THE RIGHT EXCHANGE 105

105

9 g4! �al
9 . ..toeS ? ! 10 gS and then :
a) 1O . . .hxgS ? 1 1 hxgS tOd3 1 2 gxf6 tOxf4+ B
1 3 exf4 'it>xf6 ( l 3 . . . gxf6? 14 fS+ ! +-) 14 tOgS
+-.
b) 1O . . .tOd3 1 1 Mg4 ( 1 1 hS+ ! ?) 1 l . . .�b2
( l l . . .�al 12 'it>h l ±) 12 hS+ 'it>f7 13 'it>g3 ±.
10 Mel �b2 11 Me2 �al 12 a4 ! ? tOeS 13
tOxeS+
White can no longer avoid the exchange, but
his attack continues unabated.
13 ... �xeS 14 MfS Mdd7? ! 15 f4 Me7 (D)

Now there i s nothing to be done against the

white attack. Only 20. . .g6 still offered any
chance.
21 'it>f3 Mf7 22 MhS eS 23 �g4 exf4 24 MdS !
With the fatal threat of �hS#.
24...fxe3+ 25 'it>g3 1-0

Exercise
(So lut io n on page 304)

16 Md2
White evades the exchange, so as not to di­
minish his attacking potential .
16 ... �e3 17 Md6 'it>h7 IS gS hxgS?
Further resistance was possible only with
I S . . . g6.
19 hxgS �b4?
An error, but his position was already very
difficult.
20 MddS? (D)
Missing a forced mate : 20 g6+ ! ! 'it>h6 2 1
�fS exfS 2 2 MhS#. E4 . 3 9 **/
20...'it>g6?
Is 1 bS a good choice?
5 Th i n k i n g i n Sc h e m es

I know at sight what a position contains. What

could happen? What is going to happen ? You
A) Ta rget Position s
figure it out. I know it!
JOSE RAUL CAPABLANCA I f you have everything under control, but are at
a loss how to make progress, it is sometimes
Strategy without tactics is the slowest route to beneficial just to visualize desirable target posi­
victory. Tactics without strategy is the noise be­ tions and only then to examine whether and
fore defeat. how you can achieve them.
SUN Tzu

On the one hand, chess is a very concrete game,

in which even the smallest alteration to a posi­
tion can have important repercussions; on the
other hand it would be hardly possible for hu­
mans to master the game if they could not make
use of certain schemes and rules of thumb. In
the endgame, a schematic way of thinking is
particularly appropriate. By this we mean the
ability to recognize desirable positions and
piece set-ups, and then work out a plan to reach
them. If you ask a grandmaster about the as­
sessment of an endgame position, you won't
generally hear any concrete variations at first,
5.01
but rather phrases such as "If White succeeds in M . Ka rttunen - V. Rasik
exchanging the rooks he should win" or "If Rethymnon E C C 2003
Black manages to put his bishop on the long di­
agonal, he has a certain draw". In view of the wrong rook' s pawn, White
Typical themes and issues that lend them­ cannot allow his knight to be exchanged. But
selves to schematic thinking are, for example: then how can he win? If you try to find a win
• how to match the pieces with the pawn-struc­ here by concrete calculation alone, you will im­
ture mediately realize that this is by no means easy.
• optimal deployment of the pieces or improv- Instead of wasting time on difficult computa­
ing the piece set-up tions, it is much more here for White to
• consideration of the right exchange think about the best possible arrangement for
• fortresses the pieces . Black' s bishop is already restricted
weak squares
as much as possible; the knight will need to
•

knight outposts
support promotion threats, so the idea comes of
•

pure opposite-coloured bishop endings

sticking the king on bS and the knight on dS. In
•

Of course, when applying this way of think­

fact, once this set-up has been achieved, the win
ing, it is absolutely essential not to think about
is quite easy.
the position in too static a manner - on no ac­
count should you neglect the dynamic factors. 1...WaS 2 WdS Wb7 3 �d4 WaS 4 CLlgS Wb7
5 CLle6 �h2 6 Wc4 �g3 7 WbS �h2 S �b6
107 Ho wTHINKING
TO PLA Y IN SCHEMES
CHESS ENDGAMES 107

�g3 9 'Zlc5+ Was 10 'Zle4 �h2 1 1 'Zlc3 Wb7 2 2 �d4 Wc8 23 �a7 Wd8 24 'Zld5
12 �d4 �g3 13 'Zla4 �h2 14 �f2 �f4? ! The first step on the W; the knight will end
This makes it easy for White. 1 4 ... �c7 ! 1 5 up on g7, via e7-fS, each time preventing the
'Zlc3 (D) and now the black bishop can twist black king from heading back to the safe as­
and tum all it likes, but there is no escape, as comer. From e7 the knight controls cS and from
Hecht showed in CBM: g7, eS.
24 We8 25 Wd6 Wf7 26 'Zle7 Wf6 27 �e3
•••

Wf7 28 �d4 We8 29 We6 Wd8 30 �b6+ We8

31 'Zlf5 Wf8 32 �c7 We8 33 'Zlg7+ Wf8 34 Wf6
B Wg8 35 �d6 Wh7 36 'Zlf5 Wg8 37 Wg6 Wh8
38 �c5 1-0

B) M a ke a Wish !

This theme is closely related to the last section

on target positions, only in this case you allow
your creativity completely free rein.

a) 1 5 . . . �h2 1 6 'ZldS �d6 ( l 6 . . .�eS 1 7 'Zlb6

is similar) 17 'Zlb6 Wxa7 I S 'ZlcS++ +-.
b) IS . . .�eS 1 6 'ZldS �h2 17 waS (zugzwang)
+-.
c) 1 5 . . .�dS 16 'ZldS �gS ( l 6 . . .WaS I 7 Wa6
+-) 1 7 'Zlc7 +-.
15 'Zlb6 Wxa7 16 'Zld5+ Wb7 17 'Zlxf4 Wc7
18 �c5 Wb7 19 'Zld5 Wb8 20 Wc6 Was 21
'Zlc7+ Wb8 (D)

5. 02
N . Vlasov - D. Sokolov
Alushta 1 994

At first sight White's defensive set-up seems

impregnable, but we should not give up so eas­
ily. Which black piece could still be improved?
That's right : the king. But where would the
king go if he had a free choice of any square on
the board? Obviously: to as , since then the bS­
pawn would fall . From this a clear plan can be
formed, one which is achievable:
1...We4 2 �d3+ We3 3 �f1 �e6 4 �d3
Now we have achieved the standard posi­ �g4 5 �g6 �e2 6 �e8 �f1 7 �c6
tion, in which the famous 'W-manoeuvre' of Also after the more tenacious 7 �d7 ! ? there
the white knight comes into effect (incidentally, is a way through: 7 . . .We2 S Wc2 �g2 9 �fS
this is also very schematic) . We i 10 �g6 �f3 1 1 Wc l ( I I �fS �d l + 1 2
108 Ho wTHINKING
TO PLA Y IN SCHEMES
CHESS ENDGAMES 108

�c3 jJ,e2 13 jJ,d7 �d 1 -+) 1 1 . . .jJ,e2 1 2 jJ,e8 6 �h6

jJ,d3 13 jJ,c6 - so far Vlasov and Shipov in White has reached the haven of his dreams
Informator. Our analysis continues 13 . . .�e2 and can now set the b-pawn rolling.
14 jJ,d7 �e3 15 jJ,c6 �d4 16 �b2 c4 17 jJ,e8 6...l::!.d 5 7 b4
c3+ 18 �b3 jJ,e4 19 jJ,h5 �d3 20 jJ,d 1 �d2 7 l::!.b 7+ �f6 8 ctJe7 l::!.d 3 9 ctJxg6 l::!.x g3 1 0
-+. ctJf4 +-.
7...�e2 8 �c2 jJ,h3 9 jJ,e8 jJ,f5+ 10 �c3 7...jJ,dl (D)
�dl 1 1 jJ,h5+ �c1 12 iLe2 �bl 13 iLc4 7 . . .jJ,f5 8 b5 jJ,e4 9 l::!.b 7+ �f6 10 l::!.b 8 �f7
1 3 �b3 jJ,e6+ 14 �c3 �a2 -+. 1 1 b6 l::!.b 5 1 2 l::!.b 7+ �f6 1 3 l::!.d7 +-.
13 �aI 14 �b3 jJ,bl ! 15 �a3
••.

15 �c3 jJ,a2 16 jJ,d3 jJ,d5 intending . . .

�a2-
a3 -a4-a5 -+ and 15 jJ,e2 jJ,a2+ 16 �a3 jJ,d5 1 7
jJ,d3 jJ,e6 (zugzwang) 1 8 jJ,e4 jJ,c4 1 9
jJ,c6
�b l - + are both lines given by Vlasov and
Shipov.
15...jJ,c2 16 jJ,e2 �b1 17 jJ,fl
The stalemate trick 17 jJ,d l ! ? is answered by
17 . . .�c l ! .
17 ... �c1 18 jJ,c4 jJ,d1 19 jJ,e6 jJ,e2 20
jJ,d7
�c2 21 jJ,c6 �c3 22 jJ,d7 c4 23 jJ,c6 �d2 0- 1

8 b5 jJ,a4 9 ctJa7 jJ,c2 10 l::!.c6 jJ,e4 11 b6

1 1 ctJc8 would have won more quickly.
1 1...l::!.d 8 12 l::!.c8 l::!.d313 ctJc6 jJ,xc6 14 l::!.c 7+
jJ,d7 15 b7 l::!.b316 l::!.xd7+ �e6 17 l::!.g 7 �f5 18
l::!.xg6 l::!.xb7 19 l::!.g 5+ �f6 20 �xh5 1 -0

C) I m provi ng the Pos ition of

a P i ece a n d M a kogo n ov's
Pri nci pl e
5 .03
I f there i s n o question o f a race, and i f there is
Av. Bykhovsky - M . Roiz
no obvious candidate move, then Makogonov 's
Israeli Cht 2004
Principle, that you should improve the position
The white king would very much like to sup­ of your worst-placed piece, can come in very
port the attack, without being harassed by the useful. You could even get into the habit of ex­
enemy rook. But before the king can set off for tending this to sounding out all your pieces, to
g5 or h6, the knight must be posted in a more check whether they are happy and have options
central position : for the future. If a piece is in a very flexible po­
l ctJd4 l::!.d l 2 �e3 l::!.e l+ 3 �f4 l::!.d l 4 ctJc6 sition and can quickly come into action, any­
l::!.fl + 5 �g5 l::!.f5+? where on the board, then generally this piece is
Only the bold 5 . . .�e6 ! ! held out any promise well-placed. But if a piece is going to be tied
of salvation, based on the activity of Black' s down for a long time performing a not so impor­
pieces .
tant function, then it should always be regarded
109 Ho wTHINKING
TO PLA Y IN SCHEMES
CHESS ENDGAMES 109

as a candidate for improvement. Quite often 7 . . .tDxc5 8 tDxc5 J::1a8 was more tenacious,
improving the knight is the key to solving the but of course Black is strategically busted.
problem of the position. S tDb4 2l.b7 9 tDxb7 J::1bxb7 10 J::1xc6 as 1 1
J::1xb6 axb4 1 2 J::1xb7 J::1xb7 13 J::1xb4 1-0

In the next example too, things very much

tum on the prospects of both sides' knights.

5 . 04
J . Capabla nca - J . M ieses
Bad Kissingen 1 928 B
In this ' Karlsbad' pawn-structure, d3 and d6 5.05
are ideal squares for knights, since from there E m . Lasker - J . Ca pabla nca
they can be redeployed quickly and t1exibly on St Petersburg 1914
both wings.
1 tDe1 ! �eS 2 tDd3 J::1b S 3 f3 J::1e 7 4 �f2 1...2l.b7?
tDfS 5 J::1b 3 tDd7 6 g4 b5? Now the e6-square and the d6-pawn will be
If Black remained passive with, for example, very weak. Instead it was much better to keep
6 . . .�d8, then White could open a second front the pawn- structure as it is and bring all the
on the kingside, when tbe d3-knight would again pieces to good squares, as Capablanca indi­
render good service : 7 h4 as 8 tDb6 tDxb6 9 cated after the game : 1 . . .2l.xf4 ! 2 J::1xf4 c5 ! 3
J::1x b6 J::1a8 10 tDf4 �d7 1 1 g5 ±. J::1d l 2l.b7 4 J::1f 2 J::1a d8 5 J::1fd2 J::1x d2 6 J::1xd2
7 tDac5 (D) tDc6 7 J::1d7 J::1c 8 and in view of his activity and
the great e5-square for his knight Black stands
no worse. This schematic evaluation works well
here. Of course, you should still always verify
this with some concrete variations.
2 2l.xd6 cxd6 3 tDd4 J::1a dS?
3 . . .2l.c8 was necessary. Capablanca underes­
timates the power of the knight landing on e6.
4 tDe6 J::1d7 5 J::1a dl tDcs 6 J::1f2 b5 7 J::1fd2
J::1de7 S b4 �f7 9 a3 2l.aS? !
9 . .J::1xe6 1 0 fxe6+ J::1xe6 was by far the least
.

of the evils, even though White retains good

winning chances.
10 �f2 J::1a7 1 1 g4 h6 12 J::1d3 as?
Ultimately the opening of the a-file just helps
White, whose forces can more easily exploit it
owing to his space advantage.
7...tDb6?
110 Ho wTHINKING
TO PLA Y IN SCHEMES
CHESS ENDGAMES 110

13 h4 axb4 14 axb4 Mae7 15 �f3 Mg8 16

�f4 g6 17 Mg3 g5+ 18 �f3 ttJb6 19 hxg5 hxg5
20 Mh3! Md7 21 �g3 �e8 22 Mdh1 �b7 (D)

5 . 06
Origi nal
the d5-pawn : 4 . ..�d 7 5 �e5 �c6 6 � e6 �c7 7
�xd5 +-.
23 e5! 3 ttJf5+ �d7 4 �e5 �e4 5 ttJe3 �c6 6 �e6
A beautiful clearance sacrifice, which allows �f3 7 ttJf5 �g4 8 �e5 �d7
the last white piece to j oin in the attack power­ Now both black pieces are out of position
fully via e4. and 9 ttJe3 +- is decisive.
23 ... dxe5 24 ttJe4 ttJd5 25 ttJ6c5 �c8 26
ttJxd7 �xd7 27 Mh7 Mf8 28 Mal �d8 29 Endgames with rook vs bishop are similar to
Ma8+ �c8 30 ttJc5 1-0 knight vs bishop endgames where the side with
the knight is in control .

D) Pl an s
Developing very elaborate long-term plans in
the opening and in the middlegame is generally
inappropriate, since too much depends on what
the opponent does. In the endgame, however, it
can be a very different story, especially if the
opponent has no real counterplay.
In the following diagram, White 's winning
plan consists of the following steps:
Step 1 : knight to f5 ;
Step 2: king to e5 ;
Step 3 : king to e6;
Step 4: threaten to transfer the knight to e7,
5.07
which leads to the diversion of both defenders
and consequently to the win of the d5-pawn
J . S peel m a n - I . Nataf
Esbjerg 2001
which, with the active white king, will decide
the game. Speel man 's plan is as follows : first he fixes
1 ttJg6 �g2 2 ttJh4 �h1 the enemy pawns on dark squares, and then he
The pawn ending after 2 . . �h3 3 ttJf5+ is
. keeps using the same zugzwang-based scheme
hopeless for Black, since after 3 . . .�xf5 4 �xf5 over and over again to force back the black
the white king is on a key square with respect to king:
 111 HowTHINKING
TO FLA Y IN
CHESS ENDGAMES
SCHEMES 111

1 f4! gxf4 2 �f3 J/.e5 3 � e 4 � e7 4 �bl �f6 �h7 20 �f7 J/.d4 21 �xd6 J/.e3 22 �g6 J/.f2
5 �xb6 �e6 6 f3 (D) 23 �f6 J/.e3 24 �g5 J/.d2 25 �d6 J/.e3 26
�d7+ �g8 27 �f6 J/.b6 28 �g6 1-0

Of course, a plan can also be very long-term

and include many intermediate goals.

Now, owing to White's light-squared domi­

nation and zugzwang, Black will be forced fur­
ther and further back. The underlying scheme is
typical of the struggle of a rook against a bishop
5 . 08
which lacks targets.
N . Rashkovsky - A. Orlov
6...�f6 7 �c6 �e6 8 �a6 �f6 9 �d5 �f7 St Petersburg 1 999
10 �a7+ �f6 11 �d7 �g5 12 �d8 (D)
1 �b5 !
White should exchange all the major pieces,
so that his king can then penetrate the queen­
side unhindered. In contrast, the minor pieces
should all remain on the board, at least for the
time being, so that White 's space advantage can
play its part.
1...�xb5 2 ct:Jxb5 ct:Jb6
Also after 2 . . .J/.e7 3 ct:Jc4 ± Black would not
be able to delay the exchange of rooks for long.
3 �xa8 ct:Jxa8 4 ct:Jc4 J/.e7 5 �f2 ct:Jac7 (D)

w
12...�h4
1 2 . . .�f6 1 3 �e8 �f7 14 �e6 �f8 15 �e4
�f7 16 �f5 �f8 17 �g6 (zugzwang) +- was
given by Hazai in CBM. We continue the analy­
sis as follows: 17 . . .J/.c3 1 8 �xd6 �e7 ( l 8 . . .J/.e5
19 �d5 J/.c7 20 �d7 J/.e5 21 �f5 +-) 1 9 �d l
�e6 20 �g5 J/.e5 2 1 �el �d5 22 �f5 J/.c7 23
�e4 +-.
13 �g8 �h5 14 �g4 �h6 15 �e6 �h5 16
�f5 �h6 17 �g2 �h7 18 �g6 �h8 19 �e6
112 HowTHINKING
TO FLA Y IN
CHESS ENDGAMES
SCHEMES 112

6 lLla7!
Naturally White declines to exchange, since
the black pieces are treading on each other's B
toes. In fact this example could also have ended
up in the section ' Converting an Advantage
,
(Space Advantage) .
6...h5
6. . .lLlf6 7 �f3 �fS S �d2 h5 9 �a5 lLlceS 1 0
h3 ± .
7 �e2 �f7 8 �d3 lLla6 9 lLlc8 lLlac7 1 0
�d2 lLlb5 1 1 lLla5 lLlf6
l l . . . l L lec7 lLlc6 �fS 1 3 �a5 lLleS 1 4
lLldS+ �f6 1 5 lLle6 �e7 1 6 c3 (zugzwang) + -
(Alterman in CBM).
12 lLlb7 lLle8 13 �a5 lLld4 14 c3 c4+ 15 E5.0 2
�xc4 lLlc2 16 �b5 lLle3 17 �c6 lLlxg2 How did Black demonstrate to his opponent
18 lLlbxd6+ lLlxd6 19 lLlxd6+ �f6 20 lLlc4 the hopelessness of his situation?
g5 21
d6 g4 22 dxe7 �xe7 23 lLlxe5 lLle3 24 �c7
lLlfl 25 lLlg6+ �f6 26 lLlh4 �g5 27 lLlg2 1-0

In conclusion, we should once more like to

give you a warning :
1 ) No amount of thinking in schemes can
completely rule out the need for concrete calcu­
lation.
2) You should not think in too static terms
and should never rule dynamic possibilities out
of consideration.

Exercises E5.03 *�
(Sol utions on pa ges 3 0 5 - 7 ) How can White further improve hi s position?

E5.01 **/ E5.0 4 * * */

How should White make progress? Work out a winning plan for White.