrets 97

hand he eliminates the dangerous

h-pawn and makes . . . l:.g7 possible.

Owing to his "bad" i.b6 Black can

protect the M6 while the "good"

i.d2 can't support the pawns on e4

and g4. After 43 i.xh6 .:c7 44 i.d2

.:g7 45 @f3 .:d7 or 43 l:.fl i.d8 44

i..xh6 @c6 followed by . . . a6-a5 or

. . . @b5-c4 Black puts up a stiff resis­

tance. White still has to prove the

win.

42 @c6?
Diagram 3 1 . 9
Black ignores the chance. After
Black to move
42 . . . a5 ! 43 h6 axb4 44 cxb4 i.d8

(Kasparov) Black dissolves the Black can 't stop all the white

queenside under favourable circum­ pawns. Except for the mistake on

stances. He shouldn 't lose then. move 42 Kasparov played very

43 c4! forcefully, especially the moves 23

White attacks on both wings. It ' s i. c l ! ? and 27 .: a 2 ! are remarkable.

another drawback of 42 . . . @c6? that As a comparison the training

in sorne lines the cjf6 can be taken game Arnd Lauber - Mikhaljo

with check. Prusikhin is interesting. The game

43 i.d4 started at position 31.5 with one

Allowing a forced win, but alter­ hour per player: 25 b4 i.a7 2 6 . b5

natives aren 't really better: h 5 ? ! (Black stops g2-g4 but refrains

I) 43 . . . .:d4 44 h6 .:xe4+ from occupying the d-file) 27 .:dl ? !

(44 .. . gxh6 45 .:xf6+ @b7 46 .:xb6+ (Better is 27 bxa6 bxa6 and only

and i.e3) 45 @f3 (Kasparov) and then 28 .:dl as will become appar­

hxg7. ent) 27 . . . .:xdl 28 @xdl .:.ds+ 29

11) 43 . . . h6 44 .: a l .:d4 (44 . .. @b7 @e2 .a'.d6 3 0 bxa6 [ 3 1 . 1 O]

45 i..c3 followed by l:.dl is hope­

less as the kingside is fixed) 45 c5 !

i.c7 46 .:xa6+ @b7 47 b5 .:xe4+

48 @f3 .:d4 49 c6+ @b8 50 b6

ltxd2 51 .:as+! @xa8 52 bxc7

(Kasparov), and the pawn promotes.

44 h6 @b7

45 c5! a5

46 lt c l ! l:.c7

Or 46 . . . @c6 47 b5+! @xb5 48 c6

ltc7 49 hxg7 (Kasparov).

47 c6+ [ 3 1. 9 ] 1-0

Diagram 3 1 . 1 0

Black to move
98 Endgame Secrets

30 . . . bxa6?! (Here 3 0 . . Jha6 31 Black to move has a slight edge

l:.xa6 bxa6 is the correct choice. (weak squares d4 and b4, better

Black quickly plays . . . @d7-e6 fol­ king). How does he continue?

lowed by . . . g7-g6 and . . . f6-f5. He

shouldn't lose the bishop ending Commentary

then. As Kasparov-Gelfand showed

i t ' s desirable for White to keep pre­ 23 g5!

cisely one pair of rooks on the board For a brief moment I considered

while Black prefers to exchange ei­ 23 .. .Ct:lb4 to play against the weak­

ther all rooks or no rooks) 3 1 i..d2 ened queenside after 24 a3?! lt:lc6.

@b7 3 2 l:. fl ? ! (White plans l:.f3-h3, But after 24 @fl Black has no threat

but better is 32 b4 to fix the �a6) (24 . .. l:.xdl 25 l:.xdl lt:lxa2? 26

3 2 . . . a5 3 3 l:. f3 a4 34 l:.h3 g6 3 5 l:.g3 l:.d7+). Thus I carne to the conclu­

f5? (Loses a pawn. 3 5 . . . g5 is better, sion that I have to push my majority

even though the f5 square is weak­ on the kingside. 23 . . . e 5 ? ! 24 ctJc2

ened) 3 6 exf5 gxf5 37 l:.g5 l:.f6 3 8 only weakens the d5 square,

l:.xh5 @c6 39 l:.h4? (A strange 23 . . . h 5 ? ! or 23 . . . f5 allow 24 h4.

oversight. White saw that 39 l:.h6 That' s why the g-pawn has to

wins but wanted to provoke the ad­ advance.

vanee of a pawn first) 3 9 . . . f4 (But 24 @fl h5

now the diagonal cl-h6 is closed. 25 ctJc2 f5

-Black gets heavy counterplay) 40 26 @e2 @f6

l:.h 7 i..c5 4 1 l:.h5 e4 42 l:.e5 f3+ 43 27 l:.xd8 l:.xd8

gxf3 exf3 44 @fl l:.d6 45 i..e3 28 l:.dl l:.xdl

i..xe3 46 l:.xe3 l:rd2 47 l:.e4 1/2-1/2. 29 @xdl @e5

29 . . . h 4 ! ? has to be considered as

Endgame 32 well.

L.Keitlinghaus-C.Lutz 30 @d2 ctJd4

German championship 1 9 9 7 31 ctJel @d6

Unpublished. 32 b4 e5 [32.2]

Diagram 3 2 . 1 Diagram 3 2 . 2

Black to move White to move

 Endgame Secrets 99

Black has the better chances be­ Now the difference is obvious:

cause of his advantage in space. The white knight can support the

Nevertheless the game ended in a defence while the black knight has

draw. no active purpose.

36 c4

Endgame 33 t/2-t/2

G.Kamsky-M.Illescas Cordoba After the exchange of the weak

Las Palmas 1 9 9 4 b-pawn neither side can win.

Earlier publication: lnformant 60/73

(M.Illescas Cordoba) Endgame34

V.Kramnik-C.Lutz

German league 1 9 9 4

Earlier publications:

lnformant 60/444 (V.Kramnik),

Practica! Endgame Play

(N .McDonald)

Diagram 3 3 . 1

White to move

The �b2 is under attack. How

should White react?

Commentary
Diagram 3 4 . 1

White to m o v e
34 lt:i d l !

lt is important to keep the knight.

Analyse the game.
The black knight can be driven back

with b2-b3. 34 ctJxa4? �xa4 leads to 27 'iVd4 lt:icS 28 'iVxd7 ctJxd7 29

a clear black advantage (Illescas .a'.dl lt:icS 30 g4 g6 3 1 @g2 rJi;g7 32

Cordoba). White is tied to the de­ .a'.d2 a6 33 :d6 :es 34 ctJd4 bS 35

fence of the pawns on b2 and e4. h4 b4 36 .a'.b6 aS 37 kib5 ctJd3 38

Black can start an attack with either .a'.xa5 lt:iel + 39 rJi;g3 ctJc2 40 lt:ib3

. .. rJi; g 7 and . . . g6-g5 or . . J�[d4 and ltJa3 41 :a4 :c4 42 ctJd4 ltJc2 43

. . . f7-f5. 34 b3? ctJxc3 leads to the ltJf3 :es 44 .a'.a7 g5 45 h5 rJi;gS 46

same problems for White. ltJd2 ltJa3 47 lt:ie4 :c2 48 :b7

34 @g7 :xa2 49 :xb4 :c2 50 �b6 rJi;h7 5 1

35 b3 lt:ib6 :b7 rJi;gS 52 lt:id6 :c6 53 ltJxti

36 lt:ic3 ltJc4 54 :d7 :r6 55 :d4 1-0

100 Endgame Secrets

Commentary 28 'iUxd7 ét:Jxd7

29 l':.dl ( 3 4 . 2 ]

In the present game I had the

black pieces against Vladimir

Kramnik. I managed to survive both

the opening as well as the middle­

game without serious damage. In

the diagrammed position White has

no structural advantage, he only has

an active knight on c6 to the good.

And even this advantage will be

gone if Black comes to play .. . ctJ c 5 .

However, Kramnik keeps the initia­

tive alive and gradually increases

his advantage.
Diagram 34.2
27 'iHd4
B lack to move
The only move to create prob­

lems, otherwise Black plays . . . ctJ c 5 . The assets of White's position be­

27 ctJc5? come plainly visible: While the rook

Black allows an exchange of dominates the d-file, the knight re­

queens as he underestimated the sembles a giant octopus: it reaches

dangers of the resulting endgame. for the squares a7, b8, d8, e7 and

He has two other possibilities: ( owing to the fork on e7) to c8.

I) 27 . . . 'iHxd4?! 28 exd4 "and When I played 27 .. . ctJc5 I expected

White has the initiative" (Kramnik). to drive away the knight without too

Indeed, thanks to his better pieces many problems. But I was wrong.

White enjoys a small advantage, e . g . 29 ct:Jc5

2 8 .. . ctJ d 8 ! ? (28 . . . ét:J d 6 ! ? with the On c5 the black knight has a sta­

idea of 29 ctJe5 :c8) 29 ctJb4 (29 ble post and might tum to e4 or a4.

ctJe7+ @f8 30 .U. c 8 ? lhc8 3 1 ét:Jxc8 The drawback is that the e-file is

ctJc6 32 d5 ét:Jb4 is only good for closed for Black's rook. 29 . .. ctJ f6 ! ?

Black) 29 . .. ctJe6 30 d5 ctJc5 3 1 ét:Jc6 deserves attention as well:

(31 d6 :d8) 3 1... @ f8 32 g4!? (32 I) During the game I considered

:el a6 33 :e7? :d8! 34 :xf7+ 30 f3 intending e3-e4-e5 to be the

@xf7 35 ctJxd8+ rJiie7 36 ctJc6+ strongest move. But Black can de­

@d6) 3 2 . . . a6 3 3 @ g 2 ! ? followed by fend: 3 0 . . . @f8 3 1 e4

@ f3 . All white pieces are more ac­ A) 3 1... a 6 32 e5 ctJg8 (or

tive than their black counterparts. 32 . . . ét:Je8 33 l:d7) 33 :d6 b5 34

II) The correct choice is to avoid @f2 ctJe7 35 ctJd4 gives White a

an exchange of queens by means of clear advantage due to his active

27 . . . 'iY e 6 ! (Kramnik). After 28 'iUc4 pieces.

'1i'e8 there follows .. . ctJc5 and the B) 3 1... : c 8 ! (Black has to strive

ctJc6 has to retreat. Then White isn't for activity) 32 ctJxa7 :c2! (With

better at all. the threat . . . tuh5-f4. 32 . . . :a8? 33

 Endgame Secrets 101

ét:Jc6 lha2 34 k!d8+ ctJe8 35 :b8 30 g6!?

k!b2 36 e5 gives White a clear ad­ Both Kramnik and McDonald

vantage) 33 a4 ! ? ét:Jh5 34 :n ét:Jf4 give a question mark to this move

35 k!f2 k!xf2 36 @xf2 ctJd3+ 37 since it weakens the kingside. After

@e3 ébc5 followed by . . . ét:Jxa4 and a possible h4-h5 the pawns on g6

Black has no more problems. and h6 might become vulnerable. In

II) 3 0 g4! is more accurate. White my opinion the commentators over­

activates his king via f3 . 30 . .. @f8 estimate the strategical drawbacks

31 @g2 of 3 0 . . . g6. With the given material

A) 3 1... k! c 8 32 ébxa7 :c2 (or balance (k!+ét:J v k!+�) activity and

32 . . . k!a8 33 ét:Jc6 k!xa2 34 k!d8+ co-ordination of the pieces are at

ét:Je8 3 5 k!b8 k!b2 3 6 e4 f6 3 7 ctJd4) least as important as strategical fea­

33 @f3 ! k!xa2 34 k!d8.+ ébe8 ( or tures. Therefore 3 0 . . . g 6 ! ? shouldn't

34 .. . @e7 35 ét:Jc6+ @e6 36 ét:Jd4+ be criticized: It prepares . . . @g7 to

@e7 37 k!b8) 35 ctJb5 and Black's avoid the fork ébe7+. The real error

pieces are completely tied down. occurs sorne moves later. Let's look

B) 3 1... a 6 ! ? 3 2 k!d6 h 5 ! ? (32 . . . b5 at the altematives :

33 @f3) 33 @f3 (33 g5? ctJe4 34 I) 3 0 .. . k!c8?? 31 éi'Je7+ is out of

k!d5 k!c8 and . . . ét:Jc3) 3 3 . . . hxg4+ 34 question. The same goes for

hxg4. Black has problems in view II) 3 0 . . . b5? 3 1 k!d5 (McDonald).

of g4-g5. III) 30 . . . a5 (intending 31 k!b 1

30 g4! [ 3 4 . 3 ] éba4) is suggested by McDonald.

White kills two birds with one But after 3 1 @ g 2 ! ? @f8 32 h4 k!c8

stone: On the one hand he gains 3 3 éba7 and ét:Jb5 it isn't clear what

space on the kingside before Black Black attained apart from weaken­

sets up his pawns at f7, g6 and h 5 . ing his queenside.

On the other hand the @g 1 can be IV) 30 . . . @ f8 ! ? "with a slight

centralized via g2 and f3 . White has white advantage" according to

a slight but distinct initiative. Kramnik. McDonald takes a closer

look at the position and comes to

the conclusion that Black still has

problems: 3 1 @g2

A) 3 1... k! c 8 ? 32 ctJxa7 k!a8 33

ébc6 k!xa2? 34 k!d8# (McDonald)

B) 3 1... @ e 8 32 h4 ét:Je6 (32 k! c 8 ?

33 ét:Jxa7) 33 h5 ét:Jd8 (or 33 :cs

34 k!d6 k!c7 3 5 f4 with the idea of

f4-f5) 34 ét:Jd4 and ét:JfS, with a ele ar

edge according to McDonald.

C) 3 1... a 6 looks better. 32 lid6

(32 @f3 k!c8 3 3 :d6? étJb7) 3 2 .. . b5 !

(planning . . . ét:Je4) 33 @f3 (33 ét:Jd4

Diagram 3 4 . 3 b4 and . . . a6-a5, .. . ét:Ja4-c3) 3 3 . . . ctJb7

Black to move 34 k!d7 ét:Jc5 3 5 lld5 (35 :c7 ctJe6

102 Endgame Secrets

36 l:.b7 'Dc5 37 l:.e7? l:.c8) White protects f2, but after

3 5 . .. 'Da4! 36 l:.d6 (36 a3 ctJb6 37 34 . . . @ f6 ! [ 3 4 .4 ]

l:.d6 'Dc4, 3 6 'De5 @e7! 37 l:.d7+?

@e6 38 @f4 g5+ 39 @e4 'Dc5+)

3 6 . . . 'Dc3 37 a3 CDbl 38 l:.d3 l:.c8

with counterplay. Perhaps White

should prefer 34 l:.d5 l:.c8 3 5 ctJd4

with an initiative, he might play a

later h3-h4-h5 and 'Df5.

31 @g2 @g7

Now . . . l:.c8 is a threat.

32 l:. d 2 !

Kramnik played this rather

quickly, obviously he saw through

Bl a c k ' s intentions. The text protects

the Lsa2, for that reason 32 . . . l:. c 8 ? Diagram 34.4

isn't feasible. Prophylactic moves White to move

like 3 2 l:.d2! can be highly unpleas­

ant as they disturb the flow of the Black is OK:

opponent's play. Indeed, over the A) 3 5 'Dxf7 l:.c2 3 6 @f3 ctJd2+ 3 7

next three moves Black misses his @g3 (37 @f4 g5+, 3 7 @e2 'De4+ 3 8

chances to equalize the game. For @d3? l:.d2+) 37 . . . 'De 4 + with a per­

that he should have ventured on tac­ petua} check as 38 @f4? 'Dc5 39

tical complications. l:.xa7 l:.xf2+ 40 Wg3 'De4+ 4 1 @h4

32 a6 g5+ 42 @h5 l:.h2 even wins for

Black renews the threat of . . . l:.c8, Black.

but softens up his queenside. B) 35 f4 l:.c2+ 36 @f3 (36 @fl

3 2 . . . 'De4! is more active: 33 l:.d7 l:.xa2 cuts off the white king)

(Not 3 3 l:.d4 when the black knight 3 6 . . . ctJd2+ 37 @g3 (37 We2 'Dc4+

may tum to c3, c5 or f6 . Kramnik 38 @d3 'Dxe5+ 39 fxe5+ @e6! 40

thinks the white position is clearly l:.d6+ @xe5) 3 7 . . . 'Dfl+. Once again

advantageous in view of the weak White has to agree to a draw be­

Ls f7 . But Black has a hidden tactical cause 3 8 @h4 'Dxe3 39 l:.xf7+ @e6

solution) 3 3 . . . l:. c 8 ! 40 l:.xa7 l:.c3 i s n ' t worth it. Tactical

I) 34 'Dxa7 l:.c2 or tums like this are quite typical for

11) 34 ctJd4 a6 followed by . . . 'Dc3 the combination of rook and knight.

don't pose any problems. The white pieces seem to cooperate

III) 34 'Dd8 leads to a repetition well, but after the surprising

of moves after: 34 . . . 'Dc5 3 5 l:.xf7+ 34 . . . @ f6 ! the black pieces are no

@g8 36 étJb7! l:.b8 ( 3 6 . . . @xf?? 37 less effective.

étJd6+) 3 7 étJd6 l:.d8 3 8 l:. f6 @g7 39 33 l:.d6

l:.f7+ W g 8 . White points his rook to the en­

IV) 34 'De5 is the critica} con­ emy queenside.

tinuation. After 34 . . . l:.c2? 3 5 l:.xf7+ 33 l:.c8

 Endgame Secrets 103

Black wants to get rid of the é'.bc6,

but the knight intended to retum to

d4 in any case. There were two pos­

sibilities to get a reasonable

position:

I) 33 .. . b 5 ! ? (pushing the queen­

side majority)

A) 34 h4 (allowing B l a c k ' s knight

to harass White's rook) 34 . . . é'.be4 3 5

1id7 (35 lid4 é'.bc3 and . . . lic8)

3 5 . . . é'.bf6 36 lid4 lic8 with

counterplay.
Diagram 3 4 . 5
B) 34 1id5 é'.be4 (threatening
White to move
.. . ctJc3) 35 1id7 é'.bc5 36 lic7 (36

lie7 1ic8 37 é'.be5? @f6) 36 . .. ctJe6 3 7

. .. followed by .. . a4-a3 and/or
li b 7 (37 lie7 lic8) 3 7 . . . é'.b c 5 . Once
. . . b4 - b 3 .
again the knight chases the rook.

C) 34 @f3 é'.bb7 3 5 lid5 (35 1id7 If you watch closely the lines

é'.bc5 36 lid5 CiJ a 4 ! ? ) 3 5 . . . 1ic8 36 starting with 32 . . . é'.be4, you will cer­

CiJd4!? (36 é'.be5 li c 5 ! ? with the idea tainly find

of 3 7 @e4 lic2) 3 6 . . . li c 4 ! ? followed II) 3 3 . . . é'.be4!. After 34 1id7 lic8

by . . . lia4. we reach the same position as after

D) 34 CiJd4. We reach the same 32 . . . é'.be4 3 3 1id7 lic8, but with the

position as in the game without a-pawn on a6. This difference is un­

Black having played . . . J:k8. This is important, Black keeps the balance.

to his advantage since he can make 34 CiJd4 b5?

better use of this tempo. On a8 the Now Black's pieces are com­

rook supports the queenside pawns. pletely misplaced: The 1ic8 is

After 34 . . . b4 Black creates active blocked by the é'.bc5 which can't

play, e.g. 35 h4 (35 lib6 a5 and move in view of the vulnerable �a6 .

. . . CiJa4-c3) 3 5 . . . a5 36 h5 gxh5 37 White has a great advantage. The

gxh5 a4 38 CiJf5+ @g8 39 é'.bxh6+ active 34 . . . é'.ba4 is better, by plan­

@f8 [ 3 4 . 5 ] ning . . . lic3-a3xa2 Black seeks to

justify 3 3 . . . li c 8 . This manoeuvre

not only nets a pawn, Black also

creates threats against f2 . Sorne ex­

emplary línes: 3 5 h4 lic3 36 h5 lia3

3 7 hxg6 fxg6 3 8 é'.be6+ ( 3 8 e4 lixa2

104 Endgame Secrets

3 9 e5 CLic5 40 .a'.xb6 CLid3) 3 8 . . . @ f7 good defensive chances with rooks

39 CLif4 g5 40 CLid5 .a'.xa2 4 1 .a'.xh6 only. But the presence of knights

(4 1 CLixb6 CLixb6 42 .a'.xb6 @g7 is a (with or without rooks) makes the

draw) 4 1... CLi c 5 42 .a'.xb6 CLle4 and defence of such an ending highly

Black is out of trouble. difficult.

35 h4! [ 3 4 . 6 ] I) 3 7 . . . CLie6 immediately forces

the knights off, but after 3 8 CLlxe6+

fxe6 3 9 .a'.xa5 White wins d u e t o the

shattered black kingside. As a point

of reference I give the course of the

well-known rook ending

A.Karpov-V .Hort (Waddinxveen

1 9 7 9 ) [ 3 4 . 7 ]:

Diagram 3 4 . 6

B lack to move

White is ready to play h4-h5. The

white rook is ideally placed along

the sixth rank, it attacks both queen­

side (i0ia6) and kingside (i0ig6, i0ih6). Diagram 3 4 . 7

35 b4 White to move
Black loses a pawn, but there is

no real altemative: Black loses though he isn't even a

I) 3 5 . . . a5 36 CLixb5 .a'.b8 37 CLic3 pawn down: 37 h5 g5 3 8 .a'.a6 gxf4

.a'.b2 38 .a'.d5 (Kramnik) loses a 3 9 exf4 .a'.b3+ 40 @g2 .a'.b7 4 1 @g3

pawn under even less favourable @f7 42 .a'.a4 @g7 43 g5 .a'.c7 44 .a'.a5

circumstances. @g8 45 �b5 @f7 46 @g4 a6 47

II) 3 5 . . . .a'.c7 3 6 h 5 @h7 37 f3 fol­ .a'.b8 .a'.c 1 48 g6+ @g7 49 .a'.b7+ @f8

lowed by @g3, hxg6 and CLie2-f4 50 .a'.b6 .a'. g l + 5 1 @f3 .a'.fl+ 52 @e4

lets White execute his kingside at­ .a'. e l+ 5 3 @d4 @e7 54 .a'.xa6 @f6 5 5

tack without any disruption. .a'.a7 e5+ 56 fxe5+ .a'.xe5 57 .a'.a6+

36 .a'.b6 as 1 - 0 . A detailed analysis of this end­

37 .a'.bS CLid3 ing can be found in Jan Timman's

Bl a c k ' s knight enters the white The Art of Chess Analysis.

camp to create sorne confusion. II) 37 . . . b 3 ! ? is another try. 38

Black aims to liquidate the knights CLlxb3 (38 axb3 a4 39 bxa4 lt:Jxa4

and the queenside pawns. An ending lets Black liquidate the queenside,

"4i0i v 3i0i on the kingside" offers but things remain difficult after 40
 Endgame Secrets 105

h5 étJc5 41 l:Ib6) 3 8 .. .lbxb3 39 The final mistake. Black stops

:xb3 (Not 3 9 axb3? :as followed g4-g5, but by doing so he weakens

by . .. a5-a4 and Black has serious himself even further. After 44 . . . lba 3

drawing chances) 3 9 . . . l:Ic4 40 @g3. I feared 45 g5 h5 46 ctJd2 [ 3 4 . 8 ].

The presence of the a-pawns renders

Black's position cheerless. After

suitable preparation White's king

will advance to the queenside.

38 :xa5 lbel +

39 @g3 lbc2

40 CD b 3 ? !

A slight inaccuracy. Obviously

White avoids an exchange, but bet­

ter is 40 ébf3 !. White plans :a6 fol­

lowed by h4-h5 and lbh4, have a

look: 40 . . . :c3 (threatening :a3)

41 ctJe5 ! ( attacking f7) 41 @f6

Diagram 3 4 . 8
( 4 L.:a3 42 :xa3 bxa3 43 ébd3 ! is
Black to move
given by Kramnik. White's king

will win the stray lb c2 . Aft er . . . 'tJa 1 White threatens mate with CDe4-f6

White has lbb4) 42 @f4 g5+ 43 and :as. McDonald now analyses

hxg5+ hxg5+ 44 @e4 (Kramnik). I) 46 . . . :es 47 :b7 and

Black is without a defence against II) 46 . . . :c2 47 ctJe4 @f8

the double threat of l:Ia7 and :a6+, (47 . . . :xa2 48 ébf6 b3 49 l:Ia8) 48

@f5. :b7 :xa2 49 :xb4 lbc2 50 :b7

40 lba3 followed by ébd6. White has a deci­

41 :a4 :c4 sive edge in both cases.

4 1... : c 2 ? 42 :xb4 :xa2 43 l:Ia4 III) 46 . . . lbb5 is strongest. 47 :as

leads to a deadly pin. (after 47 :d7 lbc3 or 47 l:Ib7 lbc3

42 ltJ d 4 ! 48 :xb4 CDxa2 49 l':tb7 CDc3

After sorne thought Kramnik cor­ White's knight doesn't get to e4)

rects his mistake. The knight 1s 47 . . . :c2 (47'. . . :ds 48 l'be4 étJc3 49

needed at the kingside. :xd5 ébxd5 50 @f3 and the white

42 lbc2 king marches to c4) 48 CDe4 (48

43 ltJf3 :es :xb5 :xd2 49 :xb4 :xa2 isn't

After 43 . . . :c7 there follows 44 convincing either) 48 . . . lbc3 with

:a6 and h4-h5, lbh4. 43 . . .lk3 will counterplay as 49 ctJf6? llxa2 is im­

be answered by 44 lbe5 ! , in analogy possible. That's why White should

to the line 40 ctJf3 :c3 41 lbe5. lt play differently, e.g. 46. ctJd4 @f8

doesn 't make a difference that 47 :b7. He keeps a large advantage,

White's rook is on a4. but he hasn 't won yet.

44 :a7 g5? 45 h5 [34.9]

1 0 6 Endgame Secrets

Diagram 3 4 . 9 Diagram 3 4 . 1 O

Black to move Black to move

Now the knight retums into play

In view of the awkward weakness
and White remains two pawns up.
on f5 the game is decided.
After 55 . . . Q)xe3 there follows 56
45 ... @g8
tbxh6+.
46 ibd2 tba3

47 tbe4 :c2
Endgame 35
Or 47 . . . �c6 4 8 :b7.
C.Lutz-R.Hübner
48 :h7 :xa2
Gerrnanleague 1 9 9 4
49 �xb4
Unpublished.
Black has managed to exchange

the queenside pawns, but his king­

side lacks protection.

49 :c2

50 b6!

Another subtlety. White occupies

the seventh rank with gain of time.

50 :b7? tbc4 is weaker.

50 ... @h7

50 . . . @g7 5 1 ibd6.

51 :h7

5 1 tbd6? tbc4 52 :c6 tbxe3 isn't

worth looking at.

51 ... @g8 Diagram 3 5 . 1

52 ibd6 :c6 White to move

53 tbxf7 tbc4

54 �d7 :r6 Analyse the game.

55 11d4! [ 3 4 . 10 ] 1-0
30 'i-Vxd7 éuxd7 3 1 �c7 tbf8 32

h3 b5 33 �a7 a4 34 :b7 :es 35

:xb5 tbg6 36 g3 Ci:Je7 3 7 :b7 ibf5

38 éuh2 ibd6 39 :d7 tbe4 40 tbg4

 Endgame Secrets 107

iícl+ 41 @g2 iíc2 42 :b7 fS 43 31 iíc7 'Df8

iíb8+ @g7 44 :b7+ @f8 45 étJeS 32 h3?!

é2:Jxf2 46 é2:Jg6+ @e8 4 7 h4 é2:Je4+ 48 A rook on an open seventh rank

@gl é2:Jxg3 49 hxgS é2:Je4 50 iíb8+ usually combines very well with a

@d7 5 1 é2:Jf8+ @c7 52 g6 @xb8 53 knight. Just think of the well-known

g7 é2:Jf6 54 é2:Jd7+ @b7 55 é2:Jxf6 drawing scheme wiíd7 + é2:Jf6 v

:es 56 gS='iV iíxg8+ 57 ltJxgS @c6 b@f8: White gives perpetua! check

58 é2:Jf6 @bS 59 é2:Jd7 @c4 60 é2:Jb6+ by 1 é2:Jh7+ @e8 2 ébf6+ @f8 3

@xd4 61 t2:Jxa4 @c4 62 b3+ @b4 é2:Jh7+ @g8 4 tbf6+, since 4 . .. @h8?

63 @f2 d4 64 @e2 @xb3 65 étJcS+ 5 :h7# leads to mate. When playing

@c2 66 é2:Jd3 @b3 67 étJcS+ @c2 68 the text move I had ideas like this in

é2:Jd3 @c3 69 étJeS f4 70 @dl d3 7 1 mind. Via h2 and g4 the knight

é2:Jf3 d2 72 é2:Jxd2 @d3 73 é2:Jf3 @e3 heads for f6 . But in positions with

74 étJeS @e4 75 é2:Jg4 @f3 76 étJeS+ rooks and knights one has to watch

@f2 77 @d2 f3 78 é2:Jd3+ @g3 79 closely which threats are for real

@e3 es 80 é2:Jf2 @g2 t/2-t/2 and which are not. One moment the

pieces might cooperate nicely, the

Commentary
next moment they are spread hap­

This encounter originates from a hazardly all over the board. 32

team match between SG Porz and iíxb7 regains the pawn. After

Bayem Munich. In the first half of 32 . . . iíc8 33 @fl é2:Jg6 34 g3 fol­

the 1990s Bayem Munich was the lowed by @e2 and maybe é2:Jd2-b3

dominating team in Germany while White' s structural disadvantages

my club, SG Porz, usually had to (weak pawns on d4 and h2) are

play the part of the "eternal sec­ counterbalanced by his activity. But

ond", The direct clash of the two in the back of my head there was the

tea ms usually was a tough fight and famous game Em.Lasker­

usually Munich won. But this time J.R.Capablanca (World champion­

things were different... ship La Habana 1921, 10th game)

In the present position White is a [ 3 5 . 2 ]:

pawn down, but he can regain it im­

mediately. The game is about bal­

anced, but I spoilt everything.

30 'iVxd7

After 30 'iVxa5 iíc8 31 'iVd2 the

game should e n d i n a draw, but dur­

ing play I was worried about my

pawns. The �d4 is isolated and

Black has more space at the king­

side, that's the reason why I headed

for activity. But it's only this mis­

guided endeavour that creates real

problems.
Diagram 3 5 . 2
30 é2:Jxd7
Black to move
1 0 8 Endgame Secrets

Black won by instructive play: 36 .:xe7+ (36 lt:Jd7 a4 37 @fl b4

3 5 . . . .:b6 36 .:d3 .:a6 37 g4 hxg3 3 8 followed by . . . a4-a3 leaves White in

fxg3 .:a2 3 9 t2Jc3 .:c2 40 tZJ d l t2Je7 a deadly pin). Black is clearly better

41 t2Je3 : e l+ 42 @f2 t2Jc6 43 tZJdl in this knight ending (weak l>id4,

.:b 1 44 @e2 .:xb3 45 @e3 .:b4 46 dangerous passed pawn at the

t2Jc3 t2Je7 47 t2Je2 t2Jf5+ 48 @f2 g5 queenside), e.g. 36 . . . lt:J x e 7 ! ? 37

49 g4 t2Jd6 50 t2J g l lt:Je4+ 51 @ fl lt:Jh7 (or else White loses the l>id4)

.:b 1 + 52 @g2 .:b2+ 53 @fl .:f2+ 37 . . . a4 (Black has to give a pawn if

54 @el .:a2 55 @ fl @g7 56 .:e3 he wants to make progress, viz.

@g6 5 7 .:d3 f6 5 8 .:e3 @f7 59 .:d3 37 . . . @g6 3 8 lt:Jf8+ @f5 3 9 g4+ @f6

@e7 60 .:e3 @d6 6 1 .:d3 .:t2+ 62 40 t2Jh7+ <it>g6 4 1 lt:Jf8+) 3 8 t2Jxg5+

@ e l .:g2 63 @fl .:a2 64 .:e3 e5 65 @f6 3 9 étJf3 b4 40 @fl a3 4 1 bxa3

.:d3 exd4 66 .:xd4 @c5 67 .:d 1 d4 bxa3 42 t2Jd2 a2 43 lt:Jb3 lt:Jc6 44

68 .:e 1 + @d5 0-1. This game has lt:Jal lt:Jxd4, and it's questionable

been commentated by several whether White can draw. When I

authors (e . g . by Shereshevsky in saw 34 . . . .: e 7 ! I immediately dis­

Endgame Strategy or by Speelman missed 3 3 étJh2. In a practica! game

in Endgame Preparation ). Because one shouldn't really bother to calcu­

of this example I panicked. late such a line to the end. It ' s

32 b5 enough to understand that the knight

The doomed pawn comes back to ending is highly dangerous. The text

live again. 32 . . . .:b8 33 lt:Jh2 is should keep the equilibrium.

rather passive, for instance 3 3 . . . f5 33 a4

34 t2Jf3 lt:Jh7 3 5 .:e7. 34 .:b7

33 .:a7 Now 34 étJh2? lt:Jg6 3 5 t2Jg4 .: e 7 !

Only here I realized that my origi­ is even less desirable.

nal plan simply doesn't work: 33 34 .:es

lt:Jh2 lt:Jg6 34 t2Jg4 .: e 7 ! 35 lt:Jxf6+ 35 .:xb5

(35 .:es f5 and . . . .:b7) 3 5 . . . @ fl Finally White regains the pawn,

[35.3] but Black takes over the initiative.

35 lt:Jg6!

This knight is going for a long

walk.

36 g3

Against . . . t2Jf4.

36 lt:Je7

[ 3 5 .4 ]

Diagram 3 5 . 3

White to move
 Endgame Secrets 109

Diagram 3 5 .4 Diagram 3 5 .5

White to move Black to move

37 :1 b 7 ? ! If Black's knight returns he hasn't

Another impulsive move. Black's achieved anything. Black has to

knight wanted to go to f5 in any move the king which involves the

case. 3 7 @fl is better. sacrifice of the c'.se6.

37 ... ltJf5 A) 39 . . . @g6 40 .U.b6 .U.el+ 41

Black's threat is not only . . . .U.c4 @g2 .U.c2 42 l:.xe6 .U.xb2 43 i:ia6

but mainly . . . é'.bd6-e4, attacking f2 . .U.b4 (43 . . . .U.a2 44 é'.be5+, 43 . . . é'.be3+

38 é'.bh2? 44 @gl) 44 é'.be5+ @g7 45 .U.a7+

White is hypnotized by the idea of @h6 (45 . . . @g8 46 é'.bg4, 45 . . . @f8 46

playing é'.bg4, but now Black attains é'.bd7+) 46 é'.bf7+ @g6 47 ltJh8+ with

a large advantage. a perpetual check.

I) 3 8 .U.b4 é'.bd6 39 .U.xa4 .U.el+ 40 B) 3 9 .. . @g8 40 .U.b6 .U.el+ 41

@g2 é'.be4 followed by . . J � c 2 leads @g2 .U.c2 42 .U.xe6 .U.xb2 43 i:ia6

to the fall of the é'.sf2, e. g. 4 1 é'.bh2 (not 43 .U.xf6 a3 44 i:ixf5 a2)

i:ic2 42 ltJg4 f5 43 i:ia8+ @g7 44 43 . . . .U.a2 44 .U.xf6 'De3+ 45 @gl

i:ia7+ @f8 45 .U.a8+ @e7 46 .U.a7+ lbc4 46 .U.a6 a3 47 é'.bxg5. White is

@d8. OK, e . g . 47 . . . .U.c2 48 h4 a2 49 @g2

II) 3 8 i:ib6 is correct. Attacking e6 'Dd2 50 .U.a8+ @g7 5.1 4Je6+ @f7

provides White with enough coun­ 52 4Jc5 'De4 53 é'.bd3 and 4Jb4.

terplay. 3 8 . . . @f7 39 .U.b7+ [35.5] 38 ... é'.bd6

(39 g4 ? ! é'.be7 drives the knight back Certainly not 3 8 . . . é'.bxd4 39 'Dg4

but weakens the squares h4 and f4, .U.el+ 40 @g2 f5 41 é'.bf6+ @f8 42

Black has é'.be7-g6-f4) 4Jh7+ @e8 43 4Jxg5 and White is

out of danger.

39 l:td7

Now it's too late for an attack

against e6, e.g. 39 .U.e7 é'.be4 40

i:ixe6 .U. e l + 41 @g2 i:ic2 42 é'.bg4

(42 .U.a6 .U.xf2+ 43 @gl l:tx b 2 44

1 1 O Endgame Secrets

J:lxa4 étJxg3) 42 f5 43 ctJf6+ (43 4 7 .. . J:lxf2+ 48 @g 1 @h6 49 étJf7+

ctJh6+ @h7) 43 @f7 and Black is @g6 50 ctJe5+.

winning. 39 .l:.b4 can be answered 45 ctJe5

by either 3 9 . .. ctJe4 or 3 9 .. . J:lc4. 45 J:lb8+ @e7 46 J:lb7+ @d8

39 ... ctJe4 doesn't help: 47 ctJe5 ctJxf2 48

40 ctJg4 :c1 + ctJf7+ @e8 49 h4 (49 ctJxg5? ctJe4+,

41 @g2 .l:.c2 49 ctJd6+ ©f8) 49 .. . ctJh3+! leads to

[35.6] positions similar to those we look at

later, but White's knight is worse on

f7 than on g6.

45 étJxn
The next move Black's knight

might go to d l , d3, e4 or g4; in each

case it's a discovered check. The

game attains its clímax.

46 ctJg6+

The knight has to avoid an ex­

change. In a pure rook ending

White's tactical possibilities are

more reduced.

46 @e8

Diagram 3 5 . 6 [35.7]

White to move

With time trouble behind me, I

had a closer look at the shattered re­

mains of my position. The �b2 is

under attack and after .. . f6-f5 the

étJg4 has to abandon the �f2. In

short, White has a bad position. In

situations like these there is only

one strategy: complicate matters and

wait for a miracle.

42 J:lb7

After 42 J:la7 f5 43 J:la8+ @g7 44

J:la7+ @f8 45 ctJe5 J:lxf2+ 46 @ g l Diagram 3 5 . 7

J:lxb2 47 J:lxa4 ctJxg3 Black enjoys White to move

two extra pawns.

42 5 47 h4!?

43 J:lb8+ @g7 By creating a passed pawn White

44 J:lb7+ @f8 completes the chaos. I couldn't find

44 .. . @g6? permits a perpetua! a direct refutation, thus 47 h4

check: 45 étJe5+ @h6 46 ctJf7+ @h5 looked like the best try. During the

47 étJh8 (threatening mate on h7) game I dismissed 47 J:le7+ @d8 48

 Endgarne Secrets 111

:xe6, because the white rook aban­ @gl, 48 . .. ltJe4+ 49 @gl ltJxg3 50

dons control of the seventh rank in h6 ltJe2+ 51 @fl g3 doesn't work

exchange for the unimportant l::,e6. on account of 52 :b8+ @d7 53

Nevertheless this line isn't clear. In ltJe5+ @c7 54 h7 ltJf4 55 :c8+!

view of the reduced material and the @xc8 56 h8=¡f+ and 'li'g7xg3) 49

possibility h3-h4 White has drawing @fl f4 (49 .. . ltJg5 50 h6 :h2 51

chances, e.g. 48 . . . :xb2 49 :a6 (49 ltJh4) 50 lDxf4 lDxf4 51 gxf4 @f8

:d6+? @c7 5 0 :xd5 ltJg4+ 5 1 @ g l 52 :a7 :xb2 53 :xa4 .:i.h2 leads to

:b1+ 52 @g2 ltJe3+) 49 .. . ltJe4+ a rook ending which offers good

(49 .. . l:ra2 50 h4 ltJh3+?! 5 1 @f3) 50 drawing chances to White.

@ g l tt:Jc3 5 1 h4. V) 47 . .. ltJdl+ is promising: 48

47 C{le4+? @gl :xb2 (48 . . . ltJxb2 leads to line

The position is amazingly compli­ III)

cated, it's no wonder that Black A) 49 :a7 ltJc3 50 :as+ (50 h5

goes wrong. I give a small excerpt lDe2+ followed by .. . ltJxg3+ or

from the multitude of possible .. . ltJf4+) 50 . . . @f7 5 1 ltJe5+ @f6 and

variations, in order of their approxi­ Black wins.

mate value (i.e. Black's weakest B) 49 l:re7+ @d8

continuation can be found under 1, B l ) 50 :xe6 a3 5 1 :a6 a2 52 h5

the strongest under VI). ltJe3 53 h6 : b l + 54 @f2 (54 @h2

I) 47 .. . gxh4? 48 gxh4 isn't dan­ ltJg4+ 55 @g2 :b2+ and . .. ltJxh6)

gerous; the white king now has ac­ 54 .. . al=¡f 5 5 :xal :xal 56 h7 :hl

cess to the squares g3 and f3 . 57 h8=¡f + :xh8 58 ltJxh8 ltJg4+

II) 47 .. . ltJg4+? fails to impress af­ and .. . @e7-f6(f7) leads to a won

ter 4 8 @ g l gxh4 (48 . .. lDh2 49 ltJe5) knight ending.

49 gxh4 ltJf6 5 0 :e7+ and :xe6. B2) 50 h5 50 .. . ltJe3 51 :xe6 ( 5 1

III) 47 .. . ltJd3+? wins the l::,b2, but h 6 : b 1+ and .. . ltJg4xh6) 51...:g2+

this isn 't enough to win either. 48 52 @h 1 :xg3 with a decisive

@gl l:rxb2 (48 .. . ltJxb2 49 h5 gives advantage.

White a mighty passed pawn, e.g. VI) 47 . .. ltJh3+! [ 3 5 . 8 ]

49 .. . a3 50 h6 a2 51 :a7 :c1+ 52

@g2 al=¡f 53 :xal :xal 54 h7

and White wins. 48 .. . f4 49 gxf4

gxf4 5 0 l:re7+ @d8 5 1 :n : x b 2 52

h5 a3 53 h6 a2 54 :a7 f3 55 h7

:g2+ 56 @fl is a draw) 49 :e7+

@d8 50 h5 a3 (50 .. . ltJel 51 h6

ltJf3+ 52 @fl :h2 53 h7 a3 54 :a7)

51 h6 a2 52 :a7 :b1+ 53 @g2

a l = ¡f 5 4 :xal :xal 55 h7 ltJel+

56 @f2 (56 @h3? g4+ and .. . ltJf3+)

56 . . . ltJd3+ 57 @g2 with a perpetual.

IV) 47 . . . g 4 ? ! 48 h5 ltJh3+
Diagram 3 5 . 8
(48 .. . tt:Jd3+ 49 @gl, 48 . . . ltJdl+ 49
White to rnove
112 Endgame Secrets

. .. is the move I like bes t. Black is

wmnmg:

A) 48 @xh3 g4# puts an immedi­

ate end to White's suffering.

B) 48 @fl f4 49 gxf4 g4 results in

a mating attack: 50 .: b 8 + @d7 5 1

l2:J e 5 + (5 1 h5 g3 52 l2:J e 5 + @c7 53

l:t g 8 .a'.f2+ 54 @el g2 55 h6 .s'. fl +

56 @d2 g l = 'i' 57 .a'. x g l lb x g l 5 8 h7

.s'.f2+ and . . . .s'.h2) 5 1... @ c 7 52 l:r g 8

l:txb2 53 h5 a3 54 h6 a2 55 .s'. a 8

.s'. b l+ fo ll o w e d by . . . a l = 'iV , . . . .a'. x a l

Di ag ra m 3 5. 9

and . . . .s'. a 8 .

White to move
C) 48 @hl f4 49 gxf4 g4 is simi­

lar to line B. . . . and W hite c a n 't stop b oth

D) 48 @f3 (the toughest defence) passed p aw n s.

4 8 . . . g4+ ( 4 8 . . . l:rf2+ 49 @e3 f4+ 50 48 @gl

gxf4 gxf4+ 5 1 @d3 isn't clear) 49 48 @f3 CLi d2+ and . .. 4J c4+ .

@e3 l:tg2 50 h5 (50 .a'.e7+ @d8 5 1 48 4Jxg3

.a'.xe6 .a'. x g 3 + 52 @d2 l:.b3 and the 4 8 . . . g4 l ea ds t o li ne IV , n ot e to

g-pawn advances) 5 0 . . . .S: x g 3 + 5 1 B la c k's 4 7t h mo v , e while a fter

@d2 l:tb3 52 .s'. e 7 + (52 ¡txb3 axb3 4 8 . . . gx h4 49 gxh 4 4Jg3 50 ¡te7+

53 h6 l2:J g 5 ) 52 . . . @d8 53 h6 �xb2+ @ d8 5 1 lhe 6 .s'. xb 2 52 .s'. a6 lb 2 e +

54 @el (54 @ c 3 .s'.h2 55 ¡tg7 l2:Jg5 53 @f2 W h it e's a c ti v e ki ng and the

56 lbf4 ltJ e4+ , 54 @ d l ¡th2 55 .s'.g7 h- pawn s e cu r e the d ra w, e. g .

4Jg5 56 4Jf4 a3 with the po int 57 53 . . . 4Jx d 4+ 54 @ 3 e 4Jc2 + 55 @f4

.S: xg5 .s'. hl + 58 @ d2 a 2 5 9 .s'.g 8+ a3 56 h5 a2 5 7 h 6 al ='iV 58 .a'. xa l

@ 7 e 60 .s'. a8 al= 'iV 6 1 !i xal .s'. xal étJ xal 59 h 7 .s'. h2 6 0 h8 ='iV + .a'. x h 8

62 h7 .s'.h l 6 3 h8= 'iV l:t x h 8 64 l2:J g6+ 6 1 lb xh8 d4 62 4Jf7 + @c 7 63 lb e 5 .

@ f6 65 lbx h8 @ g7, 54 @ e3 f4+ 55 49 hxg5 ltJe4 [ 3 5 . 10 ]

@ d3 étJf2 + 56 @ c3 3
a and . . . 4Je 4+ ,

. . . í:td2#, 5 4 �d3 .s'. h2 55 l:t g7 étJf2 +

56 @ e3 ltJd 1+ and . . . l:t xh6)

54 . . . .s'. h 2 55 _:g7 étJg5 56 étJf4 a3 57

@ bl ( 57 .:xg5 l:t h l+, ... a3 - a2)

5 7 . . . 4J e4 5 8 lb xe6+ @c 8 59 l:tc 7+

@ b8 60 h7 g 3 [ 35.9 ]

D iag ra m 3 5 . 1 O

White to move
 Endgame Secrets 113

After 49 . . . t'.be2+ 50 @fl t'.bxd4 5 1 5 1... @ e 7 isn't better: 52 g6 t'.bf6

étJf4 White draws thanks to his 53 t'.bh7 étJh5 54 l:.b7+ @d6 55 g7

g-pawn. tbxg7 56 l:.xg7 �xb2 57 l:.a7 l:.b4

50 l:.b8+! 58 ltJg5 l:.xd4 59 l:.a6+. After the

When I saw this move my mood gain of the e-pawn White attains the

markedly improved. White can draw quite easily.

draw in brilliant style. 50 l:.e7+?! But what now? Rook and g-pawn

@d8 51 laxe6 is at least dubious, are under attack.

e.g. 5 1... l:. x b 2 52 l:.a6 l:.a2 53 étJf4 52 g6!

ltJxg5 54 ltJxd5 a3 55 ltJc3 l:.al + 56 The point. White doesn't care

@g2 (56 @f2 t'.be4+ 57 ltJxe4 fxe4 about his rook as the g-pawn wins

followed by . . . a4-a3 and . . . e4-e3-e2 the race.

loses at once. 56 @h2 f4 57 étJb5 is 52 @xb8

also bad in view of 57 . . . a2 5 8 ltJc3 Checks with the rook (now or one

l:. c l 59 ltJxa2 �c2+ 60 @ g l f3 6 1 move later) only lead to a repetition:

étJb4 l:.g2+ 62 @fl étJh3) 56 . . . f4 52 . . . 1:. c l + 5 3 @g2 l:.c2+ 54 @ g l .

(56 . . . étJe4 57 étJd5) 57 �a7 f3+ 58 53 g7 ct)f6

@ f2 and White scrapes the draw. 54 étJd7+!

The text is more straightforward and After the rook also the knight of-

more elegant. fers itself.

50 @d7 54 @b7

5 0 . . . @f7 51 ltJe5+ @e7 52 l:.b7+ 55 étJxf6 l:.c8

@d8 53 g6 troubles Black only. 56 g8=¡v l:.xg8+

51 étJf8+ 57 ltJxg8 @c6

It is important to block the eighth 58 ct)f6

rank. 5 1 ltJe5+? @c7 52 g6 (52 l:.g8 Now White even has a nominal

l:.xb2 53 g6 a3 54 g7 étJf6 loses a material edge, but on account of the

rook) 52 . . . @xb8 53 g7 l:.c8 stops threat . . . @b5-c4-b3xb2 he still has

the pawn. to be careful. His knight hurries to

51 @c7 [ 3 5 . 11 ] capture the i0ia4. If 5 8 @f2? then

5 8 . . . @b5 59 @e3 @c4 intending

. . . @b3xb2 . .

58 @b5

59 ctJd7 @c4

60 étJb6+

60 étJf8? @b3 6 1 ltJxe6 @xb2 and

Black wins.

60 @xd4

61 ltJxa4 ( 3 5 . 1 2 ]

Diagram 3 5 . 1 1

White to move
114 Endgame Secrets

Diagram 3 5 . 1 2 Diagram 3 5 . 13

Black to move White to move

I had this position in mind when I 66 ét:Jd3!

played 50 :b8+. lt takes only sorne Getting around the final obstacle.

circumspection to reach the draw. After 66 lt:Jxe6? d3+ 67 @e3 (67

The b-pawn supports the knight on @f2 d2 68 ét:Jd4+ @d3) Black does­

a4 and limits the mobility of Black's n 't play 67 . .. d2? 68 ét:Jd4+ @c3 69

king. lt:Je2+ with a draw; instead he

61 @c4 comes up with the nasty 67 .. . f4+!

6 1... e 5 62 b4 isn't better. and the d-pawn promotes.

62 b3+ @b4 66 ... @b3

63 @f2 d4 67 lt:Jc5+ @c2

63 .. . e5 64 é2Jb6 d4 65 lt:Jc4 e4 66 68 ét:Jd3 ! @c3

ét:Jd6. 69 lt::ie5 f4

64 @e2 @xb3 70 @dl d3

64 .. . e5 65 @d3 f4 66 ét:Jb6 f3 67 71 ét:Jf3

lt:Jc4. Black loses a pawn, the rest is

65 lt:Jc5+ @c2 easy.

[ 3 5 . 13 ] 71 ... d2
@.d3
72 ét:Jxd2

73 ét:Jf3 @e3

74 lt:Je5 @e4

75 lt:Jg4 @f3

76 lt:Je5+ @f2

77 @d2 f3

78 ét:Jd3+ @g3

79 @e3 e5

80 ét:Jf2 @g2

1/2-1/2
The Porz team won the fight 5- 3

and later on also the German team

championship.
 Endgame Secrets 115

Endgame 3 6 II) 2 f5 c4 3 dxc4 bxc4 4 ctJa5

R.Fischer-J.Rubinetti i.c5 5 lig8+ @d7 6 'tJxc4 i.d4 !

Buenos Aires 1 9 7 0 (6 . . . lixf2+ 7 @ e l �f4 8 �g4! i.d4

Earlier publication: Encyclopaedia 9 �xf4 exf4 1 O @e2 leads to a diffi­

of Endgames (Rooks 1.()/endgame cult minor piece ending) 7 .:g7+

1 4 6 2 ( editorial remarks) @c6 8 @d3 �xf2 9 J:íxh7 Jif3+ 10

@c2 :t2+ 11 @b3 l:ff4 12 tt:Jd2

i.e3 and Black is still alive.

III) 2 tt:Ja5 c4 ! ? and once again

the position lacks clarity.

2 d4 @d8

Or:

I) 2 . . . c5 3 e5 c4 4 'Da5 wins the

i.d6 ( according to the

Encyclopaedia).

II) 2 . . . f3+ 3 @e3 leads nowhere.

3 ctJa5! c5

4 e5 .i.f8

5 étJc6+ @e8
Diagram 3 6 . 1
5 . . . @c8 6 lig 8 @d7 7 d5 @e 8 8
White to move
e6 is unpleasant as wel l.

6 li xc7 [ 36.2] 1-0

White to move. At the moment
6. . . cxd4 7 e6 leads fo mate. It 's
his position can't impress. How can
s urprising how qu ickly F isc he r co­
he change t h a t ?
ordinated his loos e pieces.

Commentary

1 f4!

Now White's central pawns get

going.

1 exf4?

The only chance to fight is

1... c 5 ! ? :

I) 2 fxe5 i.xe5 3 kig8+ @d7 4

tt:Jxc5+ @e7 (Black threatens . . . i.d4

and . . . i.xb2) 5 @e3 (5 tt:Jxa6 i.d4)

5 . . . .i.xb2 6 tt:Jxa6 i. c l+ 7 @e2

.i.xa3 8 tt:Jxc7 i.c5 9 ctJxb5 �xf2+ Diagram 3 6 . 2

with excellent drawing chances. Black to move

116 Endgame Secrets

Endgame 3 7 Commentary

I.Papaioannou-C.Lutz
I reached position 37.1 twice in
Elista Olympiad 1 9 9 8
toumament games, each time play­
Earlier publication:
ing Black. This is strange as White
SCHACH 1 1 / 1 9 9 8
doesn't have an advantage at all. It ' s
(R. Tischbierek)
even stranger that both of my

opponents committed the same

mistake.

13 c4

1 3 'Dd2 l:rfd8 is equal, too.

13 'Dxd4

14 i.xd4 'iVe4

15 1'Vxe4 'Dxe4

[37.2]

�.i
.y� m � ��• �

Diagram 3 7 . 1
�
& -

�--
1/,,,,,�- �-� �-
l'f
�

0-,-. -1/um�-✓
& �
� �
/ &

White to move

Analyse the game.

·�,��,�·�
�.,�-.-.,

13

1'Vxe4
c4

'Dxe4
'Dxd4

16
14 i.xd4

'Dd2 'Dd6
1'Ve4

17
15

b3
�n
�m�-1
■■ n � ■ 1/,,,,,�m �
l:rfd8 18 'Df3 f6 19 i.c5 @f7 20

l:rd2 b6 21 i.xd6 l:rxd6 22 l:radl

Diagram 3 7 . 2
l:rad8 23 l:rxd6 i.xd6 24 @fl e5 25
White to move
@e2 @e6 26 'Del h5 27 'Dc2 h4 28

'De3 g6 29 'Dd5 f5 30 CDc3 l:rc8 31 There is stubbom prejudice that in

'Db5 i.c5 32 'Dc3 i.d4 33 'Db5 positions with a pawn structure as in

i.c5 34 'Dc3 a6 35 'Dd5 b5 36 cxb5 the present case White is automati­

axb5 37 'Dc3 l:[bS 38 @fl b4 39 cally better due to his queenside ma­

'Dd5 lfü7 40 'Lle3 i.xe3 41 fxe3 jority. But this assessment is plainly

l:rc7 42 líd8 l:rcl + 43 @f2 l:rc2+ 44 wrong. Black's pawn structure has

@fl l:rcl + 45 @f2 l:rc2+ 46 @fl merits as well, thanks to the �e6 he

l:rxa2 47 l:rb8 @d5 48 l:rxb4 g5 49 has a greater influence in the centre.

l:rc4 l:ra3 50 l:rb4 l:ral+ 51 @e2 Whether White can create a passed

l:ra2+ 52 @fl l:ral + 53 @f2 l:ra2+ pawn on the queenside or Black can

54 @fl l:rb2 55 l:ra4 l:rxb3 56 e4+ develop play in the centre depends

fxe4 57 @el l:rb2 58 @fl l:rc2 59 on concrete facts, i .e. on the place­

l:rb4 l:ra2 60 @gl e3 61 @fl l:rf2+ ment of the pieces.

62 @gl l:rc2 63 @fl l:rc4 0-1 16 CD d 2 ?

 Endgame Secrets 117

And there it is, the mistake. White !il.xe7 @xe7 Black controls the

wants to exchange the l'.be4 but for­ d-file.

gets that Black has different ideas. 21 .á'.xd6

After 1 6 'tJc3 'tJxc3 1 7 !il.xc3 .á'.fd8 22 11'.adl !:rad8

both players centralize their kings, 23 .á'.xd6 !iLxd6!

exchange all the rooks and the draw [37.3]

can be signed. Black can't play

1 6 . . . ctJd6 as 1 7 !iLc5 .a'.fd8 1 8 ctJe4 is

favourable for White. With the text

move White blocks the l:.dl for a

short moment.

16 ctJd6!

17 b3 11'.fdS

Now Black is slightly better as he

can fight for the squares along the

d-file. Especially the d4 square is a

target after . . . ctJf5.

18 'tJf3 f6

Taking away the squares e5 and

Diagram 3 7 . 3
g5 from the knight. This knight will
White to move
go for a long walk during the game,

without finding a decent square. It is surprising that Black refrains

19 !il.c5 from trading off the second pair of

White prevents . . . ctJf5, but maybe rooks. Here I refer to Shereshevky's

he should centralize his king in­ Endgame Strategy. In the chapter

stead. After 1 9 @fl @f7 20 @e2 e5 "The 3-2 queenside pawn majority"

21 !il.b2 @e6 Black has the better he looks at several endgames with

king, but White still has his bishop. the present pawn structure. He

19 @f7 writes: "In all the endings examined

20 l:.d2 in this chapter control of the d-file

Only now White deviates from was a decisive factor. It is true that

the game Keitlinghaus-Lutz (Ger­ one comes across endings (usually

man championship 1997). After 20 with just one pair of rooks) where

l:. e l ctJf5 21 !iLxe7 'tJxe7 22 !:redl this <loes not give any real advan­

ctJc6 23 'tJe 1 we reach endgame 3 2 . tage, but these are merely excep­

20 b6 tions which confirm the rule." The

21 !iLxd6 present position is such an excep­

Now Black remains with the bet­ tion. After 23 . . . .a'.xd6 24 !:rxd6

ter minor piece. White lacks control !iLxd6 25 @fl Black can't penetrate

over the black squares, therefore he into White's camp whereas after the

shouldn't, even think about advanc­ text Black might use his rook for a

ing his queenside majority. But after minority attack by means of .. . !:rc8,

21 !iLb4 ctJe4 22 .á'.xd8 l:.xd8 23 .. . a7-a6 and . .. b6-b5. The white rook
1 1 8 Endgame Secrets

can't achieve anything along the should take care of a subtlety. As

open d-file as Bl a c k ' s king and long as Black's pawn is on f6 he

bishop control the entry squares. shouldn ' t play g2-g3 as then after

24 @fl e5 . . . g7-g5 he has to allow either a

25 @e2 @e6 weak pawn on h4 or-if White takes

Now the king is well centralized, on g5-a passed pawn along the h­

Black's superiority is obvious. file. With the pawn still on g2

26 cbel . . . g7-g5, h4xg5, . . . f6xg5 shouldn't

The white knight continues his be feared; Black can't create a

odyssey. passed pawn then. Compare with

26 h5 [ 3 7 .4 ) endgame 1 2 .

27 h4

.�.�J.�-�
�.,�o,,,,,,��
1/mn� - � ��r.l'.
I briefly pondered over the correct

order of . the

. . . h5-h4. After
moves

27 . . . f5
. . . f6-f5

there could
and

-'l. � f!t � 'l. � follow 28 h4.

■----"■'··-■---··� 28

29
ébe3

étJdS
g6

f5

■8■ ■ � Black ignores the ctJd5 and con­

■8■ ■ ■8 tinues to increase his space

·�· .¡gw •• advantage.

, •• ,ADid 30

In the
ébc3

search for counterplay

White poses the threat of ctJb5, but

Diagram 3 7 .4
Black easily parries this.
White to move
30 &Cc8

This move took me sorne time on 31 ébb5 ..tes

the clock. Before Black starts his 32 ébc3

queenside attack, he improves his 32 cbxa7? �a8.

space advantage. I dispensed with 32 ..td4

26 . . . g5? as then 27 g4 and 28 ctJc2, Before starting further action

29 ébe3 blocks the Black pawns. Black repeats moves. Repeating

27 cb c 2 ? ! moves is an appropriate means to

White ignores B l a c k's plan. Per­ gain sorne time on the clock and to

haps he didn't like the fact that plumb the depths of the position.

Black seizes the h-file after 27 g4 33 étJbS ..tc5

hxg4 28 hxg4 �h8, but 27 h4 ? seri­ ! 34 ébc3 a6

ously deserves attention. White puts Now the threat 1s .. . ..td4 and

a pawn onto the colour of the . . . b6-b5.

bishop, but he avoids a restriction of 35 ébd5 [ 3 7 . 5 )

his kingside. However, White

 Endgame Secrets 119

A) 3 8 . . . cxb3? 39 .:.xc8 bxa2 40

'Llc5+ @d5 4 1 'Llb3 or 40 .. . i.xcS 4 1

lía8 is bad.

B) 3 8 .. . c3 [ 3 7 . 6 ].

Diagram 3 7. 5

B lack to move

Certainly Papaioannou didn 't like

35 a4 to prevent . .. b6-b5. Black

might then continue with 3 5 .. . a5 Diagram 3 7 . 6

and .. . ltg8, . . . g6-g5-g4. White to move

35 ... b5?!

3 5 . . . i.d4 might be more accurate. The �c3 severely restricts the

Then 36 'Llf4+ @f7 37 'Lld5 b5 white pieces. lt can't be attacked as

doesn't help White. Better is 36 the 'Lla6 is misplaced, e.g. 39 l2Jb4

'Llb4!? b5, with a branch: (39 a4 �c6 40 'Llb4 l:[b6) 39 . . . �a8

1) 37 cxb5 axb5 3 8 l:ld2. For the B 1 ) 40 :a 1 i.c5 4 1 'Lld3 (4 1 'Llc2

moment, . .. b5-b4 is prevented and líd8) 4 1... i. a 3 .

after the exchange of rooks (with B2) 40 ltc2 @d6 4 1 'Lld3 (41 @ d l

Itc2) White should no longer lose. e4) 4 1... e 4 42 'Llcl �c5 43 a4 (43

But Black has 3 8 . . Jkl 39 ltc2 lí g l . @el @b4 44 'Lle2 i.e5) 43 . . . @b4.

Black will capture the tg2 and pos­ In both cases Black is better.

sibly the �h3. White can create 3 5 . . . i.d4 is stronger than 3 5 .. . bS,

counterplay on the queenside, but since White now has a chance to

nevertheless he is in trouble, e . g . 40 equalize.

:c6+!? (40 'Llc6 ltxg2 41 'Llxd4+ 36 cxb5?

exd4 42 @d3 @d5 43 b4 g5 and the 36 b4! is best. After 3 6 . .. i.d4 (or

black pawns become dangerous) else 3 7 c5) 37 'Llf4+ exf4 (37 . .. @fl

40 .. . @fl 41 .:.c7+ @f6 42 líc6+ 3 8 c5 i.xc5 3 9 bxc5 exf4 40 ltd7+

@g5 43 líc2 and now not is to no avail) 38 líxd4 :xc4

43 !txg2? 44 @fl :h2 4 5 f4+, but ( 3 8 . . . bxc4 39 @d2) 39 lhc4 bxc4

43 @h5 and White can't protect the pawn ending is drawn: 40 f3

g2. (forced in view of 40 @d2? f3 41

11) 37 'Llxa6 bxc4 38 It c l (38 gxf3 g5 and . .. g5-g4) 40 .. . @dS 41

'Llb4 cxb3 3 9 axb3 ltc3 and White's @d2 @d4 42 a4 g5 43 @c2 g4 44

knight is in trouble) and now: @d2 [ 3 7 . 7 ].

120 Endgame Secrets

I) 3 8 a4 bxa4 3 9 bxa4 (39 'Llxa4

iLd4, the knight is out of play)

3 9 . . . :b3. The a-pawn is lost.

11) 38 l:íd5 ! ? is interesting:

3 8 . . . iLd4 (38 . . . l!tc8 39 l!t d l ) 39

�xb5 l:[c8 40 tbd 1 l!tc2+ 41 @e 1

�xa2 42 b4 [ 3 7 . 8 ].

Diagram 3 7 . 7

Black to move

As soon as Black plays . . . c4-c3

White answers with b4-b5, e.g.

44 . . . gxf3 45 gxf3 c3+ 46 @c2 @c4

47 b5 axb5 48 axb5 @xb5 49 @xc3

@c5 50 @d3 @d5 51 @c3 with a

draw.
Diagram 3 7 . 8
36 ... axb5
Black to move
Now the 'Lld5 has lost its support

and the queenside is vulnerable.

Though king and knight are cut
White faces serious problems.
off at the edge, things aren't clear as
37 tbc3
the 'Lldl controls the squares f2 and
3 7 �d2 iLd4 3 8 tbb4 leads to the
b2. The b-pawn might create
previous note. 3 7 b4 iLd4 or 3 7 a4
counterchances.
bxa4 3 8 bxa4 iLd4 don't bring any
38 b4 [ 3 7 . 9 ]
relief either.

37 l!tb8

Black has two weaker moves at

his disposal:

1) 3 7 . .. iLd4 38 'Llxb5 �c2+ _3 9

@d3 followed by tbxd4+ and �d2

needn 't be feared.

11) 3 7 . . . b4 3 8 tbd5 and the knight

can't be driven away that easily.

Neither 3 8 . . . �d8? 3 9 'Lle?+ @e7 40

:xd8 @xd8 4 1 'Lle6+ nor 3 8 . . . �c6

3 9 ¡(d2 �d6 40 'Llc7+ @d7 4 1 tbd5

do the trick.

38 @fl Diagram 3 7 . 9

To answer . . . iLd4 by 'Lle2 . White to move

Altematives:
 Endgame Secrets 121

The queenside is tied down. Black lacks the tempo .. . @d5. White

39 t2Jd5? might play 45 klb5 or 45 e4.

This is already the decisive mis­ 42 kld8

take. It ' s better to play 3 9 t2Je2 and White switches to active defence.

:d2. Waiting passively loses as well: 42

39 klb7 kld2 klc3 43 @e2 g5 44 @f2 f4 45

The idea of 37 . . . :b8 becomes ap­ e4 (45 exf4 gxf4 is hopeless)

parent. Black bring his rook to the 45 .. . :e3 46 :e2 lhe2+ 47 @xe2

d-file while keeping the c7 square @d6 48 @d2 (48 @f3 @c5 49 @g4

protected. After for instance 3 8 @e2 @d4 50 @xg5 @xe4 5 1 @xh4 @e3

kld7 39 t2Je3 i.xe3 the pawn ending 52 @g4 e4 53 h4 f3) 4 8 .. . @c5 49

is a simple win for Black, because @d3 g4 ! 50 hxg4 f3 51 @e3 fxg2

he is virtually a pawn up. White's 52 @f2 h3 53 g5 @d6 with a win.

next move is forced. 42 ... :el+

40 t2Je3 i.xe3 43 @f2 klc2+

The knight is exchanged before it 44 @fl klcl+

can get to c4. During the game I Y ou should never forget to repeat

wasn't completely sure whether moves.

Black is winning, but this seems to 45 @f2 l::rc2+

be the case. It is especially helpful 46 @fl klxa2

that Black's h-pawn is already far 47 klb8 @d5

advanced. 48 klxb4 g5

41 fxe3 [ 3 7 .1 0 ] [ 3 7 . 11 ]

Diagram 3 7 . 1 0 Diagram 3 7 . 1 1

Black to move White to move

41 :c7! We come to the final stage of the

Owing to the repetition of moves I game. The material balance is even,

now had enough time to become en­ but Black's positional advantage is

grossed in my position. The text is overwhelming. He can make use of

better than 41...kld7 42 kl c l l::rd2 43 zugzwang ideas. The @fl can't go

:c4 klxa2 44 klxb4 g5, as then to either e 1 ( on account of .. . klxg2)

 122 Endgame Secrets

or gl (on account of . . . :e2xe3). 11) But 49 . .. :c2 is simple and

That's why only White's rook can strong. After 50 :b4 (other moves

move. But the rook has to defend neglect the e4 square) 50 . . . :b2 we

both the fourth rank and the ¿jb3. reach the game.

49 :c4 49 :a3!

49 : a 4 ! ? is tricky. As is shown by After 49 . .. �b2 50 :b4 it's not so

Tischbierek the transition to a pawn easy for Black to lose a move.

ending is only a draw: 50 :b4 :al+

I) 49 .. . :xa4? 50 bxa4 e4 51 @e2 :a2+

(50 .. . @c4? 51 e4! is only trouble­ 52 @fl

some for Black. White's king gob­ 52 @d3 J:txg2 loses without any

bles up the enemy pawns) 51 g3 ! ado.

(White has to break up Black's 52 :al+

pawn wall. After 5 1 @e2? f4 it's too Another repetition of moves . . .

late) 5 1... hx g 3 ( 5 1... @ c 4 ? 52 gxh4 53 @f2 :a2+

gxh4 53 @e2 @b4 54 @d2 and 54 @fl :b2!

White is winning) 52 @g2 [ 3 7 .1 3 ]

A) 52 .. . @c4 53 a5 f4 (53 .. . @b5 54

@xg3 @xa5 55 h4) 54 a6 fxe3 55 a7

e2 56 a8='i' e l ='i' 57 'i'c6+ with a

perpetua! check.

- B) 52 .. . f4 53 exf4 @d4 (53 .. . gxf4

54 a5 e3 5 5 @ f3 g2 56 @xg2 @d4

57 a6 is drawn as well) 54 f5 e3 55

f6 e2 56 f7 e l = 'i' 57 f8='i' [ 3 7 . 1 2 ].

Diagram 3 7 . 1 3

White to m o v e

The first zugzwang.

55 :a4

White abandons the pawn. 55 :b8

@e4 56 :b5 g4 and . .. g4-g3 is lost.

55 :xb3

56 e4+

Diagram 3 7 . 1 2 56 @f2 :b2+ 57 @fl :c2 58 :b4

Black to move :a2 leads to the second zugzwang

position.

Only Black has to worry. 56 fxe4

57 @el
 Endgame Secrets 12 3

57 <it>e2 :b2+ 58 @e3 lhg2 59 Endgame 38

kla5+ <it>e6 60 <it>xe4 �e2+ 61 @f3 J.Timman-C.Lutz

.:e 1 and the two extra pawns win Wijk aan Zee 1 9 9 5

the game. Earlier publication:

57 ... .s'.b2 Informan! 62/569.

58 @fl .:e2

59 �b4 .a'.a2

The third zugzwang.

60 @gl e3

61 @fl klf2+

62 @gl kle2

63 @fl .a'.e4

[37.14]

Diagram 3 8 . 1

Black to move

Analyse the game.

29 .. JieS 30 f3 �e3+ 31 'iVxe3

:xe3 32 @f2 :es 33 f4 @f7 34

:dS klb8 35 <it>e3 <it>e6 36 �dl f5

37 g4 fxg4 38 hxg4 ctJf6 39 fS+

Diagram 3 7 . 1 4
@d7 40 g5 :es+ 41 @f4 ctJh5+ 42
White to move
@g4 g6 43 fxg6 hxg6 44 .a'.dS ctJg7

45 kXxbS ctJfS 46 :b7+ @e6 47 .:e7

0-1
:hS 48 @f3 :h3+ 49 @f2 .:h2+ 50
Black continues with . . . @e4 and
@el .:hl + 51 @f2 :h2+ 52 @fl
. . . .s'.d4-d2(+).
:hl + 53 @g2 :dl 54 @f2 .s'.d4 55

@el .:h4 56 @d2 :h2+ 57 @e3

:h3+ 58 @xe4 lt:Jxd6+ 59 <it>e5

ctJe4+ 60 @bS ctJxgS 61 a4 tZ)f3 62

as ctJd4+ 63 @e4 :h4 64 i.es

ctJe6+ 65 @bS ctJxaS 66 @xa5 g5

67 @bS g4 68 :g7 g3 69 .a'.xg3 .a'.hl

70 @e6 .a'.el 71 i.d4 :el+ 72 i.e3

:dl 73 kle3+ @fS 74 @e5 .:dS 75

124 Endgame Secrets

i.e5 :es+ 76 @d5 :as 77 ¡(f3+ 30 f3 �e3+?!

@g4 78 ¡(f7 :as+ 79 @e4 :a4+ 80 Still on the wrong track. 3 0 . . . 'iVc6

i.d4 @g5 81 :g7+ @h4 82 @e5 is better.

@h3 83 :gl ¡f,b4 84 i.e3 lig4 85 31 'ifxe3 :xe3

:a1 @g2 86 .itf4 :gs 87 :a2+ 32 @f2 :es

@f3 88 :a3+ @e2 89 @e4 :es+ 90 33 f4!

i.e5 :e7 91 :a2+ @el 92 @d4 The difference now is the passive

@fl 93 i.f4 :e2 94 :as :e7 95 placement of the black pieces.

@d3 @g2 96 :f8 :e6 97 :f7 :es White prepares :d5, an immediate

98 i.e3 :as 99 i.c5 :a4 1 0 0 @e3 33 :d5 :es is to no avail. Gener­

_:g4 10 1 i.d6 :g6 102 :n+ @h3 ally speaking White should ex­

103 i.e5 @g4 104 @e4 @h5 105 change rooks only if his king gets to

i.f6 @g4 106 :r4+ @g3 107 @e3 d5.

@h3 10s :rs :g3+ 109 @f2 :g2+ 33 @f7

11 O @fl :c2 11 1 :gs :c4 112 3 3 .. . : c 8 ! ? (planning 34 :d5 c3)

Ji.es @h4 11 3 :gs :e4 114 i.g3+ is answered by 34 @e2. After

@h5 11 5 @ f2 :a4 11 6 @f3 @h6 3 4 . . . @f7 35 :d5 :bs ( 3 5 .. . c3? 36

11 7 i.e5 :b4 11 8 .itf4+ @h7 11 9 @dl) 36 @e3 we transpose to the

:gs :a4 120 @g4 :b4 121 @f5 game.

:bs+ 1/2-1/2 34 :ds :bS

Certainly the rook isn't happy to

Commentary protect a little pawn. White's advan­

tage is big, but perhaps not big

The main feature of position 3 8 . 1 enough to win.

obviously is the passed pawn on d6. 35 @e3

The lt:Jd7 is heavily engaged with its With might and main the king

blockade. White has the advantage, makes his way to the centre. 3 5 f5?!

but Black holds sorne trump cards (to control e6) is weaker due to

as well: the control over the e-file 3 5 . . . @e8 and . . . lt:Je5, . . . @d7.

and the passed pawn on c4. 35 @e6

29 :es?! 36 :dl

By threatening a kingside attack White's threat is simple but

(with . . . :g5) Black forces his oppo­ strong: @d4 followed by : e l+ . Af­

nent to exchange queens. This is a ter e.g. 3 6 . .. :cs 37 @d4 @f7 38

clear misjudgement as White in­ :el :es 39 :xe8 @xe8 40 @d5

tends to exchange queens anyhow. Black can resign.

It ' s better to wait: 29 .. . @ f7 ! ? 3 0 f3 36 f5

�f5 3 1 �d5+?! 'ii'xd5 3 2 :xd5 :e5 Weakening e5, but seizing control

3 3 :d 1 ( after exchanging rooks the over e4. After 37 @d4 4Jf6 38

t0id6 is vulnerable) 3 3 . . . @e6 34 @f2 :el+ 4Je4 39 g4 g6 Black is be­

f5 3 5 f4 :e4. Black's rook and king yond the worst.

are much more active than in the 37 g4! [ 3 8 . 2 ]

game, he shouldn't have problems Creating huge problems for

then. Black.
 Endgame Secrets 125

42 :el+ @ f6 43 .id4+ @fl 44

:es) 42 : e l + (42 �d4+ @xd6 43

.ixc3 g6 isn't clear) 42 . . . @f6 43

.id4+ @fl 44 :es c2 45 .ib2 ttJf6

(45 . . . b4 46 :xd5 bxa3 47 .icl a2

48 :as :bl 49 :a?+ followed by

d6-d7) 46 :e7+ (Altematively,

White might try 46 g5 ttJe8 47 d7

ltJd6 48 g6+ hxg6 49 fxg6+ @f8 50

:e6) 46 .. . @f8 47 :c7 :d8 48 .ie5

ltJe8 49 :xc2 ltJxd6 50 :d2 ttJb7

(50 . .. @e7 51 .ixg7) 51 :xd8+

Diagram 3 8 . 2
ltJxd8 52 @e4 [ 3 8 . 3 ] and owing to
Black to move
his space advantage White will win.

37 fxg4

Black has to abandon his strong­

ho ld in the centre and allow a dan­

gerous pawn duo on the kingside,

but otherwise prospects are grim:

I) 37 .. . g6 3 8 g5 locks in the ltJd7.

There is still the nasty threat of @d4

and : e l + .

II) 37 .. . ltJ f6 38 d7! :d8

( 3 8 .. . ltJxd7 3 9 gxf5+) 3 9 :d6+ @fl

40 g5 ltJxd7 4 1 : d s ! (The simplest.

Playing for tricks with 41 @f3 ! ?

leads nowhere after 4 1.. .g6 and Diagram 3 8 . 3

42 . . . @e8) 4 1... @ e 6 42 :xb5 and Black to move

White wins. The �a3, the weak �c4

and the vulnerable kingside are too II) 39 . . . @d7 is better. Now 40

much for Black to cope with. :el can be answered by 40 .. . :e8

38 hxg4 lDf6 and after 40 g5 ltJe8 41 @e4 (4 1

39 f5+ :hl ltJxd6) 41...:b6 Black keeps

Later on, Jan Timman didn 't like the equilibrium by means of the

this move and proposed 39 @f3 ! ? threat .. . :c6 and .. . c4-c3, e.g. 42

instead. But Black will survive even @e5 (42 @d5 c3) 42 .. Jic6 43 f5 (43

then. Against the threat of :e 1 + : c 1 ltJxd6 44 :d1 c3) 43 .. . c3 44 f6

there are two moves : gxf6+ 45 gxf6 c2 46 :c 1 ltJxd6 4 7

I) 3 9 .. . ltJd5?! (somewhat frivo­ .ixd6 :xd6 48 f7 (4 8 :xc2 :e6+

lous) 40 f5+ @e5 4 1 �c5 (41 .iel 49 @f5 :e3 is also drawn)

look crushing, but Ruslan Siatdi­ 48 .. . :e6+ 49 @f5 :e2 50 f8=ltJ+

nov' s proposal 4 1.. . g 5 ! complicates (50 @f6 :t2+ 51 @g7 :g2+)

matters, e.g. 42 d7 :d8 43 .ig3+ 50 . .. @e7 51 ltJxh7 :e3 [ 3 8 .4 ] and

ltJf4 44 .if2 ltJd3) 4 1... c 3 ( 4 1... : c 8 Black captures the �a3.

1 2 6 Endgame Secrets

Black's position falls apart:

43 . . . hxg5 (43 . . . gxf6 44 gxh6 :h8

45 h7 f5 46 �c5 'Llf6 47 l:.h6 lt:Jxh7

48 @d5 results in a deadly pin) 44

fxg7 :g8 (44 . . . lt:Je3 45 @c5 lt:Jf5 46

l:.fl) 45 :g 1 ét:Jh2 46 l:.g2 lt:Jf3+ 4 7

@e4, White wins.

41 �f4 CtJh5+

Knights don 't like being at the

edge, but Black doesn't have access

to e4:

I) 4 l...�e4+ 42 @f3 loses at once.

Diagram 3 8 .4
II) 4 1... lt:Je4 42 :e 1 lt:Jf2
White to move
(42 lt:Jxd6 43 l:. d l ) 43 �xe8 lt:Jd3+

(43 @xe8 44 f6 gxf6 45 gxf6

39 �d7
lt:Jd3+ 46 @f5 lt:Jxb4 47 axb4 c3 48
40 g5 :es+
@e6 loses) 44 @e4 @xe8 45 i.d2
Moving the knight loses:
(45 f6? gxf6 46 gxf6 lt:Jxb4 4 7 axb4
I) 40 . .. 'Lle8 is too slow: 41 ©f4
c3 48 @d3 h5 and White lo'ses. 45
lk8 (41...l:.b6 42 : h l ) 42 @e5 c3
.1Lc3 ! ? �d7 46 @d5 lt:Jf4+ 4 7 @c5
43 f6 gxf6+ 44 gxf6 c2 45 l:.c 1 and
lt:Jh3 isn't completely clear)
46 f7 .
45 . . . @d7 46 @d5 and White wins
II) 40 . . . lt:Jg4+ 4 1 @d4 (otherwise
by 47 f6.
the knight gets to d3 via e5 or f2)
42 @g4 g6 [ 3 8 . 6 ]
4 1... � e 8 (41...l:.f8 42 ¡t e l ) 42 l:.hl!

h6 (42 . . . 'Lle5 43 :xh7 lt:Jc6+ 44

@c3 l:te3+ 45 @d2 �d3+ 46 @el

leads to a lost rook ending, 42 . . . lt:Je3

43 f6 lt:Jc2+ 44 ©d5 lt:Je3+ 45 @c5

isn't better) 43 f6 ! [ 3 8 . 5 )

Diagram 3 8 . 6

White to move

We arrive at another critical posi­

tion. White has to choose between

two tempting possibilities.

Diagram 3 8 . 5
43 fxg6
Black to move
 Endgame Secrets 127

Winning a pawn but not the game. ending. But now Black' s rook can

43 f6 is the main altemative. After operate along the h-file. The �c4 is

43 . . .l:!e4+ I can't find a win for protected d u e t o . . . :h4+.

White, though he has two passed . 48 @f3 líh3+

pawns on the sixth rank. White's 49 @f2 :h2+

king has two squares at his disposal: Rook and knight have woven a

I) 44 @h3 'Llf4+ 45 @h2 ( 45 @g3 net around the white king. The only

'Lle6 46 lífl líe5) 45 . .. 'Lle6 46 lífl way out of the checks is taking the

(46 líd5 @c6 47 l:íd2 @d7) �c4, but then the �d6 can be taken.

46 . . . líh4+ 47 @g3 l:Ih5 48 �d2 After 49 . . . c3 White shouldn 't play

l1Jxg5 49 f7 (49 �xg5 l:Ixg5+ 50 either 50 :xc3 :xc3 51 �xc3

@h4 líh5+ is a perpetual, since 5 1 @xd6 or 50 �xc3 �xd6 51 :es

@g4? :fs 52 líxf5 gxf5+ only trou­ @d7 52 :es ctJd6 53 @g2 l1Je4 54

bles White) 49 . . . 'Llxf7 50 :xf7+ :ds @c6. Instead, 50 d7 :d3 51

@xd6 with a probable draw. líxc3 líxd7 52 líc6+ @fl 53 �c5

II) 44 @f3 :es 45 :n (45 lí g l or leads to a highly difficult position

45 �d2 can be answered by 45 . . . c3 ) for Black; his king is driven back.

45 :fs+ 46 @g2 ctJf4+ 50 @el lí h l +

(46 :xg5+ 47 @h2 loses) 47 @h2 51 @f2

'Lle6 48 :xf5 gxf5 should lead to a It ' s very professional not to play

draw. He can give his knight as 51 @d2 immediately. Repeating

White has the wrong bishop, for ex­ moves gains time on the clock and

ample 49 f7 ctJf8 50 @g3 @e6 5 1 d7 gives a closer insight into the posi­

(5 1 @f4 @xf7 52 @xf5 @e8 and tion. Black has no other choice but

.. . @d7, 'Lle6) 51...'Llxd7 52 f8='iV to give checks.

'Llxf8 53 �xf8 @e5 and Black 51 líh2+

won't lose. 52 @fl : h 1+

43 hxg6 53 @g2 :dt

44 :ds 'Llg7! 54 @f2 l:Id4

The only chance. Black parts with 55 @el líh4

a pawn to bring the knight back into 56 @d2

play. 44 . . . :b8 45 :es loses at once. Finally Timman tums his atten-

45 líxb5 'Llf5 tion to the c-pawn.

Black has to reduce the enemy 56 l:Ih2+

pawns so that White will end up ei­ 57 @c3 :h3+

ther with the wrong bishop or with 58 @xc4 él)xd6+

an ending ": + � v :". For that The dangerous d-pawn is

purpose the knight can be sacrificed. eliminated.

46 líb7+ @e6 59 @es

47 :c7 líh8 During the game both players felt

Not 47 . . . 'Llxd6? 48 :c6 :d8 49 that 59 �xd6 @xd6 60 lía7 [38.7]

@f3 with transition to a lost pawn should lead to a draw.

128 Endgame Secrets

Diagram 3 8 . 7 Diagram 3 8 . 8

Black to move White to move

But Black has to play carefully to 60 @b5

pro ve that: 60 . . . l:rg3 ! ( 60 . . . @e5 61 After the loss of the Lsg5 White

l:ra5+ @f4 62 a4 loses; White sim­ can only win if he shields the a­

ply pushes his a-pawn) 6 1 l:ra5 ( 6 1 pawn against the sacrifice of the

l:ra6+ @d7 leads to draw, but not knight. I'm not sure whether the po­

6 1... @ c 7 ? 62 l:rxg6 l:rxa3 63 @d5 sition can already be considered to

@d7 64 @e5 @e7 65 l:rf6, because be a draw. 60 @b6!? ctJxg5 61 a4

the white king goes to g6) might be an improvement, after

6 1 . . . l:rg4+ 62 @d3 (62 @b3 l:rg3+ 6 1... l:r b 3 62 @b5 Black's rook

63 @b4 l:rg4+ and the king has to needs quite sorne time to get to the

retum) 62 . . . @ e 6 ! (It's important to eighth rank.

keep the king cl o s e to the g-pawn. If 60 étJxg5

the black king goes to the queen­ 61 a4 ctJf3

side, White might exchange Lsa3 and If 6 1 . . . ctJe 4 then 62 a5 ctJd6+ 63

Lsg6) 63 @e3 l:rc4 64 @f3 @d6 ! 65 @a4 threatening l:r c 6 .

a4 (65 l:ra6+ @e5 66 l:rxg6 @f5 67 62 a5

l:ra6 @xg5 is drawn) 65 . . . @e6 66 Other moves:

@e3 l:rb4 67 @d3 l:rg4 68 @c3 I) 62 jtc3 @d6 63 l:rc6+ @d7 64

l:rg3+ 69 @c4 l:rg4+. White can't go l:rxg6 l:rh5+ 65 @b6 ctJe5 66 jtxe5

on. His king can't cross the fourth (or else 66 . . .éLlc4+) 66 . . . l:rxe5 67 a5

rank while his rook has to guard l:r e l with a draw.

both the fifth rank and the a-pawn. II) 62 jtc5 ! ? looks promising, af­

59 ctJe4+ ter 62 . .. ctJd2 63 a5 ctJb3 64 a6 'Llxc5

[38.8] 65 l:rxc5 White is winning. The

black king is cut off along the sixth

rank and can 't support the g-pawn.

62 ... 'Lld4+

63 @c4
 Endgame Secrets 129

There are two other moves with @c7 �c6 77 .tf2 g3 ! 78 .txg3 @c5

the king that deserve closer 79 @b7 @b5 80 .tc7 @c5 8 1 .tg3

inspection: @b5 82 .t e l ltJd8+ with a draw.

I) 63 @a4 :h2 64 :c3 ( other 63 ... :h4

moves aren 't dangerous: 64 a6? 64 .tes

:a2+ 65 .ta3 :xa3+ 66 @xa3 On account of the many discov­

ltJb5+, 64 @a3? ltJb5+, 64 .ta3 :h4 ered checks White can't keep the

65 :c4 g5 66 .tb4 @d7 and . .. lt:Jc6 a-pawn. 64 ..te7 :f4 65 @d3 ét:Jb5

or 64 :c4 :a2+ 65 .ta3 lt:Jc2, with 66 l:Ib7 ltJd6, and Black secures the

a draw in either case) 64 . . . : h S ! ? draw.

(64 . . . :a2+ 65 .ta3 lt:Jc2 66 a6 64 ... ltJc6+

ltJxa3 67 :xa3 :d2 68 a7 :d8 69 65 @b5 lt:Jxa5

@b5 :as 70 @b6 @f5 71 @b7 Mission accomplished, Black has

:xa7+ 72 :xa7 loses) 65 a6 g5 eliminated the enemy pawn. At last,

with a probable draw, e.g. 6 6 @a5 one can have a clear conscience

g4 67 a7 :as 68 @b6 :xa7 69 when assessing the position as

l:Ie3+ @f5 70 @xa7 ltJc6+. drawn. My only problem now was

II) 63 @b6 :b3 64 @c5 ltJc2 65 the pace at which I had to play: I

.td2 l:I d 3 . Black forces the rooks had only twenty minutes left for all

off: 66 .t c l l:Ic3+ 67 @b6 l:Ixc7 68 the remaining moves.

@xc7 @d5 69 a6 ltJd4 70 @b6 ltJc6 66 @xa5 g5!

[38.9] Black quickly sacrifices the g­

pawn. As I knew the defensive

methods in an ending "l¡ + .t v l:I" I

decided to get rid of obstructing

"details".

67 @b5 g4

68 :g7 g3

69 :xg3 [ 3 8 .1 0 ]

Diagram 3 8 . 9

White to move

71 .tf4 g5 ! (Without g-pawn

Black would lose, but now it can be

sacrificed to get the king to either

d7 or b5) 72 .tg3 ( or 72 ..txg5 @d6

73 .tf4+ @d7 74 .tg3 ltJe7 75 @b7

Diagram 3 8 . 1 O
ltJc6 76 .tc7 ltJ a 7 ! ) 72 . . . g4 73 .th2
Black to move
lt:Je7 74 .tf4 lt:Jc6 75 .tg3 ltJe7 76
130 Endgame Secrets

Now we reach the aforementioned

endgame in its pure form. In Secrets

of Pawnless Endings John Nunn de­

votes 62 pages to this material con­

stellation, using Ken Thompson's

computerized analysis. But in prac­

tice you don 't need to have such an

enormous knowledge to survive. To

draw a position like the present one

it's enough to know two defensive

ideas: the Cochrane position and the

seventh-rank defence. In the further

Diagram 3 8 . 1 1
course of the game we will encoun­
White to move
ter both methods.

A brief remark conceming chess This is the Cochrane position. The

history: Thompson's analysis led to black king is forced to the edge, but

the conclusion that in certain posi­ White's bishop is pinned. As soon

tions the stronger side can force a as White's king unpins the bishop,

win in at most 59 moves. As a con­ Black's king flees to the other direc­

sequence FIDE expanded the tion. 8 1... @ h 5 ? 82 @f5 loses, as the

50-move rule to 75 moves in this bishop c a n ' t be taken.

-special ending. However, a few 82 @e5 @h3!

years later this rule was revoked, i.e . That' s the point. In the same way

after 50 moves without capture or 82 @e3 will be answered by

pawn move either side can claim a 82 . . . @ h 5 ! . 82 l::tg8 l::tb4 or 82 :d7

draw. @g4 don 't help White either.

69 :ht 83 .l:!gl l::tb4

It ' s usually a good idea to keep 84 i.e3 .l:!g4

the rook far away from the enemy 85 Ilal

king. White can 't prevent the black

70 @c6 :el king from escaping.

71 .td4 .l:!cl+ 85 @g2

72 .tc3 .:dl 86 .tf4 Ilg8

73 :e3+ @f5 87 :a2+ @f3

74 @c5 Ild8 88 .l:!a3+ @e2

75 .tes :es+ 89 @e4 :es+

76 @d5 .l:!a8 90 .tes .l:!e7

77 l::tf3+ @g4 91 �a2+ @el

78 l::tti Ila5+ Once again we encounter the Co­

79 @e4 !!a4+ chrane position, rotated by 90

80 .td4 @g5 degrees.

81 :g7+ @h4 92 @d4 @fl!

( 3 8 . 11 ] 93 .tf4 :e2
 Endgame Secrets 131

94 :as :e7 rook can't be taken on account of

95 @d3 @g2 stalemate and after 108 @e6 :g2

96 :f8 :e6 109 :r4 @g6 110 l;Ih4 :e2+ 1 11

97 :n :es i.e5 :g2 1 1 2 :h8 Ilg4 Black once

98 i.e3 :as again establishes the seventh-rank.

99 i.c5 :a4 106 ... @g3

100 @e3 :g4 107 @e3 @h3

This move doesn 't lose, but Or 107 . . . : g 8 .

100 . .. @g3 1 0 1 l;Ig7+ @h4 1 0 2 i.d4 10s :r5 :g3+

@h5 ! leads to the Cochrane position 109 @f2 :g2+

again. In a practica! game it is easier 11 0 @fl :c2

to stick to one defensive scheme 11 1 :g5 l:[c4

only. 112 i.e5 @h4

10 1 i.d6 :g6 113 :gs :e4

102 :n+ @h3 Or 1 13 . . . :g4.

103 i.e5 @g4 114 i.g3+ @h5

104 @e4 @h5 11 5 @f2 :a4

105 i.f6 @g4 Now the king has to stay at the

[38.12] edge. Instead, l 1 5 . . . :g4 and . . . @g6

is simpler. With the draw approach­

ing also my concentration faded.

11 6 @f3 @h6

117 i.e5 :b4

11 8 i.f4+ @h7

119 :g5 :a4

120 @g4 :b4

121 @f5 [ 3 8 . 13 ]

Diagram 3 8 . 1 2

White to move

Now Black applies the second im­

portant defence, the seventh-rank

defence.

106 l;If4+

After 106 :g2+ @h5 White's

rook has to leave the g-file and

Black's king can retum to g4. The Diagram 3 8 . 13

main defensive idea shows itself af­ Black to move

ter 106 :n @h5 107 �f5 :g5+ ! .

Giving check is the only move. The 121 ... :b5+?

132 Endgame Secrets

After playing this move I claimed @h4 1 2 8 :g7 @h3 1 2 9 :g3+ @h4

a draw because of the 50-move rule. 130 :d3 Wh5 13 1 i.d4 @h4 132

Isn't it ironic that a losing move se­ i.f6+ leads to a faster mate) 127

cures the draw? Strictly speaking :g7 :a6 128 i. c 7 ! (a surprising

Black should play 1 2 1 . . . :b6 (in­ move, now the rook can't go to a5

tending 122 i.e5 :g6 ! 123 :h5+ anymore) 1 2 8 . . . :c6 ( 1 2 8 . . . Wh6 1 2 9

:h6 1 2 4 :g5 :g6 1 2 5 :xg6 stale­ :d7 Wh5 13 0 :d3 :a4 13 1 i.d8 !

mate) or 1 2 1... : b l 1 2 2 i.e5 xn-, @h6 132 :ct7 :a6 1 3 3 i.g5+) 129

since after 121..Jlb5+ White can i.f4 :es+ 1 3 0 i.e5 lic6 1 3 1 :g5+

win: 122 i.e5 :b6 (Other moves @h6 (after 13 1... @ M 132 :g2

lose in an even faster way: White controls the c3 square) 132

1 2 2 . . . :a5 1 2 3 :h5+ :g8 124 @g6, :g3 @h7 1 3 3 :g7+ @h6 1 3 4 :d7

1 2 2 . . J � b 7 1 2 3 :h5+ @g8 124 :h8+ @h5 (or else 13 5 i.f4+) 13 5 i.d6!

@fl 1 2 5 :h7+, 1 2 2 . . . Wh6 1 2 3 :gl @h6 13 6 i.f4+ Wh5 137 :h7+ :h6

@h7 124 :g7+ @h6 1 2 5 :d7 :as 1 3 8 :xh6 mate.

1 2 6 :dl @h7 1 2 7 :d8) 123 lig7+

@h6 124 :g8 @h5 [ 3 8 . 1 4 ] Endgame 39

U .Andersson-C.Lutz

Pamplona 1 9 9 7

Earlier publication:

Informan! 71 /520.

Diagram 3 8 . 1 4

White to move

This is the most important win­

ning position, which has been ana­ Diagram 3 9 . 1

lysed by Philidor back in 1749. Black to move

Black's defence consists of checks

from the rear. By subtle manoeu­ Analyse the game.

vring White forces the black rook

from the b-file (where it is placed 35 ... 'Llf8 36 ctJd2 ét:Je6 37 'Lle4

best) to the a-file and then to the e­ 'Llxc5 38 'lixc5 :d6 39 a4 i.e7 40

file (where it is placed worst): 125 axb5 axb5 41 :a7 @ti 42 :b7

: g 2 :b4 1 2 6 :gl (a waiting move, :d5 43 Wg2 h5 44 :xb5 e4 45 @fl

Bl a c k ' s rook has to leave the b-file) :dt+ 46 @e2 :bt 47 @e3 g6 48

1 2 6 . . . :a4 (1 2 6 . . . :c4 127 :g5+ :b7 @e8 49 'lia6 :et+ 50 @d2

 Endgame Secrets 133

.l:lbl 51 t'i:Je7+ @d7 52 'Dd5+ @e6 Apart from a3-a4 White intends

53 .l:lxe7 @xd5 54 .l:ld7+ @e6 55 t'i:Jxf6+ and i.e7.

.l:lb7 @d5 56 .l:ld7+ @e6 57 �g7 37 t'i:Jxe5

.l:lxb4 58 .l:lxg6+ @f5 59 :e6 :b2+ 38 'Dxe5

60 :e2 .l:lb5 61 @e3 .l:lb3+ 62 @d2 Though the position is rather

:d3+ 63 @e2 h4 64 :es+ @f6 65 open, White's knight is stronger

:e6+ @g5 66 :es+ @f6 67 gxh4 than Black's bishop. It ' s the ¿je5

.l:lxh3 68 h5 @e6 69 :as @f6 70 h6 which is to blame.

@g6 71 .l:la6+ @h7 72 :e6 .l:la3 73 38 :d6

@fl :b3 74 @g2 :a3 75 :xe4 39 a4 i.e7?

@xh6 76 :g4 .l:lb3 77 f3 :b8 78 Black only needs the additional

@g3 .l:la8 79 @h4 .l:la3 t/2-1/2 move .. . :c6 to be out of trouble.

But it' s White' s turn and he can

force the win of a pawn. Anyway

Commentary
Black: is already hard-pressed:

I) 3 9 . . . :b6 40 axb5 axb5 avoids

The diagrammed position doesn't
losing a pawn, but after 4 1 :a7 fol­
promise anything good for Black.
lowed by @g2-f3-e4-d5 Black's
The ¿je5 is weak and blocks the
passive pieces give him a big head­
black minor pieces, while the white
ache. Please note that the black
knight is ready to jump to e4. Sup­
bishop is placed awkwardly, . . . i.g5
ported by the i.c5 and the :a2
(with the threat of . . . i.d2) can be
White might start an attack at the
countered by 'De4.
queenside with a3-a4. If you add
II) 3 9 .. . bxa4 is the appropriate
Ulf Andersson's well-known end­
move. Black immediately sheds the
game technique it's easy to under­
¿ja6 in return for sorne activity: 40
stand why I was rather pessimistic.
.l:lxa4 .l:ld4 41 'Dxa6 i.e7. Now I
35 'Df8?
don't see anything better for White
Black wants to take on c5, thus
than 42 .l:la5 i.xb4 43 �xe5 i.d6 44
destabilizing the i. c 5 . But now it' s
.l:la5 ( 3 9 . 2 ]
the white knight which comes to c 5 .

In view of the weakness a6 this is

rather unpleasant. Better is 35 . . . e4

3 6 CDd2 .l:le8. The white knight can't

go to e4 and Black' s pieces beco me

active, . . . 'De5-d3 is already threat­

ened. After 37 @ g 2 i. g 5 ! ? 38 h4

i.h6 Black intends trading off his

weak e-pawn with . . . e4-e3. If then

39 'Dfl or 39 t'i:Jb3 Black has

3 9 . . . 'De5 and . . . 'D d 3 . Black has an

almost equal game.

36 'Dd2 t'i:Je6

37 t'i:Je4 Diagram 3 9 . 2

Black to move
1 3 4 Endgame Secrets

White will have to allow the ex­ II) 43 . . J![b2 is better. 44 éZJd7 (if

change of knight and bishop. In the White wants to achieve anything

resulting rook ending White can substantial he has to keep the minor

play on for many more moves, but pieces on the board) 44 . .. il.xb4 ( or

without any realistic winning 44 . . . il.d6 45 l:.b6 il.xb4) 45 l:.b8+

chance. In a bad position it's sorne­ @f7 46 'lJxe5+ @e7 (46 @e6 47

times suitable to give up material in éZJd3 l:.b3 48 éZJf4+ or 46 @f6 4 7

order to transpose into a defensible l:.xb5 il.c3 48 'lJg4+ result in the

ending. Compare the note to the same type of position) 4 7 l:.b7+ @f6

next move. 48 l:.xb5 il.c3 49 t2Jg4+ [ 3 9 . 3 ]

40 axb5

The position after Black's 39th

move has been played out in a train­

ing game between Christian Wil­

helmi and Amd Lauber. Both

players had one hour to finish the

game. Wilhelmi chose the weaker

40 @ g 2 ? ! , after which Black liqui­

dated to a draw in the aforemen­

tioned way: 40 . . . bxa4 4 1 l:.xa4 l:.d4

42 'lJxa6 @f7 43 @f3 ! ? (White

plans 44 @e3 and f2-f4 to uproot

the l:.d4. 43 .:.as il.xb4 44 l:.xe5

il.d6 leads to the endgame we al­ Diagram 3 9 . 3

ready looked at) 43 . . . h5 44 l:. a 5 ! ? Black to move

(44 @e3 h4) 44 . . . il.x b 4 45 l:.xe5

il.d6 46 l:. x h 5 ! ? (leading to a draw Once again we encounter an end­

as well) 46 . . . �a4 47 t2Jc5 .:.as 48 ing "3� v 2� on the same side", but

'lJe4 l:.xh5 49 t2Jxd6+ @e7 50 @g4 this time with minar pieces. This in­

l:.e5 and sorne moves later the game creases White' s winning chances as

was drawn. White has no more win­ knights are stronger than bishops

ning chances. when fighting on one wing only.

40 axb5 Nevertheless Black should be able

41 l:.a7 @f7? to draw. However he shouldn't wait

This move is as bad as it looks. passively but instead strive to place

lt ' s n o t only that Black pins himself, his pawns on g6 and h 5 . For a better

it' s also the fact that the @f7 can be understanding of the position I refer

checked from d6, e5 or g 5 . 4 l...il.f8 to the chapter "l:. + 'lJ + 4� v l:. + .i.

is the natural choice: 42 l:.b7 l:.d2 + 3� on the same wing" of Winning

43@g2 Endgame Technique, of special in­

I) 43 . . . il.xc5 44 bxc5 l:.c2 terest are the games D. Y anofsky­

(44 . . . l:.b2? 45 c6) 45 l:.xb5 leads to R.Fischer (Stockholm interzonal

a difficult rook ending, White will 1962) and A.Beliavsky-A.Yusupov

continue with l:. b 8 - c 8. (USSR 1 9 8 0 ) .

 Endgame Secrets 135

same time. Black shouldn 't lose

then.

43 h5

Black prepares . .. h5-h4 to either

trade off pawns or to fix the c0i h 3 .

Otherwise it' s difficult to suggest a

constructive move.

44 :xb5?!

White pockets an extra pawn, but

I wonder whether he (in analogy to

43 @g2!) can improve his position

even further. In this respect 44

@f3 ! ? or 44 h4 ! ? should be pre­

ferred. After the text the e-pawn can

advance and Black gets sorne

Diagram 3 9 .4
activity.
White to move
44 e4

The rook looks rather clumsy on 45 @fl

d5, but actually Black sets a little Now 45 . . . :es can be defused by

trap. 46 @e2, therefore Black plays

Altematives: differently.

I) 42 . . Jídl 43 :xb5 .:.bl 44 :b7 45 .:.dt+

leaves White a pawn up. The line 46 @e2 nbt

44 . . . e4? 45 ltJxe4 nxb4? 46 t2Jd6+ . . . .i.xc5 isn't a threat yet (the c0ih5

shows the bad placement of the is unprotected), but . . . .:.b2+ is in the

@r/. air.

11) 42 . .. :d2 43 @g2 :b2 44 t2Jd3 47 @e3

and once again the @f7 is utterly Should lead to a draw only, but

mis placed. other choices aren't more

43 @g2! promising:

White looks through his oppon­ I) 47 h4 líb2+ 48 @e3 g5 ex­

ent' s intentions. Andersson acts ac­ changes sorne pawns.

cording to the principle "Don't 11) 47 nb7 @e8 48 .:.b6 líb2+ 49

hurry", since Black can't do much �e3 .i.g5+ isn 't very effective ei­

anyway. 43 :xb5? e4 allows black ther. The white kingside is

counterplay with . . . :es and . . . e4-e3. vulnerable.

Exchanging the e- and f-pawns is 47 g6?

favourable for Black as then the But this move is way too passive.

white kingside is vulnerable and the I) 47 . .. líxb4 48 :xb4 .i.xc5+ 49

position is opened up completely. :d4 is just lost.

Along the diagonal b8-h2 the bishop 11) 47 . . . .i.xc5+ 48 :xc5 nxb4 49

can become active, it can stop the :xh5 followed by h3-h4 and

b-pawn and attack the c0ig3 at the l::tg5-d5-d4 is won for White as well.
1 3 6 Endgame Secrets

III) 47 . .. h4 is the right move: 48 50 .:bl

l:.b7 (4 8 gxh4 �xc5+ 49 l:.xc5 The intention is 51..J:ib2+ 52

l:.xb4 or 48 g4 �xc5+ 49 l:.xc5 @e3 �g5+.

l:.xb4 both lead to a drawn rook 51 'LJc7+ @d7

ending. Black will exchange the Lse4 It looks strange to move the king

against the Lsh3) 4 8 . . . @e8 49 ¡[b6 into a discovered check, but Black

(what else ?) 49 . . . l:. h l 50 'LJxe4 takes away the e6 square from the

hxg3 51 fxg3 l:.xh3 and Black at­ knight. During the game I consid­

tains the draw with little effort only. ered 5 1.. .@d8 as too passive:

48 l:.b7 @e8 I) 52 'LJd5 �g5+ 53 @e2 (53 @c3

[39.5] h4 54 g4 l:. h l ) 53 . . . h4 54 g4 l:.b3 55

l:.g7 l:.d3 56 'LJe3 l:.d6 gives Black

good chances to save the game.

II) 5 2 'LJe6+ is stronger. 52 . . . @e8

53 b5 �b4+ 54 @e3 (54 �e2?

l:. e l # or 54 @c2 l:.fl 55 @b3 �d6

are worse) 54 . . . l:. e l+ 55 @d4 l:.e2

56 'LJg5 l:.xf2 57 'LJxe4. Though

Black managed the favourable ex­

change of the Lse4 and the fü2, he is

still much worse in view of the ex­

cellently placed white pieces.

52 'LJd5+

White transposes to a rook énding

Diagram 3 9 . 5
which seems to be drawn. But I
White to move
don't see anything better.

49 'LJa6? 52 @c6

There is the well-known saying 53 l:.xe7 @xd5

about the knight at the edge. In this [39.6]

case it proves valid once again,

Black gets the breather he desper­

ately needs. Instead, 49 l:.b6 wins.

Black can't play neither 49 . . . l:.xb4

50 l:.xb4 �xc5+ 51 l:.d4 nor

49 . . . �xc5 50 bxc5, and after 49 . . .

l:. e l+ 50 @d4 ( or 50 @f4) the loss

of the Lse4 can 't be avoided.

49 l:. e l +

50 @d2

If White advances the king, his

kingside pawns become vulnerable

after both 50 @d4 �f6+ and 50 @f4

Diagram 3 9 . 6
�d6+, followed by . . . e4-e3.
White to move
 Endgame Secrets 13 7

54 l:[d7+ Usually a rook ending "3¿j v 2¿j

Black has to watch both the ¿jb4 on one side" is drawn comfortably.

and his kingside. For that reason he But as his pawn structure is tom

shouldn 't allow his king to be cut apart Black still has to be careful. It

off. helps him that the drawish endgame

54 @e6 "l:[ + f-pawn + h-pawn v l:[" might

It is weaker to move the king arise.

elsewhere: 59 .:.c6 .:.b2+

I) 54 . . . @e5 55 @c3 followed by An impulsive check in time trou­

l:[d2 wins, as the @e5 is cut off ble, though it doesn 't do any serious

from the ¿jb4, damage. Altematively Black can

II) 54 . . . @c6 55 .:.d4, and White play 59 .. . .:.bs:

wins again. This time the king is cut I) 60 h4 .:.b2+ (now this check is

off from the kingside. possible) 6 1 :c2 ( or 6 1 @e3 .:.b3+

III) 54 . . . @c4 isn't too promising 62 @e2 .:.b2+ 63 @fl e3 64 fxe3

either, e.g. 5 5 h4 .:.xb4 5 6 @e3 @c5 @g4 65 .:.g6+ @f3) 6 1 . .. .:.xc2+ 62

57 .:.ct2 .:.a4 58 @f4 .:.d4 59 .:.e2 @xc2 @g4 63 @d2 @f3 64 @e 1 e3

and@g5. 65 fxe3 @xg3 leads to a draw.

55 .:.b7 @d5 II) 60 .:.c3 h4 (the ¿jh3 is blocked

55 .. . @e5? abandons control over mechanically) 61 @e3 @e5 62 g4

c4: 56 @c3 .:.n 57 .:.n e3 5 8 .:.e7+ ( after 62 gxh4 @f5 Black succeeds

with a win. in exchanging his e-pawn against

56 .:.d7+ @e6 one h-pawn) 62 .. . .:.as 63 f4+ exf3

57 .:.g7 .:.xb4 64 @xf3 (The ¿jh4 appears weak,

57 . .. @f6? 58 .:.b7 @e5 leads to but Black can defend) 64 .. . .:.a2 65

the previous note. .:.es+ @f6 66 .:.hs .:.a3+ 67 @g2

58 .:.xg6+ @f5 @g7 68 .:.xh4 .:.as [ 3 9 . 8 ]

[39.7]

Diagram 3 9 . 8

White to move

Diagram 3 9 . 7
The .:.h4 won't see the light of
White to move
day again. If White's king
1 3 8 Endgame Secrets

approaches, Black checks from the don't see a win for White, e.g. 66

side. .U.c6+ @g7 67 .U.e6 l:. e l 68 á'.a6 á'.bl

60 á'.c2 .U.b5 69 .U.a4 .U.e 1 70 .U.a3 l:.b 1 71 .U.e3

61 @e3 l:Ib4.

Easing Bl a c k ' s d e f e n c e . In my 61 ... l:.b3+

earlier analysis I considered 6 1 h4 ! ? 62 @d2 l:.d3+

to be more poisonous. White plans 63 @e2 h4

playing .U.c3, @e2-fl-g2 and then Trading off another pair of pawns

.U.c6-h6, winning the � h 5 . Black has secures the draw.

to watch out: 6 1.. . .U.as 62 .U.c3 .U.b5 64 .U.es+ @f6

63 @e2 (63 @e3 l:.a5 doesn't help) 65 · l:.c6+ @g5

63 .. . .U.aS 64 @fl [39.9] (64 .U.c6 or 66 l:.c5+

64 .U. e l permits 64 . . . .U.a2+ 65 @fl Or 66 f4+ @f5 67 .U.c5+ @f6.

e3 66 fxe3 @g4) 66 @f6

67 gxh4 l:.xh3

[ 3 9 .1 0 ]

Diagram 3 9 . 9

Black to move

Diagram 3 9 . 1 0

Now Black shouldn't allow his White to move

kingto be cut off from his pawns.

I) 64 . .. .U.bS? is a mistake. 65 @g2 68 h5 @e6

.U.as (65 . .. .U.b2 66 l:.c5+ @g6 67 69 l:.a5 @f6

l:Ig5+ @h6 68 @fl and .U.e5, 70 h6 @g6

@e2-e3, 65 .. . l:Ibl 66 l:Ic6 intending 71 .U.a6+ @h7

.U.h6) 66 l;:íc6 @e5 67 l:.h6 @d4. The 72 l:.e6 l:.a3

rook protects the �h5, but the king 73 @fl I;:íb3

is cut off. Whit may put his rook to 74 @g2 .U.a3

f8 and play g3-g4, @g3. Black is 75 .U.xe4 @xh6

facing difficulties. 76 .U.g4 l:.b3

II) 64 .. Jíal + (the correct choice) 77 f3 .U.b8

65 @g2 @g6. As now White's king 78 @g3 .U.as

is trapped the black king retreats be­ 79 @h4 l:.a3

fare White continues with .U.c6-h6. I ½-½

 Endgame Secrets 13 9

Endgame 40 i.bl i.h5 64 l:.el i.f2 65 xn i.e5

C.Lutz-A.Karpov 66 �el i.e2 67 i.a2 �h3+ 68 @b2

Dortmund 1 9 9 3 i.a3+ 69 @al i.d3 70 ..tbl i.b4

Earlier publications: SCHACH 6/93, 71 jácl i.d2 72 l:.dl i.e3+ 73 @a2

lnformant 57 /492 (A.Karpov), i.xe4+ 7 4 @a3 i.e2 0-1

NIC Magazine 4/93 (A.Karpov).

Commentary

It ' s well-known that young chess

players tend to be rather optimistic.

This exaggerated self-confidence is

based on a strong belief in one's

own abilities and an underestima­

tion of the opponent. Only through

defeats against stronger players can

one's own limitations be fully un­

derstood and youthful energy

change into a more mature under­

standing of the game. The present

Diagram 4 0 . 1
game is one of those painful defeats.
Black to move
After having understood about my

complete ignorance I learned a lot

Analyse the game.
from the analysis. ·

2I..J::txf4 22 gxf4 i.xf4+ 23 �bl After 21 moves a position arose

'ii'e5 24 'ii'xe5 i.xe5 25 h3 a5 26 b3 that I assessed as quite promising. I

d6 27 l:.d2 i.e8 28 @e2 i.g6 29 envisaged lines like 21...i.xf4 22

�df2 l::te8 30 @dl i.d4 31 l::ta2 gxf4 �xf4 23 e5 g6 24 i.xg6 with a

.:.es 32 l:.g2 l::th5 33 .tlf3 i.e8 34 crushing attack. However, Karpov's

@e2 g6 35 i.e2 l::te5 36 i.d3 @g7 reply put an abrupt end to my

37 .l:!g4 g5 38 .:.n i.e5 39 @b2 dreams.

i.g6 40 h4 gxh4 4 1 .tlxh4 �g5 42 21 �xf4!

jáh2 jág3 43 i.e2 i.d4+ 44 @el a4 Effectively stopping White's

45 bxa4 á'.xa3 46 @d2 �g3 47 i.d3 initiative .

.tlg5 48 .:.ful i.e5 49 �h3 i.d4 50 22 gxf4 i.xf4+

já3h2 l::tg3 51 xn i.f6 52 .:.bl 23 @bl �e5

i.g5+ 53 �e2 i.e3 54 :al i.gl 55 Centralizing the queen. White can

:d2 �f6 56 a5 bxa5 57 �xa5 i.e5 hardly avoid an exchange.

58 : a 1 @e5 59 :n i.gl 60 l:.ddl 24 'iVxe5 i.xe5

..te3 61 .:.f8 .:.g2+ 62 @b3 l:th2 63 [40.2]

1 4 0 Endgame Secrets

becomes inevitable. In the course of

the game White only looked for his

own resources, completely forget­

ting about his opponent.

25 h3

Advancing the attacked pawn

weakens the g3 square and denies

the white bishop the h3 square. This

detail is rather insignificant, but it

might become more important,

compare the note to White's 27th

move, line II. After 25 :d2 play can

Diagram 4 0 . 2
develop in a similar way, e.g.
White to move
25 . . . d6 26 :df2 ii.f6 27 :g2 ii.d4

and . . . ii.e8-g6. After a possible

N ow it' s time to take stock. How . . Jk8-c5-h5 White might be forced

to assess the present position? In re­ to advance the h-pawn anyway.

turn for the sacrificed exchange 25 a5!?

Black has a pawn and a black­ Apparently 26 . . . a4 is a threat . . .

squared bishop which has no coun­ 26 b3?!

terpart. Furthermore the ii.d3 is a . . . and White gives in to the threat.

sorry sight, it is boxed in by the I wanted to keep the option of play­

pawns on c4 and e4. Therefore ing b3-b4, but this move only opens

Black obviously has compensation up the a-file for the black rook. On

for the slight material deficit. This the other hand 26 b3?! has serious

had been my assessment during the drawbacks: the ii.d3 is restricted

game. However, I didn 't realize the even further and White loses control

fact that Black even has winning over the squares c3 and a3. The

chances. I expected Black to ex­ threats . . . .i.c5 or . . . ii.b2 are in the

change rooks with . . . :f8 and then to air. Preferable is 26 :f3. The 8b2

put his king on e 7 . Black then occu­ can easily protected by the king and

pies a fortress. But Karpov demon­ . . . a5-a4 shouldn't be feared, viz:

strates that Black can avoid the 26 . . . a4 27 l:. d fl . Black tries to acti­

exchange. His rook can become ac­ vate his bishop by means of . . . d7-d6

tive via the fifth rank while White's and . . . ii.e8, but then the 8a4 proves

rooks don't have anything to bite susceptible to attack. If Black acti­

on. vates his rook via a 5 , he c a n ' t avoid

It is interesting to see the white playing . . . d7-d6 on account of his

position deteriorate until defeat weak back rank: 27 . . . ..t.f6 28 :g3

 Endgame Secrets 141

(threatening l¡xf6) 28 . . . @ h 8 ! ? his structure is ruined completely)

(28 . . . .td4 29 l¡gf3) 29 l¡e3 :as? 30 28 . . . .tg6 29 @d2 .td4 30 e5?!

e5 and only Black has problems. .txd3 3 1 @xd3 .txe5 32 :bel h6.

26 d6! After . . . :es, . . . .tf6, . . . @f7-e7 Black

[40.3] is clearly better. Thus 3 0 e 5 ? ! is no

feasible continuation.

II) 27 kidel ! ? (White plans

.te2-g4, to pressurize against e6.

Playing . . . e6-e5 would be a major

concession by Black as this reduces

the black-squared bishop's sphere of

activity) 27 . . . .te8 28 .i.e2 .lg6 29

.i.g4 :es 3 0 @a2 (30 h4 only ere­

ates a weakness after 3 0 . . . h5 and

. . . .tf6) 3 0 . . . h5 31 .te2 @h7. Black

still enjoys the initiative, but he

committed himself to playing

. . . h7-h5. Later on we will appreciate

Diagram 4 0 . 3
the power of the construction b.lg6
White to move
+ l::ih7 . . Looking at this line it makes

A very strong move. According to sense to reconsider 25 l¡ d 2 ! ? .

Karpov's commentary in NIC White's bishop might then use the

Magazine there are three important h3 square.

considerations: 27 .tes
1 . On the queenside, the .lc6 has 28 @e2 .tg6

no future. Neither . . . a5-a4 nor 29 ¡[df2 :es

. . . b6-b5 are viable options. 30 @dl?!

2. After the transfer to g6 the Here the king is misplaced. There

bishop can pressurize against might arise a pin along the diagonal

White's centre and kingside. d l - h 5 and the l::ia3 needs protection.

3. The E::ie6 can't be attacked as That's why 30 @ b l is stronger.

the .td3 is tied to the defence of the 30 .td4

l::ie4. 31 f! a 2 ? !

27 kid2 Another move short of talent.

In the further course of the game White anticipates . . . .i. c 5 , but the

White remains passive, therefore main threat is . . J�c5. 3 1 ¡[g2 is bet­

one might wonder about more ac­ ter and only after 3 1... .t c S White

tive continuations: replies 32 �a2.

I) 27 @c2 (White centralizes his 31 :es

king. By means of e4-e5 he sacri­ 32 l¡g2

fices his weak pawn and exchanges Forced in view of 32 b4? �g5

bishops) 27 . . . .i.e8 28 libl (To pre­ with a clear black advantage (Kar­

vent . . . .tb2. Of course White can pov). At this moment I offered a

play a3-a4 at any moment, but then draw. Karpov refused without
142 Endgame Secrets

saying a word and I finally under­ h6 3 8 _¡¡g2 ( 3 8 .&(f8+? @h7 39 �d8

stood what's going on. .&(xh3 40 �xd6 .i.b4 is clearly ad­

32 �h5 vantageous for Black) 3 8 . . . �g5

33 .&(f3 [40.4] (38 . . . .i.e8 39 .&(gf2, 3 8 . . . @h7 39

.&(fg3) 39 líxg5 hxg5 40 �g3 and

after the loss of the g-pawn neither

side can win.

34 @c2 g6

35 .i.e2

3 5 �g4 intending .a'.gf4-f8 has to

be considered. Black might continue

with 3 5 . . Jk5 36 :.(gf4 .i.g7 and

. . . g6-g5.

35 �es
36 .i.d3 @g7

37 �g4

On account of .&(gf4 Black is

Diagram 40.4
obliged to play . . . g6-g5, but that's
Black to move
his idea anyway. Therefore 37

33 .i. e 8 ! ? .&(fl ! ? (intending 3 7 .. . g5 3 8 h4) is an

Karpov chooses a plan which in­ option.

volves the advance of the g-pawn, 37 g5

that's why the .i.g6 has to leave its 38 �fl

post. Black has other choices, but 3 8 h4? .ih5 isn't feasible, but 3 8

against accurate defence these tries .&(fg3!? (Karpov) eases White's de­

aren 't more effective: fence considerably: 3 8 . . . h6

I) 3 3 . . . .&(e5 (attacking the �e4) 34 ( 3 8 . . . @f6 39 h4 gxh4? 40 �f3+ @e7

�fg3. White pins the .i.g6 and in­ 4 1 �g7+ only troubles Black) 3 9 h4

tends exchanging rooks with .&(g5. .i.h5 40 �f4 [ 4 0 . 5 ] .

This would be a major success for

White. Now 34 . . . i.c5 35 a4 .:.h5

can be answered by 36 e5 .ixd3? 3 7

�xg7+ and mate.

II) 3 3 . . . .a'.h4 34 �g4 ¡( h 6 ! ?

(threatening . . . .i.h5, but White can

exploit the weak back rank) 3 5 ¡(gf4

(with a mating threat) 3 5 . . . .i.f6 36

@c2 and Black didn't achieve

anything.

III) 3 3 . . . .i.c5 (provoking a3-a4,

but White can create counterplay)

34 b4 axb4 3 5 axb4 .i.xb4 3 6 .&(b2

.i.a5 (36 . . . .i.c5 3 7 �a2 h6 3 8 ¡(a8+ Diagram 40.5

@h7 39 ¡(ff8 even loses) 37 ¡(bf"2 Black to move

 Endgame Secrets 143

White will play 41 hxg5 next III) 40 :g3 is playable: 40 .. . h6 4 1

move, after which Black has to ex­ :gf3 (White has to seek counterplay

change rooks or recapture with the against the enemy king. Both 41

h-pawn. In the latter case the �g5 kig4? d5 and 4 1 h4? gxh4 42 :g4

only serves to block the lle5. White kih5 43 :hl (43 e5 llg5) 43 .. . h3 44

isn't worse then. e5 :gs 45 :xg5 hxg5 46 :xh3

38 .i.c5 j_xd3 47 :xd3 dxe5 lead to a clear

39 @b2?! edge for Black) 41....i.xe4 (Other

Karpov didn't like 3 8 .. . .i.g6 3 9 b4 moves aren't better: 4 1.. . j_ h 5 42

as then the :es is boxed in. But l:.g3, 4 1... h 5 42 :g3 h4 43 kig2,

now White has another chance of 41...j_d4+ 42 @c2) 42 l:.±7+ @g6

playing h3-h4: 3 9 h4 h6 (39 .. . .i.xa3 43 :lf6+ @h5 44 j_xe4 (Not 44

40 :xg5+ :xg5 41 hxg5 poses no j_e2+? @h4 45 :xh6+ @g3 and

problems) 40 hxg5 hxg5 41 @b2 suddenly the black king is out of

and Black can't make any progress. danger while his opposite number is

39 .i.g6 [40.6] in trouble) 44 .. . :xe4 45 @c2

(threatening :h7) 45 .. . j_d4

(45 .. . :e2+ 46 @ d l ) 46 @d3 d5 47

cxd5 exd5 48 l:.c6. D u e t o the threat

of :h7 White keeps the balance.

IV) 40 j_ b l is the most solid

move. Though White' s position

looks rather cramped, the game is

about equal. 40 .. . j_d4+ (40 . .. h6 4 1

h4 gxh4 42 :xh4 is better for White

in comparison to the game. Black's

pawn is already on h6 which is the

reason that both j_g6 and �h6 need

protection. The game is equal) 41

Diagram 40.6 @el j_e3+ 42 @b2 �c5 (Black

White to move makes use of zugzwang ideas) 43

:g3 (43 :hl is inferior. A fantasy

40 h4? line which demonstrates the activity

White's previous play has been of Bl a c k ' s bishops: 43 .. . .i.h5 44

quite pointless, but this move is just :g3 j_d4+ 45 @c 1 j_f2 46 l:.d3

wrong. White ignores the threat of j_c5 47 @b2 j_e2 48 l:.g3 j_f2 49

40... d5. There are better ways to l:.g2 j_f3 with a win. Of course

defend: White may play 47 a4, but then his

I) 40 j_c2? d5 41 cxd5 exd5 42 queenside is fixed permanently)

exd5 :e2 43 :c4 (or 43 : e l j_e3) 43 .. . h6 44 :g4 and I don't see how

43 .. . .i.d3 44 :c3 .i.d4 and White Black can go on. After 44 .. . j_ d 4 +

loses material. 45 @el j_e3+ 46 @b2 .i.f4 4 7 h4

II) 40 : e l ? .i.h5 4 1 :g2 j_f3 fol­ h5 48 :gxf4 gxf4 49 :xf4 the re­

lowed by .. . d 5 isn't worth it either. tum of the exchange yields equality.

144 Endgame Secrets

40 gxh4 B) 43 .. . i..e3 wins at least a pawn:

41 l:[xh4 l:[ g 5 ? ! 44 l:[fg4 (44 :D ..tg5; 44 l:[f5 ..txf5

On the wrong assumption that he 45 exf5 h6) 44 .. . ..tg5 45 :h2 h5.

hasn't yet reached the time control II) 43 :el i..f2 loses the

Karpov chooses only the second­ exchange.

best move. Activating the rook leads III) 43 l:[f5 ! ? (Intending to retum

to a certain black advantage, but the the exchange and trade off the f­

better choice is 4 1... d 5 (Karpov). In pawn against the h-pawn after­

the lnformant Karpov confines him­ wards. The resulting endgame

self to the assessment "Black is win­ should be defendable then. But this

ning", But playing .. . d6-d5 sets the idea can't be realized) 43 .. . i..d4+

..td3 free, therefore one has to be (stronger than 43 .. . ..txf5 44 exf5 af­

sure that Black gets something sub­ ter which White reaches the afore­

stantial (i.e. a pawn) in retum. It is mentioned endgame) 44 @a2 (44

important to calculate sorne lines: @b 1 ..txf5 45 exf5 :e 1 + and

42 cxd5 exd5 [40.7] .. . ..t f6 . White is lost as he can't win

the i0ih7. 44 @c2 dxe4 or 44 @el

.ixf5 45 exf5 .ie3+ and .. . h7-h6 are

no improvement) 44 .. . ..txf5 45 exf5

.ic3 46 b4 (46 f6+ @xf6 47 l:[xh7

:el 48 b4 a4 49 ..tbl :e2+)

46 .. . :e3 47 f6+ @xf6 48 ..txh7

@g5. Black remains two pawns to

the good, he is winning.

IV) 43 i..c2 looks relatively best.

43 . .. dxe4 44 l:[ e l e3 (44 .. . ..tf2 45

l:[hxe4 i..xe4 46 :xe4. Trading off

rooks leads to a draw and 46 .. . l:[h5

47 :e7+ @g8 48 b4 isn't promising

Diagram 4 0 . 7 either) 45 .ll.xg6 hxg6 46 b4!?. In

White to move my opinion White should be able to

defend though Black enjoys a dis­

.. . and now: tinct advantage. It seems to me that

I) 43 l:[ff4 here Black' s position is better than

A) 43 . .. ..te7? leads nowhere after in the game, thus 4 I.. . l:[ g 5 ? ! 1s

44 l:[hg4 @h6 (44 h5 45 :xg6+ inaccurate.

@xg6 46 exd5+, 44 i..g5 45 l:[f5) 42 l:[h2 :g3

45 @c3 ..tg5 46 l:[fl dxe4 47 @d4 43 ..tc2 .id4+

and only Black can be worse. 44 @el [ 4 0 . 8 ]

 Endgame Secrets 145

Now White has three weak

pawns, but trading off pawns usu­

ally favours the defender. Further­

more White might either attack the

8b6 by means of �b 1 or exchange it

by �al and a4-a5.

46 @d2 �g3

47 i.d3 _:g5

Black manoeuvres to provoke ad-

ditional concessions.

48 �fbl

Now and later on White doubles

Diagram 40. 8
rooks along the h-file. If the black
Black to move
king goes to f6, White might retum

the exchange on h7.

We enter a new stage of the game. 48 .tes

Black's rook invades the enemy 48 .. . @ f6 ! ? 49 .:xh7 .txh7 50

camp and can attack the weakened .:xh7 offers little winning chances:

white queenside. Take a look at the 50 .. . l:tg2+ 5 1 i.e2 ©e5 52 a5 ! ? (It

construction b.tg6+8h7: Pawn and is important to trade off pawns. Oth­

bishop protect each other, that' s the erwise White still has difficulties)

reason why Black shouldn't hurry to 52 .. . bxa5 53 .:h5+ �xe4 54 �h4+

play 44 . . . h 5 . After 45 � fh l �h6 46 @e5 55 �h5+ �f4 56 �h4+ @e5 57

.tdl Black is forced into defence. .:h5+ �f6 5 8 .:xa5 .tc5 (58 . .. i.c3+

Though the i.g6 attacks the 8e4 the 59 @xc3 :xe2 60 �a6 @e7 61

pawn can't be taken: After e.g. .:a7+ 'it>d8 62 �h7 is drawish as

. . . l:te3 and . . . .txe4 White exchanges well) 59 l:ta8 @f5 60 .:h8 [40.9]

bishops and the combined attack of and White checks from the side.

the two rooks wins the l!ih7. On the Black shouldn't win.

other hand White can't simply play

e4-e5, as then after . .. .txe5 a ex­

change of bishops Black recaptures

on g6 with the pawn. Of course

Black likes to win the 8e4, but only

if he doesn 't have to make any con­

cessions ! Black has to look for other

ways to improve his position. He

would like to bring the king to e5,

but the .:n prevents this undertak­

ing. Therefore Karpov creates new

targets to attack.

44 a4
Diagram 40. 9
45 bxa4 l:txa3
Black to move
146 Endgame Secrets

Certainly Karpov pondered on :xh7 i..xh7 56 :xh7 @e5 and

48 .. . @f6, but decided that further . . . @d4 is rather awkward) 55 i..f4

manoeuvring might tire the enemy 56 :h3 (56 :xh7 i..xe4) 56 :g2

defence. And he was right! 57 :xh7 i..xh7 58 :xh7 i..e5+ 59

49 :h3 @b3 :b2+ 60 @a3 :d2 61 i.. b l

49 : f2 ! ? has to be considered, i..d4 and White still faces scrious

since 49 . . . i..e8 50 :b 1 i..d4 5 1 :h2 trouble.

i..xa4 52 :bhl loses the é'.sh7. 54 i..gl

49 i..d4 55 :d2 [ 4 0 . 1 0 ]

49 . . . :g2+ is to no avail after 50 After 55 �hl ! ? i..d4 56 :afl

@e3 (or 50 @ d l :a2 5 1 i..c2 with a :g2+ 57 @c 1 Black might finally

slight black advantage according to play 57 . . . h 5 . White isri't well­

Karpov) 50 . . . :a2 51 :bl i..h2 52 prepared for this advance, Black has

:xh2. a distinct advantage.

50 :3h2 _:g3

51 :n
White has scored a moral success

-the position is the same as four

moves before. Therefore Karpov

tries something different.

51 i..f6

52 :bl

52 : fh l ! ? i..g5+ 53 @c2 @f6 54

:xh7 i..xh7 5 5 :xh7 is less effec­

tive on account of 55 . . . :g2+ 56

@b3 (56 @c3 i..d2+ 57 @d4? :g3)

56 . . . @e5 and the black king enters

d4. In spite of the material equality

Diagram 4 0 . 10
White faces problems.
Black to move
52 i..g5+

53 @c2 i..e3

54 :al 55 �f6?!

White plans a4-a5. He can play Centralizing the king is natural,

differently: but now White can equalize the

I) 54 :n i.. g l 55 :hl (55 :d2 game. Instead:

i..e8 56 @b3 i..c6 might cause I) 55 .. . h5 56 a5 bxa5 57 e5

headaches) 55 .. . :g2+ 56 @c3 i..h2 dxe5 5 8 i..xg6 @xg6 59 :xa5 with

57 :o i..e5+ 58 @b3. As now the counterplay thanks to the passed

white king is misplaced, it might be c-pawn.

appropriate to play 5 8 .. . h 5 ! ? and II) 55 . . . i..d4 ! ? (this looks better to

. . . @h6-g5. me than 5 5 .. . <iltf6) 56 :a2 (56 :n

II) 54 :bhl (attacking the é'.sh7) i..e8 57 @b3 i..c6 5 8 :el @f6 and

54 .. . @f6 55 @c3 (with the double . .. @e5, 56 :bl i..e8 57 @b3 i..c6

threat of 56 :h3 and 56 :xh7. 55 5 8 : e l @f6 or 56 : h l @f6 57 e5+

 Endgame Secrets 147

.i.xe5 58 Zib 1 1Ig4 all lead to a e5 62 1If6+ @g5 63 1Ia6 .i.d4 64

black advantage) 56 . . . @f6 and .:xa5 Zic3+ (64 . . . .i.c3 65 1Iaxe5+)

. . . @ e 5 . The push e4-e5 isn't feasible 65 @d2 1Ixc4 66 @d3 Zic3+ 67 @e4

here, because the bishop can take on @g4 68 :a6 g5 69 :.g6 and White

e 5 . After 57 a5 bxa5 5 8 1Ixa5 .i.c5 manages to draw.

we transpose to the game with the 57 .i.c5

exclusion of White's options 56 e5+ 58 Zial

and 57 e5+. Now it's too late to sacrifice: 58

56 a5?! e5+ @xe5 59 .txg6 hxg6 60 .:a 1

Now White could have got rid of @f5 61 lid3 lig4 followed by

both the �e4 and .i.d3 under favour­ . . . e6-e5. Black is winning.

able circumstances: 56 e5+ dxe5 58 @e5

(56 . . . @xe5 57 .: e l + .i.e3 5 8 .i.xg6 [ 4 0 .1 2 ]

hxg6 59 1Id3 @f4 60 :xd6) 57

.i.xg6 hxg6 58 a5 bxa5 59 .:xa5

[ 4 0 .1 1 ]

Diagram 4 0 . 1 2

White to move

Diagram 4 0 . 1 1 Black has scored an important

Black to move partial success, he has centralized

his king. But since the material is

The c-pawn intends to advance, reduced he still hasn't won. Now

that' s why Black should be content White should have asked himself:

with repeating moves after 59 . . . .i.d4 how can Black improve his posi­

6 0 1I d 3 .:[g2+ 6 1 1Id2 :g3. tion? Then White might have found

56 ... bxa5 out that Black intends to activate his

57 :xa5? .i.g6 via h5. Unfortunately I didn't

It ' s still possible to sacrifice the ask myself this question and the

e-pawn: 57 e5+ @xe5 (57 . .. dxe5 5 8 game rapidly goes downhill .

.i.xg6 hxg6 59 1Ixa5 leads to the 59 lifl .i.gl

lines after 56 e5+) 5 8 Ziel+ .i.e3 59 60 liddl .i.e3

.i.xg6 hxg6 60 :d3 @f4 6 1 :xd6 61 1If8?

148 Endgame Secrets

Abandoning the h-file. Better is Endgame 41

6 1 .:hl. J.Lautier-C.Lutz

61 .:g2+ Horgen 1 9 9 4

62 @b3 líh2 Earlier publications: Informant

Karpov seizes the opportunity. 61/527, Credit Suisse Masters

63 i,bl ? ! Horgen 19 9 4 (edited by H.Metz).

6 3 líh8 ! ? is better.

63 i,h5

64 líel i,f2

65 .:n i,c5

66 :el i,e2

67 i, a 2 ? !

And now 67 lt c l (with the idea of

67 . . . .:h3+ 68 líc3) is preferable.

Black should then advance his

passed pawn.

67

Now it's over.

68 @b2 i,a3+
Diagram 4 1 . 1
69 @al i,d3
White to move
70 i,bl i,b4

71 :el i,d2
Analyse the game.
72 .:dl i,c3+

73 @a2 i,xc4+
31 @f2 g6 32 e4 .:c7 33 .:d4
74 @a3 i,e2
i,a6 34 .:d6 i,c4 35 .:b6 .:d8 36
[ 4 0 . 13 ]
@el @g7 37 i,dl h5 38 h4 .:dd7
0-1
39 líb4 @f6 40 a4 @e5 4 1 a5 .:d8

42 .:b6 :as 43 b4 .:d8 44 i,a4

líd3 45 4Jb5 t/2-1/2

Commentary

After a highly interesting middle­

game an endgame with an uneven

material balance arose. Black is the

exchange up, but in retum White

has two passed pawns on the queen­

side. How to assess the present posi­

tion? In the joint analysis opinions

Diagram 4 0 . 1 3 differed. Joel Lautier thought he had

White to move a slight advantage. I myself consid­

ered White's position as clearly bet­

After 7 1/2 hours of playing I fi­ ter, the passed pawns appeared too

nally decided to end the suffering. menacing. Viktor Korchnoi

 Endgame Secrets 149

regarded White's positíon as win­ g7 and h7 wasn't clear to me during

ning. Garry Kasparov (the touma­ the game.

ment winner) assessed the mutual 31 g6?

chances as equal, and after 3 5 . . . �d8 A rather strange move which has

he even preferred Black's position. two disadvantages: Firstly, the

Thus we have four different views, square f6 is weakened. This might

who is right ? After a lengthy analy­ appear irrelevant, but we will en­

sis I think today that Kasparov's as­ counter lines where White plays

sessment is the most accurate. The e4-e5 and lt:Je4-f6. Possibly the

game should result in a draw, but white king might invade f6 via f4

Lautier misplayed it and should and e 5 . Secondly (and this •is more

have lost. important) Black doesn't care about

31 @ f2 centralizing his king. The king

With time getting short Lautier might go to e5 via g7 and f6, but

hurried to play this move. This is a then he will be driven back by

wise decision as obviously he has to f3-f4+. In the game Black's king

centralize the king. Now it was my really takes up a powerful position

tum to find a move. As I deemed on e5, but this is mainly due to

my position to be in danger I in­ White's optimistic play. 3 1... @f8 is

vested sorne time to ponder on gen­ the correct move, viz.: 32 e4 �c7

eral considerations. Unfortunately I (threatening . .. �xc3) 33 Jtb5 @e7

couldn't grasp all subtleties of the 34 a4 [ 4 1. 2 ].

position. At first I asked myself how

to place the pieces. Putting the

rooks on a8 and c7

that way the passed pawns are held

suggests itself,
lfa� ■ ! ■ !
back. The Jtb7 is blocked by the �,t'■"í■""'"■
pawns on g2 and f3 (and later e4),

but maybe it might seize the diago­ ,,•.!t.•�

nal a6-fl

. . . i..c6.
or it can be exchanged by

Another

couldn't answer correctly, however,

question that I .
��,�//,.

e
��
,. y
�-�-
t�t �� ��,�
t� ���
� :, ,, ,,, !,? � , ,, ,, ,, , � ,,. .. � � 1/ , ,, ,, ,

is: How should the kingside pawns

be arranged? For a correct answer

it's necessary to look at the white

Diagram 4 1 . 2
minor pieces. At the moment their
Black to move
mobility is restricted by the pawns

on e6 and f7, therefore Black Black now has the choice between

shouldn't move these pawns. After several ideas: For example he can

. . . e6-e5? the d5 square becomes ac­ initiate an exchange of bishops with

cessible .while after . . . f7-f6? or 34 Jtc6. Another choice is playing

. . . f7-f5? a white bishop on b3 might 34 g5, 3 5 . . . h5, 3 6 . . . g4 to soften up

pressurize against the weakened the a8-hl diagonal and open up

(je6. But where to put the pawns on sorne lines for the rooks. Black has
150 Endgame Secrets

sufficient counterplay then. It ' s dif­ 33 .s'. d 4 ? !

ficult to say whether Black should White parries the threat, but this

agree to an exchange of rooks ( ei­ move marks the beginning of a mis­

ther now by means of 34 .. . .s'.dS or taken plan which only helps to

after preparatory moves like 34 .. . g5 worsen White's position. Lautier

and the exchange of bishops with obviously regarded the �b7 as infe­

.. . �a6 or .. . �c6 respectively ) . Ex­ rior to the �a4 and in the further

changing rooks is a double-edged course of the game he avoids the ex­

affair: On the one hand the white change of the bishops. White has

king might support the passed several alterna ti v e s :

pawns, but on the other hand the re­ I) 33 �c2 �a6 34 e5 .s'.b8 with

maining rook might enter the enemy counterplay against b 2 .

camp. To demonstrate the mutual 11) 3 3 é2Jb5 l:c6 has no effect.

possibilities I give sorne lines: III) 3 3 . �b5 is the most sober

34 .. . g5 (This move is useful as it move. White accepts that Black

controls the f4 square; a white pawn might exchange bishops with

on e5 can become vulnerable then. .. . �a6. There might follow 3 3 . .. @f8

34 .. . .s'.dS 35 .s'.xd8 @xd8 36 @e3 (Black's king goes to e7, thus

looks good for White, his bishop 3 1... g 6 ? tums out to be quite �tupid)

might support the passed pawns 34 @e3 @e7 3 5 a4 g5. We reach the

from e2 while the �b7 now is really same position as in the note to

bad) 3 5 @e3 .i.c6 ( 3 5 .. . h 5 ! ? intend­ 3 1 . .. @fS, but with an extra tempo

ing . .. g5-g4 is an idea. Then White for White. This is of minor impor­

has to allow an opening of lines tance, the position should be equal

(which creates new options for the here as well.

rooks) or (if he plays f3-f4 instead) 33 ..ta-6

he has to tolerate two far advanced 34 .s'.d6

pawns at g4 and h5. These two 34 @e3 .s'. b 8 , and the white rook

pawns are a kind of insurance for has to retum.

Black: Should he ever capture the 34 � c 4 [ 4 1.3 ]

�g2 or the �h2 he is in possession

of a quick passed pawn) 36 .i.xc6

.s'.xc6 37 .s'. d 3 ! ? (37 e5? is risky as

long as both black rooks are on

board. After e.g. 3 7 . .. .s'.c5 38 @e4

.s'.b8 the �e5 is vulnerable)

3 7 .. . .s'. d S ! ? (37 h5!? is possible

here as well. 37 .s'.b8 3 8 é2Jb5 .s'.c2

39 é2Jd4 doesn't help Black) 38 b4

and the white pawns are fast. I think

that after 34 a4 the position is about

even.

32 e4 .s'.c7
Diagram 4 1 . 3
Threatening . .. .s'.xc3.
White to move
 Endgame Secrets 151

On c4 the bishop is much better Let's compare positions 4 1. 2 and

placed than on b7. Therefore it's 4 1. 4 . We see that in the latter one

clear that the last two moves haven 't the i.c4 controls important squares

improved White's posinon. at the queenside. Black might use

34 . . . ltb7? (to lure the pawns to the · this to mount an attack against the

front) is suicidal: 35 b4 lic8 36 i0ib2 by means of . . . :cb7. If the at­

lixa6 lhc3 3 7 i.c6 and 3 8 b 5 . tacked pawn advances, the �a3

35 li b 6 ? ! proves weak. The i.d 1 is useless,

White still intends to advance his White should trade it off with 38

pawns, but now he definitely over­ i. e 2 . But then this exchange takes

steps the mark. Black seizes the d­ place under less favourable circum­

file and is about to enter the white stances than in position 4 1. 2 . It

camp. Let's have a look at the nor­ tums out that the manoeuvre 33

mal continuation 35 @e3 lt b 8 : ltd4, 34 ltd6 is just wrong.

Black pressurizes b2 and controls 35 ltd8

the b5 square. As a consequence the 36 @el

i.a4 is susceptible to attack. Obviously something has gone

I) 36 b4? loses a pawn after wrong if White has to play such a

36 . . . i. fl , e . g . 3 7 i.c6 (3 7 lic6 ltxc6 kind of move.

3 8 i.xc6 lic8 39 b5 i.xb5, 3 7 �d2 36 @g7

i.xg2) 3 7 .. . ltb6 3 8 b5 i.xb5. With the white king down at the

II) 3 6 lt d 2 @f8 (36 . . . lt c b 7 ! ? is in­ first rank it makes sense to recon­

teresting, with the point that after 3 7 sider the idea of playing

b4 lta7 the unstable i.a4 gives . . . @g7-f6-e5-d4. With the friendly

White a headache. The white pieces help of the opponent 3 1.. .g6? gets

are glued together and i t' s not clear its justification.

how he can advance his pawns. For 37 i. d l [ 4 1. 5 ]

that reason 35 i.dl ! ? m i g h tbe more

accurate than 35 @e3) 37 i.dl ! ?

@e7 [ 4 1. 4 ]

.�-�-�.,.,
• ■,■,■
■ ■ • ■
■A■8■ ■
� m m8■
, - � � - -�:t ·-:�

�•--•��� ��'---" Diagram 4 1 . 5

Black to move

Diagram 4 1 .4
37 h5?
White to move
152 Endgame Secrets

Once again a mistaken pawn through f3-f4 and e4-e5 rather eas­

move whose only virtue is to bring ily (Black lacks the option . . . g6-g5).

Black closer to the time control. Secondly, White might possibly

I) The natural 3 7 . . . @f6? runs into play i.. d l - a 4 - e 8 (to attack the black

a tactical refutation: 3 8 b3 (now the pawns from behind). These two

i..c4 can 't move on account of the ideas greatly help White to keep his

knight fork on d5) 3 8 .. J�d3 3 9 ttJa4 endangered position, after 3 7 . . . g5 !

(39 ttJe2? ñcd7 40 i..c2 ñd2) he doesn 't have these ideas. But 3 8

39 ñe3+ (39 . . . ñcd7 40 i..e2 or h4 has a small drawback: the g3

39 i..a6 40 ñxa6 ñc 1 41 ttJb2 square is weakened. We will appre­

ñxb3 42 @d2 simply don't work) ciate this detail soon.

40 @f2 ñd3 4 1 i..e2 ñxb3 42 i..xc4 38 ñdd7

ñxb6? 43 ttJxb6 ñc6 44 ttJd7+ @e7 3 8 . . . @f6? 3 9 b 3 .

45 i..b5 and White is winning. 39 ñb4

This explains the nervous 37... 39 a4 ñb7 discloses the idea of

h5?, but Black has better options: B l a c k ' s last move. Or 3 9 i..e2 @f6

II) Possible is 3 7 . . . ñdd7 which 40 a 4 @ e 5 .

leads to the game without the inclu­ 39 @f6

sion of 3 7 . .. h5 3 8 h4. 40 a4

III) But I like 3 7 . . . g5 ! even better. Finally Lautier sets his pawns in

Setting up the pawns like this is motion.

strategically correct. Besides, 40 @e5

3 7 . . . g 5 ! sets a little trap: Now the black king has access to

A) 3 8 a4? (looks normal, but now the squares d4, d6 and f4.

the ttJc3 lacks the a4 square) 41 a5

38 @ f6 ! 39 b3 (else . .. @e5) 41 g3 is to B l a c k ' s advantage af­

39 ñd3 and now 40 i..e2 (40 ttJe2 ter 4 1... @ d 6 ! ? , e.g. 42 i..e2 (42

ñcd7 41 i..c2 ñd2 with a clear ttJb5+ @c5, 42 ñb6+ ñc6)

Black advantage) 40 . . . ñxc3 41 42 . . . i..xe2 43 @xe2 ñb7.

i..xc4 ñc2 is good for Black; the 41 ñd8 [ 4 1. 6 ]

white king gets into trouble.

B) 3 8 g3 ! ? is the more advisable

choice. Then 3 8 . . . @f6? 3 9 b3 is still

not possible, that' s why Black

should play either 3 8 . . . ñdd7 or even

38 . . . h5!? (to open up sorne lines

with . .. h5-h4). I think that Black is

better then.

38 h4

Of course White fixes the pawns

before Black can correct his mistake

by playing . . . g6-g5 or . . . h5-h4.

Black's pawn structure has two

Diagram 4 1 . 6
drawbacks: Firstly, White can force
White to move
 Endgame Secrets 15 3

4 1... @ d 6 42 :b6+ is to no avail.

But now .. . :as is a threat.

42 :b6?

Pretending that nothing has hap­

pened. White has a number of other

moves, most of them bad:

1) 42 i.e2 i.xe2 43 @xe2 l:¡c5 44

a6 ( or 44 :b7 :xa5 45 l!xf7 :b8)

44 . . Jh8,

11) 42 i.c2 :as 43 :.a4 l:¡b8 44

b4 .ta6,

III) 42 i.a4 @f4 (here the draw­

Diagram 4 1 . 7
back of 3 8 h4 shows, Black threat­
White to move
ens .. . @g3),

IV) 42 :a4 :b8,

V) 42 lt:Ja4 @f4 or White protects the �a6 by either

VI) 42 lt:Je2 l:¡a8 all lead to 50 i.e2 or 50 i.f3, 51 i.b7. After

Black's advantage. this neither side can win.

VII) 42 i.b3 ! ? (intending 42 �a8

42 .. . i.xb3 43 :.xb3 �c5 44 :a3) is 43 b4?

a viable option. This is probably the decisive mis­

VIII) The clearest way to equality take. It' s still possible to limit the

is 42 g3 ! . In the aforementioned damage by means of 43 b3 ! ? :

lines we saw that the black .king 43 . . . i.d3 (43 . . . i.a6 44 lt:Jb5) 44

lt:Jb5 (not 44 @d2? i.fl) and White

might enter the white position via

f4. The text prevents this and pre­ can flee from the attack:

pares to boot the king out with 1) 44 . .. �d7 45 a6 (45 b4? l:¡c8)

45 i.xb5 46 :xb5+ @d4

f3-f4+ and e4-e5+, v i z . : 42 :as 43

f4+ @f6 44 e5+ @e7 (44 @f5? 45 (46 @f4? 47 @f2) 47 :as @e3 48

@f2 l:Ixa5 46 @e3 and suddenly :a3 �d3 49 :al.

11) 44 .. . i.xb5 45 :xb5+ @d4 46

there's no defence against the mat­
@d2.
ing threat of 4 7 g4+ and 48 i.c2+)
III) 44 .. . �cl 45 @d2 l:¡al 46
45 :a4 :b8 46 b4, (46 a6? :as)
@xd3 (46 b4? i. fl ) 46 .. .l:hdl+ 47
46 .. . i.a6 47 b5 i.xb5 48 lt:Jxb5
@e3 (47 @c2? : g l , 47 @e2? l:¡ad8)
lhb5 49 a6 :.a7 [4 1.7 ]
47 .. . �xa5 48 f4+ ©f6 49 lt:Jd6 l:¡a7

50 e5+ @g7 51 :bs and White

draws because of the possible

lt:Je8-f6.

IV) 44 .. . � c 5 ! ? is another interest­

ing try. After 45 lt:Jd6 �cxa5 46

lt:Jxf7+ @f4 Black has the deadly

threat of .. . @e3, .. . : a l , . .. � 8 a 2 .

Therefore White ' s king has to seek

154 Endgame Secrets

salvation in 4 7 @d2. The resulting .i.e8 @f6) 46 .. . :a2 47 ét:Jc3 :xg2

position is unclear. and .. . @d4.

43 : d 8 [ 4 1. 8 ] 11) 45 ét:Jb 1 leads to a funny mate

that we will see back again, slightly

varied: 45 .. . l:te3+ 46 @f2 :e2+ 47

�g3 :b2 48 ét:Jc3 �fl 49 f4+ �d4

50 ét:Jb5+ @e3 51 ét:Jxc7 l:ixg2+ 52

�h3 �f3 53 ..idl+ l:ie2 mate!

III) 45 ét:Je2 l:ie3 46 �dl �xe2

A) 47 .i.xe2 l:ic2.

B) 47 @f2 :a3 48 @xe2 (48

.i.xe2 l:ic2) 48 .. . l:ia2+ 49 @ e l :d7

with a mating attack.

C) 47 @d2 @f4 (47 . . . � x d l ? 48

@xe3 poses sorne technical prob­

lems) 4 8 .i.xe2 l:id7+ 49 @ e l @g3

Diagram 4 1. 8
50 l:ic6 l:ib3 51 l:ic 1 l:ixb4, and the
White to move
white pawns drop off.

Now we can see clearly the dam­ The last move was accompanied

age that White has brought about. by a draw offer, Lautier finally

The :b6 is out of play and the ét:Jc3 sensed the dangers of his position. It

, lost its support. Black can make still makes me sad that I accepted

things hot for the white king with this offer, but I miscalculated. After

. . . :dd7 or .. . l:td3. But Black has to 45 ét:Jb5 :cd7 [ 4 1 . 9 ] Black wins:

be careful, as otherwise White' s

pawns might become dangerous af­

ter all.

44 �a4

Other moves are no improvement:

I) 44 �e2 .i d 3 ! 45 �xd3 (45

ét:Jd5 : e l+ 46 @d2 �xe2)

45 :xd3 46 :b5+ (46 ét:Je2 :c2)

46 �d4 or

11) 44 b5 @d4. Black is winning

in both lines

44 :d3

Not 44 . . . �b3? 45 f4+ @xf4 46

ét:Je2+. Diagram 4 1 . 9

45 ét:Jb5 White to move

Now White's king is devoid of

defenders, but moving the knight A) 46 ét:Ja3 (my eyes tumed glassy

elsewhere loses as well: when my opponent suggested this

I) 45 ét:Jdl l:ia3 46 �c6 (46 l:ic6 move) 46 .. . l:ie3+ 47 @f2 l:id2+! 48

:xc6 47 �xc6 l:ib3 48 @d2 �a6 49 @xe3 :e2# wins on the spot.
 Endgame Secrets 155

B) 46 a6 :e3+ 47 @f2 :d2+ i s n ' t :bb2 50 :es+ @f6 5 1 @f4 ( 5 1 a6

much better. e5 or 5 1 e5+ @g7 52 :c4 :xg2+ 53

C) 46 @f2 ( during the game I @f4 g5+ 54 hxg5 h4 lose)

considered this as the best defence) s1...:xg2 52 a6 (52 :c4 e5+ 53

46 . . . :d2+ 47 @g3 ii.fl 48 f4+ @e3 :b3+ 54 ctJc3 :a2) 52 . . . :xb4

@xe4! (The aforementioned miscal­ 53 a7 :a2 [ 4 1.1 1 ],

culation. I only saw 48 . . . @f6? 49

ctJd6 :xg2+ 50 @f3 when White is

better) 49 ctJc3+ @e3 with the bril­

liant point 50 i.xd7 (50 i.c6 :c7

and . . . :xc6) 50 . . . :xg2+ 51 @h3

@f2 (threatening a deadly discov­

ered check) 52 ctJdl+ @ g l 53 ctJe3

:h2+ 54 @g3 :h3# [ 4 1.1 0 ] w i t h a

beautiful mate:

Diagram 4 1 . 1 1

White to move

. . . and Black wins as. 54 :es e5+

55 @g3 :bb2 leads to mate.

Endgame42

M.Wahls-C.Lutz

German championship 1997

Diagram 4 1 . 1 0 Earlier publications: lnformant

White to m o v e 72/262, SCHACH 1/98.

Of course, it would have been

nice to finish the game in such a

style.

D) 46 :c6 (Relatively best. Black

has to be content with a technically

won positions) 46 . . . i.b3 (46 .. . i.xb5

47 ii.xb5 : d l + 48 @e2 :7d2+ 49

@e3 or 46 . .. :e3+ 47 @f2 :d2+ 48

@g3 i.fl 49 :es+ @f6 50 @f4 are

less convincing) 4 7 i.xb3 :xb3 48

@:f2 ( after other moves there is

.. . @f4 or .. . :xb4) 48 .. . :d2+

(48 .. . :xb4 49 :es+ @f6 50 a6 al­ Diagram 4 2 . 1

lows for sorne counterplay) 49 @g3 White to move

1 5 6 Endgame Secrets

Analyse the game. f3 'Ll g 5 1 9 l:r fd 1 'Llf7. The knight

covers the squares d6 and d8. If

1 7 ñd5 ñe8 18 f3 'Llf6 19 ñd6 Black could play .. . .i.e6 and . .. a5 he

.i.f5 20 c3 a5 2 1 ñ f d l @f8 22 'Lld2 would be without any worries, but

b5 23 'Llb3 .i.c2 24 ñ a l a4 25 'Llc5 White has the annoying move 20

ñbc8 26 'Lla6 a3 27 bxa3 ñxc3 28 ñ 5 d 2 !. After 20 .. . a5 2 1 .i. c 4 ! White

.i.xb5 ñec8 29 a4 e4 30 tl)b4 exf3 seizes the diagonal a2-g8. This

31 gxf3 ñxf3 32 ñcl .i.e4 33 bishop can 't be driven away as

ñxc8+ @e7 34 @gl @xd6 35 a5 2 1... b 5 ? ! 22 .i.d5 merely weakens

ña3 36 a6 'Lld5 37 ñc6+ @e5 38 the queenside.

ñc5 1/2-1/2 18 f3 'Llf6

In the course of the game the 'Llf6

Commentary contents itself with the protection of

the squares d5 and d7.

It took only 1 6 moves in a Sicilian 19 ñd6

to arrive at an endgame. In my Along the sixth rank the rook at­

home preparation I had already tacks the queenside pawns. To com­

looked at position 4 2 . 1 and consid­ plete his development Black has to

ered it to be equal. What are its move one ofthese pawns.

main features? White enjoys a 19 .i.f5

queenside majority, but this isn't Making use of the unprotected

important now. More important is l::,c2. 1 9 .. . b5? is clearly wrong on

White's control over the d-file and account of the weaknesses on c5

the weak pawns on a6 and b6. But and a6.

the white position has drawbacks as 20 c3

well: The 'Llb3 is stalled by the 20 .i.xa6 �xc2 doesn't promise

pawns on b6 and e5 and the @hl is anything.

misplaced. If a black rook should 20 a5

ever enter the first or second rank The weakening of the b5 square is

this is highly unpleasant for White. of small relevance only. Black now

Over the next moves White seeks intends 2 1 . . . ñed8.

to exploit the weakened queenside 21 ñfdl

while Black strives to complete his Seizing the d-file. Black doesn 't

development and to fight for the have to fear 2 1 .i.b5 ñed8 22 ñfdl

d-file. ñxd6 23 ñxd6 .i.e6 (23 .. . @f8 24

17 ñd5 'Lld2 followed by él'ic4 is unpleas­

lt ' s weaker to retreat the rook to ant) 24 ctJd2 (otherwise Black's

c6. After 1 7 .. . ñd8 18 f3 'Ll f6 19 king gets to e7) 24 .. . 'Ll e 8 ! ? as 25

ñd 1 ñxd 1 + 20 .i.xd 1 @f8 the black .i.xe8 ñxe8 26 ñxb6 ñd8 yields

king expels the ñc6. The position is strong counterplay.

equal then. 21 @f8

17 ñe8 Apart from centralizing the king

The natural way to protect e5. Black can consider two pawn

Wahls suggests 1 7 . .. f6!? instead: 18 moves:

 Endgame Secrets 157

I) 2 1... e 4 ? is bad as White doesn't 22 ét:Jd2

have to take on e4. After 22 ét:Jd4 The knight heads for c4, therefore

exf3 23 gxf3! Black's minor pieces Black is forced to act. 22 i.b5 Mec8

lack decent squares while the white 23 ét:Jd2 (or 23 i.a6 Me8) 23 . . . i.e6

knight takes up a dominating post. is not dangerous.

II) 2 1... a 4 is better. After 22 ét:Jd2 22 b5

b5 and 23 . . . i.e6 the white knight is 23 ét:Jb3!

placed passively, while after 22 Jumping back carne as an un­

ét:Jc 1 b5 23 ét:Jd3 e4 Black gets coun­ pleasant surprise. White has to en­

terplay in the centre. That's why 22 able the knight to become active on

i.b5 ! ? seems to be best. After the queenside. However, after 23

22 . . . axb3 23 i.xe8 [ 4 2 . 2 ] ét:Jfl ? ! i.e6 the knight stagnates.

23 i. c 2 ! ?

Black intends . . . i.xb3 followed

by . . . b4. Altematively he can push

his queenside pawns:

I) 23 . . . a4 24 ét:Jc5 is better for

White. Compare this position to that

after 2 1... a 4 ! ? 2 2 ét:Jd2 b 5 : In the lat­

ter one Black is out of trouble, be­

cause White' s knight is badly

placed. But in the former one White

has the initiative. He can centralize

his king while Black can 't improve

his pieces that easily. After 24 . . . e4

Diagram 4 2 . 2
White has 25 @ g l exf3 26 i.xf3.
Black to move
II) 23 . . . b 4 ! ? (42.3]

Black has to choose:

A) 23 . . . bxa2? 24 i.b5! @f8 25

:d8+ :xd8 26 :xd8+ @e7 27 :as

i.b 1 2 8 i.c4, and White wins.

B) 23 . . . ét:Jxe8 24 Md8 .tíxd8 25

:xd8 @f8 26 axb3 i.e6 27 :b8 (27

b4 b5 intending . . . i.c4) 27 . . . i.xb3

28 l1xb6 i.e6 29 b3 @e7 3 0 c4 or

C) 23 . . . :xe8 24 axb3 i.c2 25 .tlcl

i.xb3 26 .tlxb6 i.e6 27 c4.

In line C Black avoids the ex­

change of rooks while in line B

Black agrees to it. In both lines

White should be slightly better. Diagram 4 2 . 3

However, Black should draw if he White to move

manages to sacrifice his knight

against White's two passed pawns. . . . keeps the equilibrium:

158 Endgame Secrets

A) After 24 c4? .i.c2 Black suc­ Seeking for counterplay Black

ceeds in exchanging bishop for oversteps the mark. He has a choice

knight, followed by a blockade of of other moves:

the queenside: 25 :al!? (25 :1d2 I) 25 .. . e4 26 :xf6 gxf6 27 ét:Jd7+

.i.xb3 26 axb3 @e7 followed by @e7 28 ét:Jxb8 :xb8 29 : e l. In

. . . :ed8 and .. . �d7-c5. The .i.e2 is a view of Black's shattered kingside

sorry sight, only Black can be better White is virtually a pawn up.

here) 25 .. . .i.xb3 26 axb3 :ed8 (this II) 25 b4 26 :d2 loses the �a4.

is possible because of White's weak III) 25 :ed8 26 :xd8+ :xd8

back rank) 27 :xd8+ :xd8 28 :xa5 27 .i.xb5 :d5 28 ..txa4 isn't feasi­

:d2 29 :xe5 :xb2 30 ..t d l �d7 ble either.

and .. . : b l , . .. �c5. IV) 25 . .. :ecS (the natural and

B) 24 ét:Jxa5 bxc3 25 bxc3 :as best move) 26 ét:Ja6 :b7 (threaten­

26 .i. a 6 ! ? ( 2 6 �c6 :xa2; 26 :a6? ing .. . a4-a3) 27 :el .i.f5 28 ét:Jb4.

:xa6 27 .i.xa6 :as 28 :d6 ét:Je8 During the game I considered this

loses material after .. . ét:Jc7) 26 .. ..&te6 continuation as too passive. But

27 :ds+ :es 28 :sd6 :e6. If though Black is tied to the defence

White doesn't want to retum the of the �b5 he isn't really worse:

pawn he has to agree to a repetition 28 @e7 29 l:ddl (29 :d2 :d7)

ofmoves. 29 .i.e6 30 @gl .i.c4 3 1 @f2 :d7

24 :al!? and Black slowly equalizes.

Thwarting Black's intention. 24 26 4Ja6 a3

: 1 d2 .i.xb3 25 axb3 b4 26 i.c4 (26 Consistent.

c4? @e7 leads to the lines above) 27 bxa3 :xc3

26 .. . bxc3 27 bxc3 :ec8 2 8 @ g l (af­ 27 .. . .i.a4 28 :c1 :ed8 29 :xd8+

ter 28 :a6? or 28 :a2? Black can :xd8 30 4Jc7 is clearly advanta­

take on c4) 28 .. . a4 29 .i.a6 .s'.xc3 30 geous for White.

bxa4 e4. Black's counterplay com­ 28 .i.xb5 :ec8

pensates for the passed pawn on a4. [42.4]

We will encounter similar positions

later on.

24 a4

Forced. Other moves are worse:

I) 24 .. . .i.xb3 25 axb3 b4 26 .i.c4

is highly unpleasant, the pawns on

a5 and f7 are weak.

II) 24 .. . b4 25 ét:Jxa5 simply loses a

pawn as <loes

III) 24 .. . :edS 25 :xd8+ :xd8 26

ét:Jxa5.

25 ét:Jc5

25 ét:Ja5? :ec8 leaves the knight

misplaced. Diagram 42.4

25 :bc8? White to move

 Endgame Secrets 159

The counterplay along the e-file II) 30 11.c 1 ? i.d3 31 l':txc3 :xc3

and the offside white minor pieces 32 @ g l e3 33 i.xd3 l':t c l + 34 i.fl

seem to yield compensation for the e2.

pawn. But White's doubled passed III) 30 11. e l ? i.d3 3 1 fxe4 11. c l 32

pawn can become dangerous rather 11.gl (32 1id8+ @ e 7 ! ) 32 .. . 11.xgl + 3 3

quickly. The ltJf6 is useless in a @xgl 11.cl + 34 @f2 ltJxe4+ and

fight against the a-pawns. Black wins material.

29 a4? IV) 3 0 fxe4 ltJxe4 ( 3 0 .. . .txe4 31

Of course White doesn 't like to ltJ b 4 ! ? ) 31 1l.d4 i.xa4 32 i.xa4

part with his extra pawn, but now 11. cl + 33 11.xcl (33 11. d l ? ll:)f2+)

Black develops counterthreats. 29 3 3 .. . 11.xcl+ 34 i.dl ll:)f2+ 35 @gl

ltJb4 leads to a clear advantage: ltJxdl with counterplay.

29 .. Jixa3 (or else Black doesn't see 30 exf3

his pawn back again, e.g. 29 .. . i.f5 After 3 0 .. . e 3 ? 3 1 @g 1 White stops

30 a4 e4 31 11.adl exf3 32 gxf3 the e-pawn. 3 0 .. . i.d3? 31 ltJxd3

1ixf3? 33 1l.d8+ 1l.xd8 34 1l.xd8+ exd3 32 :xd3, and White is simply

@e7 35 ltJc6+ @e6 36 i.c4+. two pawns up.

29 .. . e4 3 0 lic6 isn't better) 3 0 11. c l 31 gxf3 1l.xf3

1l. a c3 3 1 @gl ! ? (direct tries to ex­ [42.5]

ploit the pin along the e-file only

lead to a small advantage: 31 ltJc6

ltJe8 32 1l.d2 1l.3xc6 3 3

34 1idc2 i.xc2 35 1ixc2,

i.xc6 .ta4

31 i.c6
.m¡m.,.r
■-■ Z,""1/,-/,"//,/,"""'�/,

: c 4 ! ? 3 2 a3 i.b3 3 3 l':txc4 i.xc4 or

■ �� - ■
,,-�-;-�
3 1 1ic6 1l.3xc6 3 2 ltJxc6 i.f5 3 3 a4)

•. ,.�.,.
3 l...i.f5 3 2 1ixc3 1ixc3 33 a4 @e7

34 lib6. The a-pawn is quite menac­

ing and Black's minor pieces can't

[j ■ A ■
� m �m ■
help.

29 e4 a n
�m��
Now Black is OK. He has three

threats: 30 exf3, 3 0 .. . e3 and

Diagram 4 2 . 5
3 0 . .. i.d3. 29 i.d3? doesn't work
White to move
yet on account of 3 O 1l.xd3 l:k 1 + 3 1

11. d l. Black restores material equality.

30 ltJb4 He still has threats against White ' s

White's knight has to get into play king while White has the passed

to defend against Black's initiative. pawns to the good. It ' s an important

Sorne lines help to show the danger­ defensive idea for Black to give his

ousness of the black pieces: knight against White's a-pawns. An

I) 30 li fl ? i.d3 31 i.xd3 1ixd3 endgame wli + i. + �h2 v bli + �fl

32 1l.xd3 ( or 3 2 1l.b6 e3) 32 .. . exd3 + �g7 + �h7 is completely drawn.

gives Black a big passed pawn. 32 11.cl

160 Endgame Secrets

Or: I) 32 i.c6 allows too much On account of the discovered

play: 32 . . . i.e4 33 i.xe4 (33 l:[xf6 mate . . . :fl Black allows his rook to

xn mate, 33 @gl i.xc6 34 l:[xc6 be taken with check. I had to foresee

l:[a8) 3 3 .. . tbxe4 34 J:ld4 f5 and this tactical tum as otherwise White

Black has the initiative. is better:

11) 32 a5? is even worse: 32 . . . i.e4 I) 32 . . . <it e 7 ? ! (to follow up with

3 3 @ g l J:lc5 and . . . :gs+. . . . i.e4, but...) 33 l:[d4! l:ífc3

III) 32 tbxc2 ! ? is possible. The ( 3 3 . . . i.f5 34 l:[xc8 i.xc8 35 tbc6+

black bishop is a strong piece, along @e6 3 6 i.c4+ @f5 3 7 l:íd8 wins for

the diagonal h l - a 8 it can both attack White, as <loes 3 3 . . . :f2 34 @gl or

the enemy king and stop the a­ 3 3 . . . i.e4 34 �xe4+ check!) 34

pawn. But after its exchange Black :[ e l + @f8 3 5 a5 and the i.c2 is out

quickly occupies the second rank: of play.

32 . . . .U.xc2 11) 32 . . JHc3?! (rather passive) 33

A) 33 a5? (too optimistic) :xc2 :xc2 34 tbxc2 �xc2 35 a5

3 3 . . . :m 34 a6 (34 i.c6 @e7) l:[ x a 2 3 6 a6 @e7 37 l:[d4 J:la5 38

34 . . Jíxh2+ 3 5 @gl l:[hg2+ 3 6 @hl l:[a4 .U.xa4 3 9 i.xa4 tbd5 40 a7 tbc7

tbh5 and White is in deep trouble. 41 i.c6. After exchanging the

B) 33 lí e l (threatening l:[d8+) knight against the a-pawn the posi­

3 3 . . . g5 (all black pieces have to par­ tion is drawn. However, 33 @gl

ticipate, 3 3 . . . g 6 ? ! is worse) 34 J:le2 i.e4 34 !íxc3 �xc3 35 a 5 l:[a3 36

líc5 (Both 34 . . . :cl + 3 5 @g2 g4 3 6 a6 ! ? is a try to go for more.

l:íd3 followed by l:[f2 or 34 . . . :xe2 III) 32 . . . l:[f2?! 33 @gl l:[g2+ 34

35 i.xe2 líf2 36 i.c4 tbg4 37 a5 @xg2 i.e4+ 3 5 @f2 l:[xcl 3 6 a5 is

l:[xh2+ 3 8 @ g l líc2 39 a6 are dan­ better for White.

gerous for Black) 3 5 l:[c6 :xc6 36 33 l:[xc8+ @e7

i.xc6 l:[a3 with an equal game as 34 @gl

Black stops the a-pawns, 34 J:ld7+?! tbxd7 doesn't offer

32 i.e4! any advantages.

[42.6] 34 @xd6

35 as J:la3

36 a6 tbdS

Avoiding the draw with 36 . . . tbg4

(intending . . . tbe5) isn 't better: 37

i.c6 (3 7 tbc6 :xa2 and . . . J:la 1 +)

37 . . . i.xc6 3 8 líxc6+ and l:[b6-b7.

37 l:[c6+

37 tbxd5? i.xd5 and the l!ia2 gets

lost.

37 @es

3 7 . . . @e7?! is very optimistic, e.g.

38 líc 4 i.f3 ( 3 8 . . . tbc3 39 :xe4+

tbxe4 40 a7 J:lxa7 41 tbc6+) 39

Diagram 4 2 . 6
tbc6+ <itd6 40 tbd4 i.e4 4 1 i.c6 ! ?
White to move
 Endgame Secrets 161

:xa6 42 4Jb5+ �e5 43 lhe4+ Commentary

<it>xe4 44 i.xd5+.

38 l:[cS [42.7] White has the more active queen,

but after 23 . . . d5 24 exd5 exd5 Black

shouldn't have any problems. lt ' s

insignificant that Black has an iso­

lated pawn since he has the right

bishop. Instead I went wrong

completely.

23 b5?

With sorne more pieces on the

board ( e.g. with a rook on c8) this

move would be · appropriate. But

endgames have their own rules.

24 axb5 axb5

25 c4! bxc4
Diagram 4 2 . 7
26 ¡uxc4
Black to move
Having a distant passed pawn is

With a draw offer that I accepted an advantage in almost any type of

because of 3 8 . . . <it>d6 39 l:¡c6+. endgame. Black is close to defeat.

3 8 . . . @d4? 39 l:¡xd5+ and étJc2+ 26 ... d5

isn 't feasible. 27 ¡uc7 'iVbs

After 27 . . . 'vJi/xc7 28 i.xc7 fol-

Endgame 43
lowed by @g3-f4 it ' s questionable
I.Glek-C.Lutz
whether Black can hold the game.
Copenhagen open toumament 1 9 9 5
28 'vJilb8 ¡uc5
Earlier publication:
28 . . . 'vJi/c6 29 b4, and White's
ChessBaseExpress 4 7 (I.Glek).
pawn is very fast.

29 ¡ues 'iVb6

30 e5!

Very strong. White fixes the

weakness on e6 and closes the

diagonal.

I) 3 0 b4 d4 3 1 b5 d3 is only equal.

II) 30 exd5 exd5 3 1 �e5 d4 does-

n 't promise anything either.

30 ... d4

31 i.f2
31 i.h4 d3 32 i.e7 'iVf2! and

Black draws (33 i.xf8 �f4+).

Diagram 4 3 . 1
31 h6
Black to move
32 h4?

Black to move. Assess 23 . . . bS and Giving away the advantage. 32

23 . . . dS. i. g l ! [43.2] is suggested by Glek.

162 Endgame Secrets

Endgame 44

C.Lutz-1.Rogers

German league 1 9 9 5

Earlier publication:

Informant 65/122.

Diagram 43 .2

Black to move

N ow the bishop is defended.

Black is in difficulties: 32 . . . 'iVxb2

(32 . . . h5 3 3 @hl h4 34 b4 'iVxb4 3 5

'iVxe6+ @h8 36 �g4 .ic5 37 Diagram 4 4 . 1

�xh4+ is clearly better for White, White to move

too) 33 'iVxe6+ @h8 ( 3 3 . . . @h7 34

�d5) 34 'iVc8 'iVb4 35 'iVd8 and a Analyse the game.

'pawn gets lost.

45 �hl+ @g7 46 él:'ld7 f4 47 él:'lf6
32 ... �xb2
él:'lxe5 48 él:'le8+ @g8 49 dxe5 'iVxeS
33 'iVxe6+ @h8
50 �h6 f3+ 51 @h2 d4 52 b5 d3 53
34 'iVf7 .ia 3 !
b6 d2 54 b7 dl ='iV 55 b8='iV 'iVhS+
N ow Black has enough
56 �xh5 gxh5 57 él:'lf6+ @g7 58
counterplay.
él:'lxh5+ @g6 59 'iV e5 1-0
35 h5 d3

36 'iVe8+ @h7
Commentary
37 'iVg6+ @h8

38 'iVe8+ @h7
Queen endings have a • special
39 'iVg6+ @h8
character: Even with a bare queen it
½-½
might be possible to give perpetua!

check. If you add knights and pawns

a tactical battle is pre-programmed

where both kings should strive for

maximum protection. In the present

position White has a clear advan­

tage: On the one hand he has a

passed pawn, on the other hand he

can create threats against the enemy

king.
 Endgame Secrets 163

45 'iYhl+ @g7

46 lt:Jd7!

Playing on both wings: Apart

from b4-b5-b6 there is the threat of

lbf6.

46 f4 !

Black seeks counterplay by at­

tacking White's kingside. Other

continuations lose:

I) 46 .. . 'iYh6 4 7 'iYxh6+ @xh6 48

b 5 . White plays 49 b6 and advances

his king to the queenside.

Diagram 4 4 . 2
11) 46 . .. 'iYd2 47 8f6 lt:Je3+
B lack to move
(47 .. . 'iYh6 4 8 lbe8+ @h7 49 'iYxh6+

@xh6 50 lbd6 loses the l!:,f7) 48 @f3 . .. wins. The black knight is in

'iYd 1 + 49 'iYxd 1 lbxd 1 50 b5 and the trouble.

b-pawn queens. Cl) 4 7 .. . 'iYe 7 48 lt:Jf6 'i!Vxb4 49

III) 46 .. . lba3 is a tougher defence: 'iYhl ! (the queen retums to the scene

A) Exchanging queens by 47 f4 of the crime) 49 . .. �5 50 'iYh7+

'iYh6 48 'iYxh6+ @xh6 doesn 't help @f8 51 'iYg8+ @e7 52 'iYc8 and

now. The b-pawn is blocked, e . g . 4 9 lbg8 mate.

@f2 @g7 (not 49 .. . lt:Jc2? 50 b5) 50 C2) 47 . . . lbc2 48 'iYc3 'iY c l

lbb8 (50 @e2? lbc2 51 b5 lt:Jxd4+) (48 'iYe7 49 lbc5) 49 'iYc5 @h7

50 .. . lb b 5 5 1 @e3 @f8 and the black (49 lb e l + 50 @fl lt:Jd3+ 5 1 'iY x c l

king approaches. I don 't see how lbxcl 52 b5 lbb3 5 3 b6 lt:Jxd4 54 b7

White can win. lbc6 55 f4 followed by b8='iY is

B) 47 8f6 'iYh6 48 lt:Je8+ (48 hopeless) 50 'iVf8 lbe 1 + 51 @h3

'iYal lbc2 49 'iYc3 'iY c l gives Black 'iYh6+ 52 'iYxh6+ @xh6 53 b5 g5 54

enough counterplay to secure the f4! (54 b6 g4+ 55 @h2 lt:Jf3+ 56

draw) 48 . .. @h7 49 'iYal (49 'iYxh6+ @g2 lbxd4 is less convincing as the

@xh6 50 lbd6 @g7 5 1 b5 lbc2 loses black pawns aren 't blocked. Black

the l!:,d4 and i t ' s not easy to see how has counterplay by means of .. . @g5

the b-pawn advances) 49 . . . lbc2 50 and . . . f5-f4) 54 .. . g4+ 5 5 @h2 lt:Jf3+

lt:Jf6+ (50 'iYc3 'iY c l , 5 0 'iYa7 lb e l + ) 56 @g2 lbxd4 57 b6 lbc6 5 8 b7 and

50 .. . @g7 51 'iYa8 'iYh8 52 'iYxh8+ this time the extra knight proves

@xh8 53 b5 lbxd4 54 b6 lt:Jc6 55 decisive.

lt:Jd7 (55 b7 lbb8) 5 5 .. .tba5 5 6 @f3 C3) 47 .. . lt:Jb 5 48 'iYa5 étJxd4

(56 8c5 lbc6) 56 . .. @g7. If White's (48 .. . lbc3 49 b5) 49 'i!Vc5 'i!Vd8 50

king goes to the queenside, Black's lt:Jf6 and Black loses the knight, e . g .

king goes to f3 . I don't see a clear 50 .. . lbe2 51 'iYe 3 or 50 . . . l2Jb3 51

win. 'iYc3.

C) 4 7 'iYa 1 ! [ 44. 2] C4) 47 . . . lt:Jc4 48 'Df6 'iYh6 (after

48 . . . f4 49 'iYhl we transpose to the

164 Endgame Secrets

game) 49 b5 ( 49 'i'a8 'i'h8 50 ... leads to a lost knight ending.

'i'xh8+ @xh8 5 1 b5 lt:Jb6 isn 't com­ The white king can invade the king­

pletely clear, but now 50 'i'a8 is a side, e . g . 5 I...lt:Jb6 52 lt:Jd6 @g7 53

threat) 49 . . . 'i'h8 50 'i'a7 (Black's @g4 @f8 54 @g5 @g7 55 f4 @f8

attack on the kingside doesn 't coun­ 56 @h6 @g8 (56 . . . lt:Ja4 57 lt:Jc8

terbalance the advance of the b­ @e8 5 8 b6) 57 lt:Je8 lt:Ja4 5 8 lt:Jf6+

pawn) 50 . . . g 5 ! ? (50 . . . f4 51 gxf4 @f8 59 @h7 lt:Jb6 60 @h8 followed

'i'h4 52 'i'b8 'i'xf4 53 'i'g8+ @h6 by lt:Jh7+, @g7 and lt:Jg5xf7.

54 'i'xfl or 50 . . . 'i' d 8 5 1 'i'd7 are 48 toes+

winning for White) 5 1 'i'e7 and it's Apart from this check 4 8 dxe5!?

White who attacks on the kingside. de serves consideration: 48 . . . 'i'xe5

The lines above show that in case 49 lt:Jg4 (but not 49 lt:J e 8 + ? @f8)

of a queen exchange it's highly im­ and now Black shouldn't play

portant how far the b-pawn can 49 . . . 'i'e4+ 50 f3 'i'e2+ 5 1 lt:Jf2 'i'b2

advance. 52 'i' e l fxg3 53 @xg3 as then the

47 lt:Jf6 lt:Jxe5! b-pawn offers winning chances to

Consistent. In retum for the White. Instead 49 .. . f3+ 50 @g 1

knight Black gets two pawns and 'i' e l + 5 1 @h2 'i'xb4 52 'i'xf3 'i'd4

the lt:Jf6 loses its footing. leads to a position where White's

I) 4 7 . . . lt:Je3+ is refuted by 48 fxe3 material advantage shouldn't be

'i'xg3+ 49 @ fl . enough to win.

II) 47 . . . f3+ 48 @xf3 lt:Jxe5+ 49 48 @g8

dxe5 is insufficient: 49 . . . 'i'f5+ Forced. 48 . . . @f8 49 'i'h8+ @e7

(49 . . . 'i'xe5 50 'i'h7+ @xf6 51 50 dxe5 fxg3 51 'i'f6+ 'i'xf6 52

'i'h8+ @f5 52 g4+) 50 @g2 'i'xe5 lt:Jxf6 gxf2 5 3 @xf2 is hopeless.

5 1 lt:Jg4 'i'e4+ 52 f3 and Black gets 49 dxe5 'i'xe5

only one meagre pawn in retum. There is another possibility which

III) 47 . . . 'i'h6 48 lt:Je8+ @h7 49 just about fails: 49 .. . f3 + ! ? 50 @xf3

'i'xh6+ @xh6 50 b5 fxg3 51 @xg3 ( otherwise it' s a perpetual on c 1 and

[44.3] h6) 50 .. . 'i'xe5 (with the double

threat of . . . 'vJVe4# and . . . @f8) 51

'i'b 1 ! (Supporting the passed pawn

and protecting e4 at the same time.

51 'i'h4 g5 52 'i'g4 @f8 doesn't

help White) 5 1... @ f8 52 b5 @xe8

53 b6 @d7 54 b7 'i'b8 (forced as

the pawn ending after 54 . . . 'i'e4+ 55

'i'xe4 dxe4+ 56 @xe4 @c7 57 @e5

is just lost) 55 'i'b6 [ 44.4].

Diagram 44.3

Black to move
 Endgame Secrets 165

comparison to 63 @b5 d3 ! )

63 . . . 'iYc7+ (63 . . . e5 64 @c5 'iYf8+ 65

@b5 �8 66 'iYc6+ @d8 67 'iYxf6+

and White wins) 64 @b5 'iYe5+

(64 . . . �8 65 'iYc6+ and @b6) 65

@a6 'iYe2+ 66 @a7 with a win.

11) 60 . . . f5 6 1 @e2 (waiting is bet­

ter than 61 @d4 'iYh8+) 6 1... 'iY e 5 +

(61...@e7 62 @e3 @d7 63 @d3 and

a black pawn has to advance) 62

@d2 �8 63 @d3 e5 (63 . . . @e7 64

@c3) 64 @d2 f4 (64 . . . d4 65 @d3

Diagram 44 .4
and @c4) 65 @e2 fxg3 (65 e4 66
Black to move
gxf4, 65 . . . @e7 66 'iYc6, 65 S+ 66

@d3 e4+ 67 @e3) 66 fxg3 e4 67

This position is quite remarkable.
@d2 e3+ 68 @ d 3 ! (White's king has
In spite of his minus pawn White is
to handle the black pawns like a
winning. The reason is the mighty
juggler. For the moment the e-pawn
¿jb7 which ties down the enemy
is taboo on account of . . . 'iYxg3 with
queen. But on the other hand the
check) 68 . . . e2 (68 . . . @e7 69 'iYc5+)
white queen can 't move either with­
69 @xe2 'iYe8+ 70 @dl 'iYh8 71
out allowing . . . @c7. White's win­
@d2 [44.5].
ning plan is to bring his king to d3

and then to operate with zugzwang:

55 . . . g5 (Black won't get around this

move) 56 @e2 (56 @g4 f6 57 @h5

isn't worthwhile: 57 . . . 'iYh8+ 5 8 @g6

'iYg8+ 59 @xf6 'iYf8+ 60 @xg5

'iYf5+ 6 1 @h4 'iYh7+ is a perpetual)

56 . . . 'iYe5+ 57 @d2 �8 5 8 @d3 g4

59 @d2 (A waiting move, but White

can't proceed in a more straightfor­

ward way: 59 @d4 'iYh8+ 60 @c5?

'iYc3+ 6 1 @b5 'iYc4+ 62 @a5 'iYa2+

with a perpetual, After 59 @c3?

'iYc7+ White even loses. White's

Diagram 44. 5
king can go to c3 only if Black's
Black to move
king lea ves d7) 59 . . . f6 6 0 @d3

I) 60 . . . @e7 6 1 @c3 (now the king Though the number of pawns is

can cross the e-file. 61 @d4 'iYe5+ reduced the black problems haven't

or 6 1 'iYc6 @d8 and . . :'iic7 are less diminished. White may transpose to

convincing) 61 ...@d7 62 @b4 d4 a pawn ending, e.g. 7 1.. . @ e 7

(otherwise White plays @b5, 'iYc6+, (71...d4 72 �5+ @e7 73 'iYg5+) 72

@b6) 63 @c4 ! (gaining a tempo in 'iYc6 @d8 (72 . . . 'iYxg3 73 'iYc5+ @f7
1 6 6 Endgame Secrets

74 �xd5+ @e7 75 ¡vc5+ @fl 76 57 b8=¡v d l = ¡v _ White can't win

¡vc4+ and depending on Black's anymore.

next move White checks either on VI) 5 1 ¡v g 7 + ! ? is the critical con­

the third rank or the b-file) 73 tinuation: 51...¡vxg7 52 ébxg7 @f8

�f6+ @d7 74 ¡vf5+ @c7 75 ¡ve5+ [44.6]

@xb7 76 ¡vxd5+ with an exchange

of queens.

50 ¡vh6

The only move to pose problems.

Black's king can't go to f8, his

queen has to defend against the

mate. Possibly White might ex­

change queens on g7. After 50 � 1

fxg3 5 1 b5 gxf2 Black has at least a

perpetual.

50 f3+?

The decisive error, now Black is

lost. After 5 0 .. . fxg3 it's question­

Diagram 4 4 . 6 ------
able whether White can win as his
White to move
king lacks protection:

I) 5 1 fxg3 ¡ve2+ 52 @h3 ¡vfl + or In this funny position White has

II) 51 f3 ¡ve2+ 52 @xg3 ¡v e l+ only two pawns, but one of them is

lead to perpetual check. a passed pawn. White's knight is in

III) 51 f4 ¡ve2+ 52 @xg3 ¡ve3+ a tragi-comic position. It can 't be

53 @h4 ¡vf2+ 54 @g5 �g2+ 55 taken (52 .. . ©xg7? 53 b5), but it

@f6 �2+ 5 6 @e7 ¡vxb4+ 57 ébd6 can't free itself either: The squares

�8 certainly doesn't help White. h5 , f5 and e6 are controlled by

IV) 51 ¡v e3 �2 and White has bla ck p awns while the king guards

too few pawns left. e 8 . To make the latter square acces­

V) 5 1 b5 (now Black has to watch sible White has to advance h is

out) 5 1.. . d 4 (the only move on ac­ passed paw n to b6. Ho wever, this

count of 5 1... g x f2 52 ¡vg7+! and p awn might then fall victim to

5 1... � e 4 + 52 @xg3 �e5+ 53 ¡vf4 Bl ac k 's king. White's king can't in­

¡vc3+ 54 @g2) 52 b6 (52 f4 ¡ve2+ v ade that easi ly as he has to watch

53 @xg3 ¡vel + 54 @g4 �gl+ 55 the d-pawn. W hite should probably

@h4 � h l + ! 56 @g5 ¡vd5+ with a s eek to exchange the b-pawn against

draw; 52 ¡vg7+? ¡vxg7 53 ébxg7 d3 one the kingside pa wns. But then

and White loses) 52 .. . d3 (after the fi nal outcome depends on the

52 .. . �e4+ 53 @xg3 �e5+ 54 ¡vf4 p lacement of the remaining black

¡vxf4+ 55 @xf4 White can stop the p a wn s. P osit io n 44.6 is amazingly

d-pawn) 53 b7 ¡ve4+ (only now!) co mp lex; I c a n ' t decide whether it's

54 @xg3 ¡ve5+ 55 �f4 (55 @g2 a draw or a w in fo r White. White

�e4+ 56 f3 ¡ve2+ 57 @g3 ¡v e l + 5 8 now has to chose be tween 53 fxg3

@g2 ¡ve2+) 5 5 .. . ¡vxf4+ 56 @xf4 d2 and 53 �xg3. 53 b5 @ e7 has no

 Endgame Secrets 167

independent significance as White 4J c5 l o ses. After ex changing the

has to take on g3 anyway. �b6 and �e5 the b lack pawns h a ve

A) 53 fxg3 (not very natural, but lost their mobili ty)

White might tie down Black's

pawns by advancing the g-pawn to ·

g4 or g5) 53 . . . @e7 54 @f3 (54 b5

@d6 55 ctJe8+ @c5 56 @f2 d4 57

@e2 f5 followed by . . . e6-e5 should

be drawn) 54 . .. @d7 (54 . . . e5 5 5 g4 is

troublesome. After . . . f7-f5 White

might sacrifice the knight on f5,

otherwise White plays g4-g5, e.g.

55 . . . @d7 56 @e3 @e7 57 b5 @d7

58 g5 and b5-b6, ctJe8-d6xf7) 55

@e3 f5 56 @d4 (if Black can ad­

vanee also the e-pawn to the fifth

rank he shouldn't lose) 56 . . . @d6 57 Di agram 4 4 . 7

4Je8+ @e7 (57 . . . @d7 5 8 t'2)f6+ @d6 White to move

59 ctJ h 7 ! ? ) 58 tZ)c7 @d6 59 ctJb5+

@c6 60 tZ)c3 @d6 and . . . e6-e5+. White can trade o ff his �b6

Though White' s knight has escaped, against two pawns : 57 b6 @ c6 58

the position isn't clear to me. 'D e8 @xb6 ( 5 8 . . . f5 5 9 lbf6 is bad as

B) 53 @xg3 (the more natural Bl ac k loses his e-pawn) 59 ctJxf6

move as the king is immediately @c 5 60 ctJd7+ @d6 61 lbf8 g5 62

centralized) 53 . . . @e7 (Black can't ctJh7 g4 63 ctJf6 @ e6 64 tZ)xg4. But

take control over f4 as 5 3 . . . e5 54 b5 now B l a c k ' s k i ng is t oo active after

@e7 55 b6 @d7 56 tZ)e 8 @ c6 57 64 . . . @f5, e.g. 65 'Dh2 ( 6 5 @f3 e4+

ctJf6 d4 58 @f3 @xb6 59 ctJd7+ 66 @g3 d4, 65 tZ)h6+ @g6 66 ctJ g8

loses without a fight) 54 @f4 ( 54 b5 @f7) 6 5 . . . d4+ 66 @f3 e4+ 67 @g3

@d7 55 b6 @ c6 56 tZ)e8 @xb6 57 e3 68 f3 e2 69 @f2 d3 and B lack

ctJd6 f5 leads nowhere) 54 . . . f6 wins the f-pawn. Thus White has to

(54 . . . @d6 55 'D e8 + @e7 56 ctJc7 try something different. I an a l y sed

@d6 57 'Db 5 +; the knight escaped 57 @d3 ! ? f5 for a w hi l e but I could­

and White sh o uld win. 54 . . . @d7 55 n 't come to any clear c o n cl usions.

@e5, and W h ite wins after both Maybe you can find a wi n ?

5 5 . . . @e7 56 b5 and 5 5 . . . @c6 56 Back to the gam e:

ctJ e 8 . His king is too far advanced) 51 @h2 [ 4 4 . 8 ]

55 @e3 e5 56 b5 @ d7 [ 4 4 . 7] O b v i o usl y White doesn 't take the

( 5 6 . . . f5 57 b6 @d6 5 8 lbe 6 @ c6 5 9 pawn.

168 Endgame Secrets

If 53 .. . jVd4 then not 54 @h3? d2,

but instead 54 b7! 'ifxf2+ 55 @h3

'iffl + 56 @g4 f5+ 57 @g5 'if c l+ 5 8

@xg6 ¡vxh6+ 59 @xh6 d2 60 b8='if

d l = 'if 6 1 étJd6#.

54 b7

Certainly not 54 'ifxd2? jVh8+

with a perpetual,

54 .. . dl=jV

55 b8='iV! [44.9]

Diagram 4 4 . 8

Black to move

51 d4

Black's pawn lacks one tempo.

I) 5 1... \li'b 2 . Black attacks two

pawns, but after the cool 52 @h3

neither can be taken. White' s king

can't be bothered: 52 .. . jVal 53 @g4

'ifd4+ (53 f5+ 54 @xf3 @f7 55

jVg7+, 53 'iie5 54 b5) 54 'iff4

'ifxf2 55 'iff6 @f8 56 CDd6 'ifa7 57

'ifh8+ @e7 58 'Dc8+, winning the Diagram 4 4 . 9

queen. Black to move

II) 51...'ifd4 52 �h3 is similar to

I. The "silent" promotion is much

III) 5 1... 'ii h 8 52 'ifxh8+ @xh8 53 stronger than the "noisy" 55 CDf6+?

b5, and the newly bom queen mates 'ifxf6 56 b8='iV+ 'iidd8.

on g7. 55 jVh5+

IV) 51...f5 52 b5 @f7 53 'iig7+ What e l s e ?

'iixg7 54 ltJxg7 d4 55 b6 d3 56 b7 56 'ifxh5 gxh5

d2 57 b8='if d l = 'if 5 8 'iic7+. Once 57 ct)f6+ �g7

again White comes first. 58 étJxh5+ @g6

52 b5 d3 59 'ife5 1-0

53 b6 d2 Black loses the �f3 next.

 Endgame Secrets 169

Endgame 45 Bastian faced a difficult choice on

H.Bastian-F .Zeller move 74, for example: 74 @b6?

German championship 1 9 9 6 Wc6+ 75 @a7 Wxc5+ 7 8 @b8 Wf4+

Earlier publication: SCHACH 1 / 9 7 . 77 @a8 Wc6+ 78 �7+ Wxb7+

(winning a rook), 74 l!b8? Wfb3+

75 @a5 Wd2+ or 74 Wh8+? @a2 75

Wh2+ @a3, with a win for Black in

either case. But there is a way to

salvation:

74 We4!

White loses a pawn, but in retum

he exchanges queens.

74 ... Wtb3+

� Wb4 Wd�+

76 @c6 [ 4 5 . 2 ]

Diagram 45 . 1

White to move

And now, finally, something dif­

ferent. This bizarre position arrived

in an actual game of the German

championship. How should White

play?

Commentary

At first you probably looked for a Diagram 4 5 . 2

winning continuation. After ali, Black to move

White has a nominal advantage. But

in fact White has to be happy that 76 ... Wd5+

there is a small path to the draw. In 76 . . . 'iVbxb4 77 .ixc4 Wxc4 or

the game White got confused: 74 76 . . . Wcxb4 77 ét:Jxb3 Wa4+ 78 .ib5

�e4? Wfd3+ 75 @b6 (or 75 @b4 Wxb3 don't achieve anything either.

Wa3+) 75 . . . Wd6+ 76 @b7 (or 76 77 @b5

@a7 Wxc5 77 @b7 Wdd5+ 78 @b8 Black can't make any progress.

Wd8+) 76 . . . W3d5+ 77 @a7 W5xc5+

78 @a8 Wd8+ 0 - 1.
 Index of Players

The numbers refer to games. Bold numbers signify that the player was

White. Numbers in brackets refer to endgames within the commentary.

Adams 3 Kamsky 33

Alexander 26 Karpov 1 2 , ( 1 5 ) , (34), 40

Almasi 13 Kasparov 27,31

Andersson 22,39 Kavalek (15)

Bastian 45 Keitlinghaus 32

Beliavsky (1 5 ) Kramnik 27,34

Bertok 4 Lauber ( 3 1 ) , (39)

Bischoff 14 Lautier 8, 4 1

Bol bochan 29 Lasker (35)

Capablanca (35) Leko 12

Chemin (19) Lutz 1 , 3 , 5 , 6, 9, 1 1 ,

Christiansen ( 1 9 ) , 25 1 3 , 1 5 , 17, 18, 19,

Dautov 30 30, 3 2 , 34, 3 5 , 3 7 ,

Degraeve 2 3 8 , 3 9 , 40, 4 1 , 42,

Dvoirys 16 43,44

Ekstroem 9 Luther 19

Ellers (27) Maiwald 14

Finkel 21 Nimzowitsch (19)

Fischer 24,29, 36 Nunn 23

Forintos 24 Olafsson (11)

Gelfand 31 Papaioannou 37

Ginsburg 1 1 Piket 7, 8

Glek 43 Prusikhin (31)

D.Gurevich 20 Roselli (1 9 )

M.Gurevich (1 9 ) Rogers 44

Hansen 2 Rubinetti 36

Hickl 5 Sadler 23

Hnbner 35 Salov 28

Hort 4, (34) Shirov 7

Illescas Cordoba 33 Schlosser 15

Ivanchuk ( 11 ) Serper 10
172 Endgame Secrets

Smyslov 26 Tseitlin 21

Sokolov 28 Tukmakov 22
Speelman (1 5 ) Uhlmann 18

Stefansson 1 Wahls 42

Suba 20 Wilhelmi (27), (39)

Timman (19), 3 8 Yurtaev 10

Tiviakov (1 5 ) , 1 6 Yusupov 25

Tondivar 17 Zeller 45

Trauth (1 5 ) Zude 6
 Index of Material Balances

l. Pawns

2 8 v 18 1.2, 1.3 , 3 . 9 , 3 . 1 4

28v28 3 . 10 , 2 1.3

3 8 v 38 3 . 2 , 8 . 3 , 1 5 . 13

48v48 3 .4 , 3 . 8 , 3 . 1 3 , 3 . 1 6 , 1 5 . 5

5 8 v 58 2 . 3 , 3 . 7 , 3 . 11 , 3 . 1 5 , 2 9 . 3

Mv5� 37.7

M v M 2 . 1 , 2 . 2 , 3 .1 , 3 . 1 2 , 3 . 1 7 , 1 5 . 8

11. Minor pieces

1 1. 7 , 3 5 . 1 2 - 3 5 .1 3 , 44.6-44.7

7 .1 - 7 . 2 , 3 2 . 2

13 . 2 , 1 4 . 2 , 27.7

8 . 1 - 8 . 2 , 8 .4 , 9 . 2 - 9 .4 , 10 . 1 - 10 . 3 ,

1 1.1 - 1 1. 6 , 1 1. 8 - 1 1.1 0 , 1 5 . 10 - 1 5 . 1 2 ,

15.14, 38.3, 38.9

.i. v � ( opposite-coloured) 6 . 1 - 6 . 5 , 22.5

.i. v � (same-coloured) 4 .1 - 4 . 2 , 5 . 1 - 5 . 6 , 1 6 . 3 , 3 0 .3 , 3 1. 7

.i_ + 'tJ V 2étJ 12.1

2.i. V .i_ + étJ 13 .1 , 1 4 . 1 , 1 5 . 6 , 1 5 . 7 , 1 5 . 9 , 1 6 . 1 - 1 6 . 2 ,

1 6 .4

1 5 .4
1 7 4 Endgame Secrets

111. Rooks

27.3

17.1-17.2, 18.1-18.5, 19.8-19.14, 28.3,

3 0 .4 , 3 1. 8 , 3 4 . 7 , 3 7 . 1 0 - 3 7 . 11 ,

3 7 .1 3 - 3 7 . 1 4 , 3 8 . 7 , 3 9 . 6 - 3 9 . 1 0

z:t+ ctJ V z:t 3 8 .4

J:[+ ctJ V J:[ + ctJ 3 3 . 1 , 34.2-34.6, 3 4 . 8 - 3 4 . 1 0 , 3 5 . 2 - 3 5 . 1 1

z:t+ .i. V .i. + ctJ 1.1

z:t+ .i. V z:t 2 0 . 1 - 2 0 . 3 , 2 1.1 - 2 1. 2 , 3 8 . 1 0 - 3 8 .1 4

J:[+ .i. V z:t + ctJ 36.1-36.2, 37.3-37.6, 37.8-37.9, 38.2,

38.5-38.6, 38.8, 39.2-39.5

l:t+ .i. v l:t + .i. (opposite-coloured)22.1-22.4, 2 3 . 1 - 2 3 . 6 , 2 4 . 1 - 2 4 . 7 ,

2 5 . 2 - 2 5 .4 , 2 6 . 1 - 2 6 . 9 , 2 7 . 1 - 2 7 . 2 ,

2 7 .4 - 2 7 . 6 , 2 7 . 8 , 4 0 . 9

:+ .i. v : + .i. (same-coloured) 2 9 . 1 - 2 9 . 2 , 3 0 . 2 , 3 0 . 5 , 3 1. 9 - 3 1.1 0

z:t+ .i. + ctJ V ¡[ + .i. + ctJ 3 9 . 1 , 42.7

l:t+ 2.i. vl:t + i. + ctJ 25.1

2:[ V J:[ + ctJ 4 1.1 1

2:[ V J:[ + .i. 4 0 . 1 1 , 4 1. 7

2:[ V J:[ + .i. + ctJ 42.2

2:[+ ctJ V 2:[ + ctJ 32.1

2:[+ .i. V J:[ + .t + ctJ 1 9 .4, 4 1.1 - 4 1. 6 , 4 1. 8 - 4 1.l O

2:+ .i. v : + 2 .i. 4 0 . 2 - 4 0 . 8 , 4 0 . 10 , 4 0 . 1 2 - 4 0 . 13

2:+ .i. v2:[ 19.3

2:[ + .i. V 2l:t + ctJ 19.7

2: + .i. v2:[ + .i. 2 8 . 1 - 2 8 . 2 , 3 0 .1 , 3 1. 4 - 3 1. 6

2:[+ .i. + ctJ V 2l:t + .i. + ctJ 1 9 . 2 , 1 9 . 5 - 1 9 . 6 , 3 7 . 2 , 4 2 . 1 , 42.3-42.6

2:[ + 2.i. V 21:t + .i. + ctJ 3 1. 2

2:[ + 2.i. V 2:[ + 2.i. 15.3

21:t+ 2.i. + ctJ V 2:[ + 2.i. + ctJ 1 5 . 2 , 3 1.3

IV. =Queens

"vJJ/ v � 3.3, 3.5, 3.6

"vJJ/ v "vJJ/ 8 . 5 , 3 7 . 1 2 , 44.4-44.5

"vJJ/ + clJ v "vJJ/ 44.8

"vJJ/ + clJ v "vJJ/ + clJ 44.1-44.3

"vJJ/ + .i. v "vJJ/ + clJ 9.1

"vJJ/ + .i. v "vJJ/ + .i. 43.1-43.2

Other material balances 1 5 . 1 , 1 9 . 1 , 3 1.1 , 3 4 . 1 , 3 5 . 1 , 3 7 .1 , 3 8 . 1.

4 0 . 1 , 44.9, 4 5 . 1 - 4 5 . 2
 Bibliography

Y.Averbakh: Turmendspiele (2 H.Metz: Credit Suisse Masters

volumes); Sportverlag Berlin 1 9 8 6 Horgen 1 9 9 4 ; Edition Olms 1 9 9 5

Y.Averbakh: Bauernendspiele; J.Nunn: Secrets of Pawnless

Sportverlag Berlin 1 9 8 8 Endings; Batsford 1994

Y.A verbakh: Ldufer- und V.Smyslov: Die Kunst des

Springerendspiele; Sportverlag Endspiels; Verlag Bock und Knbler

Berlin 1 9 8 8 1996

Y.Averbakh: Endspiele Springer M.Shereshevsky: Endgame

gegen Ldufer, Turm gegen Strategy, Pergamon 1 9 8 6

Leichtfigur; Sportverlag Berlín 1 9 8 9 J.Speelman: Analysing the Endgame

Y.Averbakh: Damenendspiele; (2nd edition); Batsford 1 9 8 8

Sportverlag Berlin 1 9 9 0 J.Speelman: Endgame Preparation

A.Beliavsky and A.Mikhalchishin: (2nd edition); Batsford 1 9 9 7

Winning Endgame Technique; J.Speelman, J. Tisdall and B . Wade:

Batsford 1 9 9 5 Batsford Chess Endings; Batsford

A.Karpov: Wie ich kdmpfe und 1993

siege (2nd edition); Schachverlag K.Thompson: Endspiel CD 1-4;

Rudi Schmaus 1 9 8 4 ChessBase 1 9 96

H.Kmoch: Die Kunst der J. Timman: The Art of Chess

Bauernführung (2nd edition); Analysis; Cadogan 1 9 9 7

Verlag Das Schacharchiv 1 9 6 7

N. McDonald: Practica! Endgame I also consulted the following

Play; Cadogan 1 9 9 6 periodicals :

A.Matanovic (editor): Encyclo­ Chess Informant

paedia of Chess Endings (5 ChessBaseMagazin

volumes), SCHACH

Chess Informant 1 9 8 2 - 1 9 9 3 NIC Magazine