FIDE Surveys – Jeroen Bosch

1

Jeroen Bosch:

A classical lesson: Trading
Bishop for Knight: gaining
control over square d5

I recently saw an instructive win by the
Indian talent Praggnanandhaa over his
countryman Gopal which I would like to
share with you. From the game it’s obvious
that the youngster has already studied and
assimilated a lot of the classical examples
that are the backbone of a strong player’s
positional understanding.

Praggnanandhaa : Gopal, Gibraltar 2018

XIIIIIIIIY
9-+rtr-+k+0
9+-wq-vlpzp-0
9-+lzp-sn-zp0
9zpp+-zpP+-0
9-+-+P+-vL0
9zPLsNQ+-+P0
9-zPP+-+P+0
9tR-+-+RmK-0
xiiiiiiiiy

20.Bf6!
Trading bishop for knight is correct here.
White needs to gain control over square
d5.
In case of the useful 20.Kh1, Black gains a
decent game with 20...d5! 21.ed5 (Not
21.Bf6? de4.) 21...Nd5, when the game is
even after 22.Nd5 (Black defends after
22.Be7 Ne7 23.Qe2 Nd5 24.Bd5 (24.Nb5?
Nf4!) 24...Bd5 25.f6!? (25.Nb5? Qc2–+)
25...Bc4! 26.Qg4 g6 27.Qh4 Kh7.)
22...Bd5 23.Bd5 Bh4.
20...Bf6 21.Kh1
This preparatory move is essential. The
king needs to move out of the a7–g1
diagonal as so often in the Sicilian. Now
White threatens to take the b5 pawn.
The immediate 21.Nd5? fails to 21...Bd5

22.Qd5?! (22.Bd5 Qc2 23.Qb5 Rb8.)
22...a4 23.Ba2 Qa7 24.Kh1 Rc2. Likewise
it is bad to play 21.Bd5? b4.
21...Rb8
In case of 21...a4, White could consider the
subtle 22.Ba2 (White is of course also
better after the direct 22.Bd5, but Black
can still try 22...Be8!?, protecting f7 and
avoiding the trade of bishops: 23.Ba2
Qc5.) and now, if 22...Ba8 (After
22...Qb7 both, 23.Bd5 and 23.Nd5 will
give White a pleasant edge.) White has
23.Bd5!, forcing the trade of the light-
squared bishops: 23...Bd5 24.Qd5.
White has a distinct 'slight' plus after
21...b4 22.ab4 ab4 23.Nd5.

XIIIIIIIIY
9-tr-tr-+k+0
9+-wq-+pzp-0
9-+lzp-vl-zp0
9zpp+-zpP+-0
9-+-+P+-+0
9zPLsNQ+-+P0
9-zPP+-+P+0
9tR-+-+R+K0
xiiiiiiiiy

An interesting moment. White has a choice
between two scenarios:
- good knight vs bad bishop, or
- good biship vs bad bishop (opposite-
coloured). Which scenario would you
choose?
22.Nd5
After 22.Bd5, going for the good knight-
bad bishop scenario, Black gains some
counterplay with 22...b4 (22...Be8 23.Ba2.)
23.ab4 Rb4 (23...ab4 24.Bc6 bc3 25.Qc3
Rdc8 26.Bd5 Qc3 27.bc3 Rc3 is
unpleasant, but a draw.) 24.b3 (24.Bc6
Qc6 25.b3 (25.Nd5 Rb2) 25...Rd4, with
definite counterplay.) 24...Rd4 and Black
has his chances too, although White
remains better. In the game Black has no
counterplay whatsoever, and Praggna
shows that there is no need to fear the
'drawing' presence of opposite-coloured

FIDE Surveys – Jeroen Bosch

2

bishops. Indeed, with more pieces on the
board it is well-known that this just brings
an extra attacking piece on a complex of
squares where the defender has none.
22...Bd5 23.Bd5 Qb6 24.c3
Black has no counterplay, and White can
calmly prepare the advance of his kingside
pawns.
24...Rdc8 25.Rfd1
It was possible to immediately execute the
plan with 25.g3, followed by h3–h4, Kh1–
g2–h3, to prepare g3–g4–g5.
25...Rc5
Note that 25...b4 is always met by 26.cb4
ab4 27.a4, when Black's 'activity' has only
given White another winning idea - a
passed pawn on the queenside.
26.Bb3 Rd8 27.g3 Rc7 28.Kg2 Qc5 29.h4
Qb6 30.Rd2 Qb7 31.Kh3

XIIIIIIIIY
9-+-tr-+k+0
9+qtr-+pzp-0
9-+-zp-vl-zp0
9zpp+-zpP+-0
9-+-+P+-zP0
9zPLzPQ+-zPK0
9-zP-tR-+-+0
9tR-+-+-+-0
xiiiiiiiiy

Praggna has executed the first part of the
plan. The king is quite safe on h3 and the
pawns protect dark squares for the moment
- of course you need avoid that they will be
blocked (Pawns on g4 and h3 and Black's
bishop on h4).
31...Rdc8 32.Rad1 Rc5 33.Qf3
It's possible to take the pawn, but actually
safer not to go for such a small gain.
33.Qd6!? Qe4? 34.Bf7 Kf7 (34...Kh7
35.Bg6 Kg8 36.Qd8 mates.) 35.Qe6 Kf8
36.Rd7+– and mate follows. However,
Black should not take on e4, when White
has removed a weak black pawn which
opens the position and might render his
own king unsafe, when he starts to prepare
g3–g4–g5.

33...Be7
Again 33...b4 is just met by 34.cb4 ab4
35.a4.
34.Rf1
It was possible to play for a direct attack
with 34.f6!? Bf6 (34...Qd7 35.Kg2 Bf6
(35...gf6 36.Qh5+–.) 36.Rd6 Qe8 37.Qf5.)
35.Rd6.
34...Qd7 35.Kg2 Qb7 36.Bd5 Qd7
37.Bb3
A little faster is 37.Qh5 Rf8 38.g4+–, but
the beautiful thing about White's position
is that there is no hurry as the opponent has
no counterplay whatsoever.
37...Qc7?
37...Qb7.
38.Qh5! Bf6 39.Rfd1
Or 39.g4! Qb7 40.g5 Qe4 41.Rf3+–.
39...Rc6 40.Kh3
Another 'slow' move just before the time
control. 40.g4 or 40.Bd5, followed by g4.
40...b4
Black finally lost patience.
41.cb4 ab4 42.a4 Rf8

XIIIIIIIIY
9-+-tr-+k+0
9+qtr-+pzp-0
9-+-zp-vl-zp0
9zpp+-zpP+-0
9-+-+P+-zP0
9zPLzPQ+-zPK0
9-zP-tR-+-+0
9tR-+-+-+-0
xiiiiiiiiy

43.g4!
White is winning on both wings. The rest
is easy:
43...Qb6 44.Re2 d5 45.Bd5 Rd6 46.g5
Bd8 47.g6 Kh8 48.gf7 Bf6 49.Kh2 Qd8
50.Rg2 Bh4 51.Rdg1 Bf6 52.Rg6 Bh4
53.Rg7 1:0.

FIDE Surveys – Jeroen Bosch

3

This was 2018, let’s go back in chess
history with the following classical Fischer
game.

Fischer : Gadia, Mar del Plata 1960

XIIIIIIIIY
9r+-+-trk+0
9+-+-vlpzpp0
9pwqlzp-sn-+0
9+p+-zpPvL-0
9-+-+P+-+0
9+LsNQ+-+-0
9PzPP+-+PzP0
9tR-+-+R+K0
xiiiiiiiiy

1.Bf6!
Fischer trades his bishop for the knight to
gain control over square d5.
1...Bf6 2.Bd5!
An excellent follow-up. The light-squared
bishops are also exchanged to obtain a
good knight-bad bishop scenario. Black
cannot dispute White's control over the d5–
square and is clearly worse due to his
backward d6–pawn.
2...Rac8 3.Bc6 Rc6

XIIIIIIIIY
9-+-+-trk+0
9+-+-+pzpp0
9pwqrzp-vl-+0
9+p+-zpP+-0
9-+-+P+-+0
9+-sNQ+-+-0
9PzPP+-+PzP0
9tR-+-+R+K0
xiiiiiiiiy

4.Rad1
Too hasty is the immediate 4.Nd5, as
Black has sufficient counterplay after
4...Qd4!
4...Rfc8?!
It was more stubborn to prevent the knight

from occupying the d5–square with
4...Qc5.
5.Nd5 Qd8 6.c3
Fischer has obtained the position he had in
mind when trading on f6.
6...Be7
How did Fischer continue?

XIIIIIIIIY
9-+rwq-+k+0
9+-+-vlpzpp0
9p+rzp-+-+0
9+p+NzpP+-0
9-+-+P+-+0
9+-zPQ+-+-0
9PzP-+-+PzP0
9+-+R+R+K0
xiiiiiiiiy

7.Ra1!
Very elegant. Fischer prepares to start a
second front by opening up the a-file, with
a2–a4. Playing for a direct attack with 7.f6
Bf6 8.Nf6 gf6 9.Qh3 was possible (Black's
kingside structure is damaged with 7.f6.)
The concrete 7.Nb4 is met by 7...Rb6,
when White can play 8.f6 after all 8.f6 (Or
repeat the starting position with 8.Nd5.)
8...Bf6 9.Nd5. A multi-purpose move, like
7.Rf3, does not spoil anything either.
White may either decide on a kingside
attack, or double rooks on the d-file.
Gadia now succumbed to the pressure:
7...f6?
A horrible move.
8.a4 Rb8??
Overlooking a double attack, Black
resigned after
9.Ne7 1:0.

1. The trade of the g5-bishop for the
knight, to take control over the d5-square,
is one useful lesson from this fragment.
2. Fischer’s subsequent trade of the light-
squared bishops with 2.Bd5 is another
point to remember. In this way he obtained
complete control over the strong square.

FIDE Surveys – Jeroen Bosch

4

In the next two fragments we will return to
both elements.

Karpov : Mecking, Hastings 1972

XIIIIIIIIY
9r+-+k+-tr0
9+p+-vlpzpp0
9pwqnzp-sn-+0
9+-+-zpPvL-0
9P+-+P+-+0
9+PsN-+-+-0
9-zP-+L+PzP0
9tR-+QmK-+R0
xiiiiiiiiy

13.Bf6
In this case a more or less compulsory
trade. White cannot keep his king in the
centre with the queens on.
13...Bf6 14.Nd5 Qa5
Black's position is also unattractive after
14...Qd8 15.0–0 0–0 16.Rc1, with a clear
edge.
15.Qd2 Qd2 16.Kd2 Bg5
Disturbing White's coordination for the
moment.
17.Kd3 0–0 18.h4 Bd8
18...Bh6 19.g4 is awkward.
19.Rac1 a5
19...Nd4 20.b4 is given by Karpov, when
Black could still try 20...a5, but White
keeps an edge. Karpov points out that
Black cannot seek relief in exchanging the
knights 19...Ne7 20.Ne7 Be7, as White
infiltrates the seventh rank with 21.Rc7.
20.Kd2 Rb8 21.g4 Nb4
Mecking seeks relief in a further trade,
hoping to draw because of the opposite-
coloured bishops. However, with all the
rooks on, White keeps sufficient attacking
chances.
22.Bc4 Nd5 23.Bd5 g5
Mecking hopes to block the kingside.
Passive defence will not save Black. For
example: 23...Kh8 24.g5 Bb6 25.Kd3 Bc5
26.Rc2 b6 27.h5 Rbc8 28.g6 fg6 29.hg6 h6
30.Bf7 and Black must give up an

exchange, to avoid a deadly f5–f6.
24.fg6 hg6 25.Kd3 Kg7 26.h5 Bb6
26...Bg5 would allow 27.Rc7.
27.Rh3 Bc5 28.Rf1
Now the threat is 29.h6, followed by the
win of the f7–pawn.

XIIIIIIIIY
9-tr-+-tr-+0
9+p+-+pmk-0
9-+-zp-+p+0
9zp-vlLzp-+P0
9P+-+P+P+0
9+P+K+-+R0
9-zP-+-+-+0
9+-+-+R+-0
xiiiiiiiiy

28...f6 29.hg6 Kg6 30.Rfh1
Karpov has a winning attack.
30...Rbe8 31.Rh7 Kg5 32.Ke2 Kf4
33.R1h3 Bd4
33...Kg4 34.Rh1 (or 34.Rf3), forces
34...Rg8 and White can win in many ways,
apart from just taking the exchange.
34.Rg7
And Black lost on time, but 35.Rf3 mate is
unavoidable anyway. 1:0.

Note that, because of the peculiar nature of
the knight move, it is possible to increase
your control over a light square (d5) with a
dark-squared bishop (Bg5xf6)!
In Fischer’s game the exchange of bishop
for knight led to a good knight-bad bishop
scenario. In Karpov’s game the future
World Champion obtained a light-squared
bishop that turned out to be no less
superior to the dark-squared bishop than
Fischer’s knight had been.
Interestingly, the attack on the kingside
remained very strong even in the absence
of queens (as opposed to our first fragment
Praggnanandhaa : Gopal.
I am sure that Praggnanandhaa knew both
these classical examples and has stored
these patterns into the ‘positional database’
inside his brain.

FIDE Surveys – Jeroen Bosch

5

In our next example the dark-squared
bishop has already been exchanged for the
knight.

Gashimov : Godena, Reggio Emilia 2010

XIIIIIIIIY
9r+-+-trk+0
9+-+-+p+-0
9pvllzp-wqnzp0
9+p+-+-zp-0
9P+-+P+-+0
9+LsN-+NzP-0
9-zPQ+-zPP+0
9tR-+R+-mK-0
xiiiiiiiiy

21.Bd5!
Trading the defender of square d5 (And of
pawn b5). Quite similar to Fischer : Gadia,
wouldn't you say? 21.Nd5 Bd5 22.Bd5 is
only a slight edge for White (In
comparison to Karpov : Mecking Black's
dark-squared bishop is much more active.)
21...Rac8 22.Bc6!
With simple means White obtains a huge
positional edge. The tactical solution does
not work here: 22.e5?! Ne5! (22...de5?
23.Ne4+–.) 23.Ne4 (23.Ne5 de5 24.Ne4
and now Black has 24...Qg6; 23.Bc6 Rc6
24.ab5 ab5 25.Qe4 fails to 25...Rc4.)
23...Nf3 24.gf3 Qf3! 25.Bc6 Rc6! 26.Qc6
and now the pin allows 26...Bf2! 27.Kh2
Bg3 28.Kg1 Bf2 29.Kh2 g4 and Black's
attack will force White to return material
with interest.
22...Rc6 23.ab5 ab5 24.Qe2
White is clearly better as both pawn b5 and
square d5 are weak.
24...Rfc8
The lesser evil was 24...Ne7 25.Nb5.
25.Nd5 Qd8 26.Qb5
White is winning.
26...g4 27.Ne1 Ne5 28.Nd3 Bd4 29.Ra4!
Rb8??
Overlooking the same knight fork as Gadia
did versus Fischer.
30.Ne7 1:0.

In our next example White will play with a
pair of knights versus a pair of bishops –
all for the sake of full control over square
d5.

Bosch : Armas, Wijk aan Zee 1995

XIIIIIIIIY
9r+r+-+k+0
9+lwq-vlpzpp0
9pzp-zp-sn-+0
9+-sn-zp-vL-0
9P+-+P+-+0
9+LsN-+-+-0
9-zPPsNQzPPzP0
9tR-+R+-mK-0
xiiiiiiiiy

15.Bf6!
The by now familiar trade to gain control
over d5.
15...Nb3
A logical in-between move, but this eases
White's task.
Perhaps it would have been more practical
to give White the choice between playing
for opposite-coloured bishops or for good
knight-bad bishop. At least it would have
given him something to think about:
15...Bf6 16.Nd5 (Even more ambitious is
16.Bd5, but Black has some counterplay
after 16...Ne6 17.Bb7 Qb7 18.Nc4 Qc6
19.Ne3 Nd4 20.Qd3 Bg5.) 16...Bd5
17.Bd5, clearly favours White.
16.cb3!
16.Nb3 Bf6 is nothing for White (The
knight is misplaced on b3.)
16...Bf6 17.Nc4!
In his chessbase comments Donev now
makes a useful remark: »An important
advantage of a pair of knights, compared to
the bishop pair, is that both knights can
attack one square at the same time (In this
case square d5.), while only one bishop can
defend this point.« (My translation from
the German.)
17...Rd8 18.Ne3! Bg5 19.Ned5!±
According to Donev the knight is as

FIDE Surveys – Jeroen Bosch

6

powerful as a rook here.
19...Bd5
19...Qc6? 20.Qg4 h6 21.h4+–; 19...Qc5
20.b4 Qc6 21.Qg4+–.
20.Rd5 Qb7 21.Rad1 Be7 22.R5d3 Rdc8
23.Nd5
Obviously White is much better, due to his
superior knight, the backward d6 pawn,
and his full control over d5.
23...Rc6 24.Rc3 Rac8
24...Bd8 25.b4 (25.Rc6 Qc6 26.Qc4 Qc5
27.g3± Bosch) 25...Rc3 26.bc3±, Donev.

XIIIIIIIIY
9-+r+-+k+0
9+q+-vlpzpp0
9pzprzp-+-+0
9+-+Nzp-+-0
9P+-+P+-+0
9+PtR-+-+-0
9-zP-+QzPPzP0
9+-+R+-mK-0
xiiiiiiiiy

25.Ne7
It's a pity to trade the superior knight, but it
brings a clear pawn and a winning position
to boot, while 25.Rc6 Rc6 26.Nb4 Rc8
27.Na6 d5!? 28.ed5 Bd6 unnecessarily
complicates matters.
25...Qe7 26.Rc6 Rc6 27.Qa6 Kf8 28.g3
Qc7 29.Qd3 Ke7
Protecting d6 to activate the rook, but now
his king will be a target as well.
30.Qe3 Rc2 31.Rd3!
Planning Rc3.
31...Rb2 32.Rc3 Rb1 33.Kg2
White has given up his extra pawn for
control over the c-file and a direct attack
on the black king.
33...Qb7 34.Qg5 Ke8
34...Kf8?? 35.Qd8#.
35.Qg4 Kd8?! 36.Qh4 Ke8 37.Qh7 g6
38.Qh4 Qa6 39.Rc4
Not 39.Rc7? Qf1; 39.Qh8 Kd7 40.Kh3!
also wins.
39...Rb3 40.Rc7
And mate is unavoidable. 1:0.

Let’s have one final example where
Black’s dark-squared bishop is more
actively placed on the queenside. The
result is the same though: Black cannot
defend on the kingside. Instead of moving
forward with the kingside pawns as
Karpov and Praggnanandhaa did, Black
will succumb because of the weakness of
f7.

Almasi : Navara, Reggio Emilia 2008

XIIIIIIIIY
9-+-+r+k+0
9+-wq-+pzpp0
9-+-zplvl-+0
9trp+-zp-+-0
9-+-+P+-+0
9+LzP-sNQ+P0
9P+-+-zPP+0
9tR-+-tR-mK-0
xiiiiiiiiy

26.Nd5
Correctly going for opposite-coloured
bishops (Remember Karpov-Mecking?)
After 26.Bd5 even 26...Qc3 is possible.
26...Bd5 27.Bd5
White's bishop is superior and with all the
heavy pieces still on the board that is a
major advantage. Moreover, Black has
weak pawns on b5 and f7.
27...Rc8 28.Re3 Qd7
28...Ra3 29.Rb1 Rc3 30.Rc3 Qc3 31.Qc3
Rc3 32.Rb5 h5!±.
28...Qe7 29.Rd1 g6 30.Red3 Kg7 31.Bb3
Ra6± was Black's best chance, according
to Navara, who still considers the position
to be strategically lost in view of the
weaknesses on b5, d6 and f7.
29.Rd1 Bd8?! 30.Qh5 Bb6?
This loses, Navara writes that even though
Black's position remains bad it was
essential to play 30...g6 31.Qh6 Bf6 32.Rf3
Qe7.

FIDE Surveys – Jeroen Bosch

7

31.Rf3+– Rf8

XIIIIIIIIY
9-+-+-trk+0
9+-+q+pzpp0
9-vl-zp-+-+0
9trp+Lzp-+Q0
9-+-+P+-+0
9+-zP-+R+P0
9P+-+-zPP+0
9+-+R+-mK-0
xiiiiiiiiy

White to play and win:
32.Bf7!
32.Rdd3 also wins.
32...Rf7 33.Rd6! Qa7 34.Qf7
Understandably transposing into a won
ending. Even stronger was 34.Qe5! Rf3
35.gf3 and, to avoid mate, Black must give
up the bishop after all: 35...h6 (35...Bf2
36.Kf1 h6 37.Qe8 Kh7 38.Qg6 Kg8
39.Rd8#.) 36.Qe6 Kh7 37.Rb6+–.
34...Qf7 35.Rf7 Kf7 36.Rb6
The rook ending is an elementary win.
36...g5 37.g3 h5 38.Kf1 h4 39.Kg2 Ra2
40.Rb5 Kf6 41.Rb6 Ke7 42.gh4 gh4
43.Rh6 Rc2 44.Rh4 Rc3 45.Rh5 Ke6
46.Rf5 Rc4 47.f3 Rc2 48.Kg3 Rc1 49.h4
Rg1 50.Kh2 Rg6 51.Kh3 Rg8 52.h5 Rg1
53.Kh4 Rh1 54.Kg5 Rg1 55.Kh6 Rg3
56.Kh7 Rg2 57.h6 Rg3 58.Kh8 1:0.

Conclusion
Our first (modern) game fragment
illustrated the usefulness of studying the
classics. We have seen classical examples
of the trade of a bishop for a knight on f6.
The gain is clearly full control over the
square d5. The backward d6-pawn and the
e5-pawn severely hindered Black’s dark-
squared bishop in most of the above
fragments.
Fischer entered a strong knight-bad bishop
scenario by means of the further trade of
light-squared bishops, while Karpov
demonstrated that opposite-coloured

bishops in such a structure may prove no
less favourable for the stronger side.
Depending upon the concrete features of
the position you may go for the one or for
the other.