Monday, September 5, 2011

Torture from Grigoriev

64, 1936
White to play and win

Some positions in this study could have arisen in my game with Mack Smith.  The solution is not easy!


Chris Falter said...

Bill, this is a fascinating study; thanks for sharing it. The basic idea, per the Nalimov tables at that I just consulted, is to bring the king to the g file, then advance the lead f pawn, then bring the king back around to the queenside. If the black king follows, white will simply use his spare tempi to zugzwang and outflank. If the black king stays on the h file, however, the win is much more involved, and involves multiple triangulations.

Ultimately the white king must come around to the queenside, because he doesn't have enough files available to outflank on the kingside. But he must first make black commit to a close engagement of the pawns on the kingside.

Will I remember this in the next tournament game that it applies? I doubt it!

Bill Brock said...

Putting concrete variations into words is always dangerous, but this looks right to me, Chris!

Chris Falter said...

The complete analysis was too long, and with too many branches! But it starts 1. Kf4 Kf6 2. Kg4 Kg6 3. f4.

Then 3...f6 4. f5 Kf7 (...Kg7 5. Kf4 and white outflanks black on the queenside, and uses the rear pawn to ward off counterattacks) 5. Kf4 Ke7 6. Kf3 (zugzwang: if black moves to the 8th rank or to f7, white outflanks him on the queenside; if ...Kd6 7. Kg4 Ke5 8. f3 and white's king will be able to reach g6, winning).

Or 3...f5 4. Kf3 Kf6 5. Ke3 Ke6 6. Kd4 Kd6 7. f3 +- (zugzwang: the white king outflanks black's).

Their are other complications, like how to handle a black counterattack on the kingside after white wins the race to the queenside, but your blog isn't big enough to handle it. :)