Another puzzle for an easy Reitstein day: Vugar Gashimov victory

White to play and win

(finding the first move is not enough: need to go deep enough to prove a convincing win, or at least an overwhelming position)

V Gashimov v B Gelfand, Spanish team ch, 2009

Solution

1 Ng7! is an obvious try, but does it work? 1…Kg7 2 Qh5! (2 Bh6+? throws too much wood away) Rh8 (2…f5, see below) 3 Bh6+ Kg8 is the 'key' position.

White has no good checks, since g4 and g5 are covered; by a process of elimination I found 4 e6! Be6[] 5 Qg5! 1-0, since both Qg7mate and Qg3+ are threatened, and both can't be prevented. Very nice.

Black has other defences. Firstly, 1…c4, when 2 Nf5! wins, since if 2…cd?? 3 Qg4+ 1-0; so white emerges a pawn up, with black's king side shattered. Or, 2…f5, 3 ef+ Kf6[] which is harder. White's natural (and Stockfish says best) move is 4 Bf4, when 4…Nc6 is again obvious:

I know in my practice I could easily mess up such a position as white: one or maybe two inaccurate moves, and bang!, the advantage goes up in smoke. Stockfish tells me it is mate in six, so the diagram is worth repeating as a puzzle:

White to play and mate in six

In practice, there is no way I would see, or even look, for a mate here: however, I might stumble on it, by the (non GM) technique of 'keep checking and see what happens'. So, here, check and take the h7 pawn off, there try freeing the g6 square up. 5 Qh6+ Kf7[] 6 Qh7+

Black has three king moves: 6…Kf6?? 7 Qg6 mate; 6…Ke8 7 Bg6+ Kd7 8 Qh3+ and mates; so 6…Ke6 7 Qh3+! and now the win becomes clear: the queen edges closer, so, 7..Kf7 (say) 8 Qh5+ Ke6 9 Qg4+ and Qg6mate, ignoring desperado Rf5 moves.

The game came to my attention through Malcolm Pein's Daily Telegraph column, which I religiously read. Most times, I can solve his puzzles by inspection or with only a bit of effort (he writes for a general audience of chess players) but on occasion, as here, I have to cut out the article, sans solution, and put time aside to solve it.

Posted with Blogsy