generror: Despite losing twice, Steinitz again choose the same opening lines, but he deviates here with <9...Be6> -- definitively an improvement over his earlier <9...Ne5?!>. Both players then wait for the other to make a mistake. Steinitz finally gives away his a-pawn with <22...Rad8?>, but he cunningly creates complications by offering his d-pawn instead with <23...d5> ("the kind of move which an experienced player makes when he has no good defence at his disposal").
Lasker takes it, and Black can't of course retake because <24...cxd5? 25.Bxa4> wins the exchange; but now White can't take the a-pawn because after <25.Bxa4? Qa6>, Black can drive the rook from the 4th rank and then would win an exchange, and <25.Rxa4??> also blunders the win because of <25...Bxf2! 26.Kxf2 (Rxf2?? Re1+ 27.Rf1 Rxf1#) Re2+ 27.Kg1 Rxc2>.
However, Lasker would have gotten very good winning chances with <26.Qf5 Re7 27.h4!>, but for some reason that only he knows, he chooses to trade his nice bishop pair for a knight.
But now Black has the initiative, and could even have won with <30...Bxf2!!>, as <Honza> pointed out above. White could have equalized with <33.Re4!>, but after <33.Bc1?>, Black could have gotten a solid advantage with <33.Rde6>; after <33.Rxf2?>, the game results in a very drawn endgame, where even Stockfish doesn't see any hidden subtlteties.
Especially compared to the previous games of this match, this is a relatively uneventful game; but I guess even Steinitz and Lasker needed to rest their nerves a little.
Kibitzing
