Fresh from his defeat in the prior round by Pillsbury in a Vienna Opening, Steinitz played the same line again. But the opening is not why Steinitz lost this game.

Mason's 6...exf4 was an improvement on Pillsbury's 6...Bd6, but Mason made no serious effort to press for a win. Steinitz' 23. Bxc7 was precipitous (he should have played 23. Re2 or 23. Re8), but Mason--strangely--did not simply play 23...Qxc2. Instead, starting with 23...Qf5, he began offering Steinitz the chance to trade Queens. Steinitz could have traded down to an even endgame on move 24, but kept trying to keep Queens on the board.

Steinitz finally capitulated to the trade with 31. Rf4, and an even Bishop versus Knight endgame was reached. Mason had a tiny edge, but Steinitz should have drawn with ease.

But Steinitz was not interested in a draw, and with 37...Bc5 (37. Kg3 was best) allowed Mason to get a passed pawn on the Queenside.

In this kind of ending, the Knight is superior to the Bishop, and Steinitz should have exercised great caution. But he played recklessly, and soon paid the price.

The Tournament Book says that Steinitz should have played 39. Kf4 instead of his actual 39. Kf2. But this comment is nonsense. After 39. Kf4, 39...Kd5 would have been crushing.

Steinitz' losing move was 41. Ke2. He still had a likely draw here with 41. Bxg7. But after 41. Ke2, Mason's 41...Nxg4 was decisive.

From this point on, Steinitz' desperate efforts to stem the tide only made matters worse. After Mason's 42...b4, Steinitz' last possible chance lay in 43. Bf8. His 43. g3 was hopeless, and now Mason's Queen-side pawns marched to victory. Once Mason got in 44...Ne3, the game was over. It was perhaps only here that Steinitz realized that 45. Kd2 would lose to the 45...Nf1+ Knight fork. Steinitz' actual move, 45. Be7, was even worse, and Mason's pawns were now unstoppable. My one regret is that Mason did not (slightly) shorten proceedings with the unusual move 48...Na1. But his 48...a4 was quite adequate.