White: David Bronstein
Black: Yefim Geller
1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3!
It puzzles me that the modern school of young grandmasters eschews this obvious move in favour of the tepid 4.Qc2 or 4.e3. Black ought to be punished for his impudence in pinning the knight, and 4.a3 is clearly the move to do it.
4...Bxc3+ 5.bxc3 0-0 6.f3
Black must now either allow e4 or play d5 and let White undouble his c-pawns.
6...d5 7.cxd5 exd5
Natural enough, but 7...Nxd5 earns my vote. If a man believes in his knights, he should use them!
8.e3 Bf5 9.Ne2 Nbd7 10.Nf4 c5 11.Bd3
A paradoxical move: White exchanges half his bishop pair, leaving behind the apparently less effective cleric. But there are deep plans afoot.
11...Bxd3 12.Qxd3 Re8 13.0-0 Rc8 14.Rb1 Qa5 15.Rxb7!
Calling Black's bluff. After 15...Nb6, the rook seems cut off. Bronstein dem-onstrates that it is well cut in.
The beauty of this move is twofold: firstly, it meets Black's threat of a breakthrough down the c-file by gratuitously weakening White's king; and secondly, White must have calculated it all to the end even before taking the pawn on b7.
16...h6 17.h4! cxd4 18.g5! dxe3 19.gxf6 Rxc3 (see diagram)
Black has two huge pawns and a pawerful initiative for his piece, right?
Wrong! In fact it's mate in three (as Bronstein indeed announced on playing the move): 20...fxg6 21.Rxg7+ Kf8 (or Kh8) 22.Nxg6 mate.Reuse content