Last week’s puzzle was taken from Aussie GM David Smerdon’s terrific new book The Complete Chess Swindler. My review will be published on British Chess News shortly.

This is Shirov-Kramnik (Groningen 1993). White’s attack has misfired and he didn’t fancy 18. Qh4 Nxg3 19. Bxe7 Nxf1 20. Bxd8 Qxe5 21. Bf1 Qe3+ 22. Kb1 Bc6 when Black stands better.

Instead he chose the queen sacrifice Bxh6, the move I asked you to evaluate last week.

The game concluded 18… Nxf4 19. Bxg7+ Kh7 20. Rxf4 Rg8 21. Rfg4 Rxg7 22. Rxg7+ Kh6 23. Rg8 Kh7 24. R8g7+ with a draw, but Kramnik had missed something.

He could have returned the queen with either 20… Qxc3 or 21… Qxc3, when taking with the pawn allows Ba3+, mating, while taking with the rook would have left him a piece ahead.

So the correct answer to my question is that Bxh6, even though it worked out over the board, should lose with best play. Whether or not it was the best practical chance is another matter. At what point, if at all, did Shirov see Qxc3?

In this position White might consider simplifying with Qh4 or sacrificing with Bxh6.

Which would you choose? Specifically, does Bxh6 win, draw or lose with best play?

Over to you.

Last week I left you with this endgame study by Richard Guy.

It could also, I suppose, have been presented as a mate in 10.

It solves like this:

1. Nf2+ Kg1 2. Nh3+ Kh1 3. Ka6 Ne1 4. Be4+ Ng2 5. Kb5 (Ka5 takes a move longer) 5… a5 6. Kc4 a4 7. Kd3 a3 8. Ke2 a2 9. Kf2 a1=Q 10. Bxg2#

Congratulations if you solved it.

This week, another endgame study composed by the distinguished mathematician Richard Guy, who died earlier this year at the remarkable age of 103. It was first published in Le Problème in 1946.

White to play and win.

Last week’s puzzle was taken from In The Zone, a new book by the prolific Cyrus Lakdawala: you can read my review here.

This is taken from a game between Edmar Mednis and Bobby Fischer, played in the first round of the 1963-64 US Championship.

Mednis has a strong passed pawn and could have won here by playing 24. Rd5 Rxe7 25. dxe7 Nc6 26. Nd6 Nxe7 27. Nxc8, with two rooks and a knight against Fischer’s queen. Instead he traded queens with 24. Qg5, exactly the sort of pathetic move I would have played. Bobby eventually rounded up the d-pawn and won the ending – followed by his next ten games, finishing the tournament with an unprecedented 100% score.

If he’d lost this game, as he should have done, chess history might have been very different.

Many thanks to all who analysed and commented on Facebook – including two IMs!

A position which changed chess history for you this week.

White to play: what would you suggest?

Do let me know if you have any bright ideas (or if you recognize the position).

Last week I left you with this position from the Four Knights Game.

Black has several reasonable moves here: 8… Be7 is most natural and the engine choice, but a6, for example, would also be fine.

But 8… d6, which looks plausible, would be a fatal error after 9. Nc6 bxc6 10. Qxc6+ Bd7 11. Nxc7+ Ke7 12. Re1+ Be6 13. Nxa8

Very easy to miss over the board.

This was a position from the lichess.org tactics trainer: in an online game Black played d6 and White missed the refutation.

Not so hard to solve, though, if you know there’s something there.

An opening position for you this week.

You’re Black, to move, in a Four Knights Opening. Which moves would you consider? Which moves would you reject, and why?

You tell me: I’ll tell you next week.

Last week I left you with this Mate in 4 composed by Fritz Giegold.

Giegold’s problems offer a very attrative introduction to the world of chess composition. If you see his name above the diagram you should always start by looking at the most improbable move.

This one solves by 1. Qg2! (threat Qxf3#) 1… fxg2 2. Rg3! (threat Rc3#) and, however Black captures, White has 3. Ng8 and 4. Ne7#.

Well done if you managed to solve this without computer assistance.

Time for a problem. White to play and mate in 4 moves (Fritz Giegold, Bayerische Problemmeisterschaft 1962, 1st Prize).

If you think you’ve solved it, do let me know.