Endgame: Lasker - Tarrasch, 1914 St. Petersburg

Moving to the next game in Lasker's Moves that Matter, another one of Lasker's endgame escapes occurred in his game against archrival Tarrasch. The 1914 St. Petersburg tournament was played in two stages. Eleven players competed in a round robin preliminary, then the five top players met in a double round robin final. The scores of the two stages were combined to determine the overall winner.

In the diagrammed position, Lasker played 36.Bxg7. The game continued 36...Bxf5+ 37.Kf7, where Kasparov started his analysis. Here Tarrasch played 37...Bxg7, and after the further 38.Bxf5 Kxf5 39.Kxg7 a5 40.h4 Kg4 41.Kg6 (Kasparov: '!'; Soltis '!!'), Lasker managed to draw. Tarrasch had been expecting only 41.Kf6, which loses.

Tarrasch had a win with 37...Be6+ 38.Kf8 Bxg7+ 39.Kxg7 Bxb3, and then, for example, 40.h4 Bd1 41.Bg6 c4 42.h5 Bxh5 43.Bxh5 b4 44.Kf7 a5.

St. Petersburg (prelim) 1914
Tarrasch, Siegbert

Lasker, Emanuel
(After 35...Kd5-e5)
[FEN "5B2/6pB/p4bK1/1pp1kP2/4b3/1P6/1P5P/8 w - - 0 36"]

Soltis gives Lasker's 36.Bxg7 without comment. The move 36.Bxc5?? is a blunder because of 36...Bxf5+. Although Soltis says nothing about 36.Kf7, I can't find anything wrong with it. If 36...Bxf5, then 37.Bxf5 Kxf5 38.Bxg7 Bxg7 39.Kxg7 is the same position reached in the game. Avoiding the Bishop exchange with 38...Bd4 39.Bf8 a5 (39...c4 40.bxc4 bxc4 41.Ba3) doesn't win either: the sequence 40...a4 41.bxa4 bxa4 isn't a real threat. The alternative 36...Kf4 37.Bxg7 Bxg7 38.Kxg7 Bxf5 39.Bg8 doesn't look any stronger. What have I overlooked? Was Lasker's 36th move a blunder?

To play through the complete game see...

Emanuel Lasker vs Siegbert Tarrasch, St Petersburg preliminary 1914
http://www.chessgames.com/perl/chessgame?gid=1271260

...on Chessgames.com.