This position appeared in tonight's club championship game between Jon Munsey and Richard Smith. White is two pawns up, but has been under heavy pressure for most of the game.
Now black could have maintained this heavy pressure with 26...f3 27.gxf3 Bxf3, with the possibility of 28...Re8 in the offing. This would have given him quite substantial compensation for his sacrificed material.
Instead, he thought he could finish the game off immediately, and played 26...Re3+??. This proved in the game to be the right decision, as white blundered in response with 27.Kf2, allowing 27...Re2#. However, the right response was the simple 27.dxe3, after which Richard's planned 27...Rd1+ 28.Kf2 Bxe3+ has one small flaw: it isn't mate. 29.Bxe3 leaves white a rook up with a winning position.
Why did Richard overlook this? It was probably a "visual error", one caused by not actually seeing how the pieces are going to be moving. The bishop on c1 isn't capable of moving to e3 in the diagram position, so it's harder to see that it will be able to do so when the d-pawn moves away. In addition to that, it's been useless and motionless for so long that it's hard to think of its playing a vital part in the game.
Openings and endings
1 year ago