In his desire to build pressure on the g6 pawn he saves a tempo by first playing Rh4! setting up for the possible sac after 36. hxg6 ... fxg6 37. Rxh6+ ... Kxh6 38. Qh4# (if Black did not free up the g8 square as he was forced to do). This allowed him just enough pressure and tempo to finalize his kingside assault with accurate play afterwards.

White's victory is amazing, can we say triumph of the will? Because it really is - Black takes every defensive advantage he can and if White didn't have this relentless will to win he would most likely not have ... after all, where did Black go wrong? The dark square weaknesses are all but unexploitable as the Knights and dark B's were exchanged early; Black's Bishop may be very bad but it is still good enough for defense up until the VERY end; Black gets in his ... c5, and he hits on the best strategy for switching blockaders against White's d-pawn.

Indeed the fact that g6 is the fatal weakness can hardly be deduced - the white squares seem too strong for Black to be conquered there! Even the strongest stone yields to the hammer in the end, though, and so it goes for Alapin. Such is usually the case in games like this where the defender has ample defensive resources and a lack of notable weaknesses, but no real choice but to defend; Capablanca's famous loss to Lasker in 1914 is another example of the insufficiency of passive defense. The attacker simply has too much latitude to probe for another assailable point (indeed, g6 turns out to only be weak inasmuch as it is a spot that force can be mustered against - per se it is really more strong than weak - but that is enough) and overload the defenders (Black's Bishop fails because it is an overloaded piece and not because it is a bad Bishop).

To this end, Black lost when he lost his chances for counterplay ... the game from move 22 through the finish is virtually a lingering death that Black fights like a lion but eventually must succumb to. Black's last chance for a tenable game I think was 21 ... e5! (c5 can wait another move, as it would take White at least 2 moves to prevent it due to the advance being supported and White having no guard at all of the square) Now Black has a very serious threat of 22 ... Kh8!! after which the advantage would even pass to Black, because b1-h7 is permanently blocked by White's e-pawn (thus making a K-side attack difficult at best), and if White were to transition to an endgame he would even lose (as d6 would be untenable, and Black has a healthy Q-side majority).

Thus, after 21 ... e5 White must knock out this pawn before Black can play Kh8 (as then the support by f6 is inevitable), and is virtually forced to play 22 f4. After 22 f4 exf4 23 e5 Kh8 (not 23 ... c5? 24 e6! and wins) 24 Qe4 c5 25 Qxf4 c4 26 Bc2 White still has the better of it, but the center is more open and Black has more open lines for counterplay which should at least make things more difficult.