Domdaniel: Instead of the risky 33.Nd6!? - when he will be reduced to a single advanced pawn, giving Black chances of counterplay - White could have won more easily with 33.Nxc5, when the obvious 33...b6 is met with the finesse 34.Ra1!. Then 34...bxc5 loses to 35.Rxa7+ Kf8 36.Rh8+ Rg8 37.Ra8+ and 38.Raxg8, winning a Rook.If Black declines the Knight offer, then 34...a5 35.Na4 Rxb3 36.Nxb6 will mop up Black's pawns.
However, as the game develops, it's difficult to find a plausible drawing mechanism for Black. Even with three queenside pawns, he never really gets a chance to do anything with them, as White is able to conjure up constant threats with Rook, Knight and pawn.
White's eventual victory is narrow, hinging on key moves such as 41.Kf3! (though 41.Rd8 Re8 42.Kf3 also does the trick) and 44.Ke4. The point is that, when Black has to give up his Rook for the Knight and pawn, White has a centralized King and an active Rook.
With its own King in support and the attacking King not close enough, one pawn is able to draw against a Rook. Yet here four pawns are not enough, as the White pieces restrain and eventually win them.
After 52.Kc3, any move apart from ...Ka5 allows mate. And after 52...Ka5 53.Rh1! the pawns are doomed. Nice play by Tartakower, though Thomas's rather stolid defence made it easier at a few key points.
Much earlier, Black could have reached a drawable position with 27...gxf6 28.exf6+ Kd7, when his weaknesses on h6 and f7 can be defended. But after 27...Ke6? Sir George had to lose a piece ... and, eventually, the game.
In the opening, 4...c5! is more active and dynamic than ...c6 -- though one could hardly expect the players to know that in 1923, before Richter and Veresov did the analysis that attached their names to this variation.
Kibitzing
