Cordingley puzzle 45
White to play and win
Solution
The first move, 1 Bg7, and black's reply, 1…Rd1 ( 1…Qg7? 2 Rd8+ is trivial), are obvious. White's second move is clearly a discovered check, but where to put the bishop? Ignoring (though checking) Bf8 and Bh8, Bh6+ is a draw by repetition after 2…Kh8 (not 2..Kf7?? because if 3Qg7+, and the LPDO Qb7 drops off; 2 Bf6+ blocks the f file, and since there is nothing clear after 2…Bf8, my attention turned to the three remaining moves, Be5, d4 and c3. In my initial appraisal, I also noted that the Rd1 is LPDO and that …e3 is a double attack, f2 and h1mate, which needs to be guarded against.
So 2 Bc3/d4/e5+, Kf8 (Kf7 the LPDO drops after being skewered by Qg-7+) 3 Rg4 and then what? Clearly white has compensation, but is it enough, say after 3 Rd7, protecting both LPDOs. Hence I landed on my choice of move 2 Bd4, liking the fact that it both hits c5 and blocks the d1 rook. I couldn't see the position through to the end, but my bones told me it was likely to be sufficient, so I plumped for it. Pleasingly, it is Houdini's preference too.
The move white played 2 Be5+, also wins, but maybe with more effort. I knew, and it proved to be the case, that Houdini would savage both my play, Cordingley's solution, and the game continuation. There are some pretty computer lines in the attached PDF. Houdini also told me that the move I least liked, 2 Bc3, is only =; the reason for my judgement was not deep analysis, but the fact that on c3, it doesn't do as much as on the other two squares; and Bb4+ is never possible, whilst the black queen guards the square. I should also add that I preferred 2 Bd4 to 2 Be5 partly because I couldn't analyse the latter sufficiently.
Finally, in a real game, as compared with leisurely home analysis, would I have dared play 2 Bd4, or would I have chickened out with 2 Bh6? Frankly, I don't know what I would have done: chickened out, probably. Certainly I wouldn't have been able to see to the end, so I might have been scared of sacrificing, in case it lost: a weakness of my playing approach.



I considered Bxg7 but after Rxd1 I just couldn’t calculate it through. Taking advantage of the pin on f file is hard to see.