The psychological cracks start to show with the clumsy 49 Bh8 when there was a forced win on the board.

I think I'd vote for Steinitz. Of the world champs I am sure he's be first in this category.

For great masters, Rubinstein and Chigorin would be up there, surely.

Of course, this is the kind of question a strong engine could answer definitively. <nimh>? <CharlesSullivan>? Any interest?

Issue 1: Why did Lasker blunder here?

Theory 1: He was already lost since he was about to lose a pawn and decided to try a "hazardous attack" to turn the game around. This is the Tournament Book's theory. But this theory is simply wrong. Lasker was NOT lost before 42...Rf1 check and did NOT have to lose a pawn. The Tournament Book has simply overlooked a Queen fork. Had Lasker played 42...dxe5 instead of his blunder, and had Mason tried to snatch a pawn with 43. Qx35 as the Tournament Book suggests he could have, Lasker would have won a Rook by 43...Rf1 check 44. RxR QxR check 45. Kc2 Qc1 check 46. Kb3 Qd1 check, winning White's Rook.

Theory 2: Lasker was in time trouble. as keypusher has correctly surmised, and as is confirmed on page xvi of the Tournament Book, the time limit was 15 moves per hour. Thus, Lasker did indeed have a time control coming up on Move 45. However, there is nothing in the accounts of the game to suggest time pressure.

Theory 3: Lasker was out of practice. True indeed, Lasker had not played competitive chess since January 1897 at the conclusion of his rematch with Steinitz. However, this was Round 5, and Lasker was about to win 18 games without a loss. While he had erred in the prior round in losing to Blackburne, his error was in overlooking a brilliant Rook sacrifice, not in tossing a piece away for nothing.

Theory 4: (My theory): Lasker's error is more subtle than we may think. Even after the loss of a piece as a result of his 42...Rf1 check, Lssker had a nasty surprise for Mason. Had Mason played the natural looking move of 47. Nf6, Lasker's sacrifice would have worked, since 47...b5 now wins for Black. Lasker may have looked ahead and expected 47. Nf6, and then won brilliantly. In fact, Mason avoided this pitfall with his careful Nc3 (Fritz avoids the entire problem by playing 47. RxB--but I doubt any human would do this). In short, Lasker's blunder was a failure to see that his brilliant winning idea five moves ahead was flawed.

Issue 2: Why did Mason blunder? The answer seems obvious from playing over the game. Lasker--though lost--presented Mason with a web of problems to solve. As noted above, Mason had to avoid the vicious surprise Lasker had in store for him had he played 47. Nf6. Even after seeing through that trap, Mason was confronted with another tricky Lssker combination on Move 60. Had he played 60. Kb3 instead of 60. Rb2, Lasker had an immediate forced draw with 60...Qb1 check. Thus, Mason--playing the World Champion--was confronted with one problem after another, and succumbed to the pressure. I know the feeling. I once was fortunate to obtain a better endgame against a famous Grandmaster. I should at least have drawn, and probably should have won, but faced with repeated tricks and traps, I eventually made my fatal mistake. Winning a won game against a great player is really hard.