Cordingley puzzle 93

White to play and win

Solution

Not the hardest of puzzles, though (i) I took some time to solve it; and (ii) see below for what Houdini found.

The first thing I noticed was of course the pressure on the a1-h8 diagonal, with the Bb2 and Nd5 focussing on it, and the queen well placed to join in the attack. I also noticed that the Ra8 was LPDO, and for a while this caused a mis-direction in my thinking: trying to get a double attack by Qf3 on a8 and f6: to no avail.

So, back to f6, and since Ne6 is clearly bad-fe- the desperado 1 Nc6! is the move; then 1…dc 2 Nf6+ gf (2..Kh8 then 3 Qh3 or probably better 3 Qd3) and all I had to do was find the right way to attack f6, and once I saw 3 Qh6! the game is over. Or so I thought, and so Cordingley thought.

Houdini switched on, purrs away, and shows (after 3 Qh6) 3…Ba6 as best, – 3.8 and 3…d5 as +1.6. At first I thought ‘oh, Ba6 is the best defence’, almost moved on, and then saw that 3…Ba6 was minus 3.6: black wins. So I looked at if further, and indeed 3…Ba6 is the best defence, but Houdini’s evaluation changes deeper into the analysis with a few plausible moves: I think both 4 Rae1 and 4 f4 are sufficient for white to win, though 4 Rfe1 isn’t, or probably isn’t.

The attached analysis of 3…Ba6 isn’t exhaustive: to quote from Jon Ludwig Hammer, who yesterday beat Wang Hao in the Norway tournament, who in the post mortem news conference said ‘I was confident that my position was winning, but not that I would win it’. Quite so: if I were playing white, and black played 3…Ba6, there would be a fair chance I would make a mess of it.

Posted with Blogsy