Sunday Chess Problem

Well, the final snow tally in my neck of the woods was a little over two feet. Impressive! On the other hand, there doesn't seem to have been much damage. There were no strong winds, we never lost power, and all of my trees seem to be intact. And it means I get another snow day tomorrow! Cleaning up after two feet of snow takes some time. I think my students might forget who I am.

In other news, it seems that the Grand Mufti in Saudi Arabia is not fond of chess.

A video clip of Saudi Arabia's top cleric saying that the game of chess is “forbidden” in Islam because it wastes time and leads to rivalry and enmity among people has provoked heated debate, and widespread criticism, among Arabic Twitter users.

The clip was shared on YouTube in December, gaining traction in recent days on social media. Some Twitter users mocked Grand Mufti Sheikh Abdelaziz Al Sheikh, saying chess is an intelligent game and that is why conservative clerics decry it. Others defended his religious advice, saying that many other Islamic scholars have also warned that the game can be addictive and cause people to lose focus from their daily prayers and remembrance of God.

What a jerk! Sunday Chess Problem isn't just fun, now it's fighting jihad as well.

I have a really fun problem for you this week. It was composed by Noam Elkies--a mathematician!--in 1994. It's an endgame study, with the stipulation that white is to play and draw.



Let's first notice that white is completely tied up in knots in the lower left corner. He does have the move Nc2, however. This obviously loses, since black will just take the knight with his pawn, but it does mean that white can't count on any stalemate defenses based on the fact that he has an f-pawn.

So white only has one chance. He has to try to promote the f-pawn. After the obvious beginning 1. f6 Qb3, white already faces a choice.

The basic plans are clear. Black is going to liberate his queen from the lower left corner by giving check on d1, and then emerging onto the kingside. Since the black queen is in time to stop the white f-pawn, white will need to move his king closer to the pawn to support it. That means that on the next two moves, he is going to move his pawn to f7 and his king to h6 (or maybe g6) to support it.

But what's the point? It seems like white is just losing. By using standard queen maneuvers known to every young chessplayer, black will force the white king to move in front of his own pawn. Black will take that opportunity to move his king up to a4. Then the white king will emerge from in front of the pawn. But black will just repeat his queen maneuvers, force the white king back in front of his own pawn, and then take the opportunity to move his king again, this time to b3. After repeating this maneuver a bunch more times black will march his king all the way up the kingside. The combined force of the black queen and king will be more than enough to take the white pawn and that will be that. It's all very tedious, but the end does not seem to be in doubt.

So what does it matter whether white plays 2. f7 and then 3. Kh6, or instead 2. Kh6 followed by 3. f7?

Well, it does matter. One move loses and the other move draws. The correct sequence is 2. f7! Qd1+ 3. Kh6.



Now play will unfold something like this: 3. ... Qf3 4. Kg7 Qg4+ 5. Kh8 Qf5 6. Kg7 Qg5+ 7. Kh7 Qf6 8. Kg8 Qg6+ 9. Kf8:



Now that the white king has been forced in front of his pawn, black has time to bring his king up with 9. ... Ka4. But now white will move his king out of the way of the pawn with 10. Ke7. Now white is threatening to promote, so black repeats his queen maneuvers to force the white king back to f8. 10. ... Qg7 11. Ke8 Qe5+ 12. Kd7 Qf6 13. Ke8 Qe6+ 14. Kf8 and black has time for another king move with 14. ... Kb3.



I think it's time to hit fast-forward. After two more repetitions, black gets to play the move 24. ... Kc1.



After three more repetitions, black plays 39. ... Kf3.



Two more repetitions brings us to 49. ... Ke5.



Now, finally we're ready for the point of all this. Play will now continue 50. Ke7 Qg7. I promise that 50. ... Qe6+ would make no difference. Play continues 51. Ke8 Ke6.



Now what? White can promote his pawn, but it will do him no good. Promoting to a queen means getting mated instantly on d7, while giving check by promoting to a knight only postpones mate for few moves. So what can white do? Has it all been for naught?

Not at all! White plays 52. Kd8! and immediately claims a draw. If you have been keeping track, there has not been a capture or a pawn move since white played 2. f7. That means fifty moves have been played without a capture or a pawn move. And that's a draw by the appropriately-named but seldom used fifty move rule.

Of course, if white had played 2. Kh6 followed by 3. f7, it would only have been forty-nine moves, and black would win after all.

Jeepers! That's just brilliant. Maybe FIDE needs to make it the fifty-one move rule just to eliminate this possibility.

Incidentally, I let Houdini analyze this position for several minutes. It claims a small advantage for black, but fifty moves is a bit outside its horizon. But when I laboriously played through the entire solution, it instantly asserted the fifty-move rule at the appropriate moment.

See you next week!

More like this

How funny! I just kept thinking " but I don't see the stalemate" at the next - to- last & last diagrams. Is this one a first? I mean, I'm no afficianado, but this looks diabollically original to me. Nice pick, Jason!

A funny thought: in a tournament game at standard time controls, if I were White I'd have been counting the moves starting around move 35 or so for sure. Alternate universe.

By Bill McNeal (not verified) on 24 Jan 2016 #permalink

52 Nc2 is a legal last move as well, is it not? Also with the 50 move rule?

Of course, it's an endgame study rather than a mate-in-X, so that's fine.

At move 25, my engine suggests ... Qe5+ rather than Qe7. That doesn't seem to make any significant difference, except that at move 47 instead of ... Kf5 or ... Ke5 my engine says that ... Qh7 is a mate in 20. The idea is that this forces 48 Nc2:
- 48 Ke7 Ke5 49 Ke8 Kd6 and now White must either play Nc2 or f8, as either Kd8 or Kf8 are mate after ... Qh8. Nc2 leads to a capture, f8 is a pawn move, so the draw is gone.
- 48 Ke8 Ke5 49 Ke7 Kf5 50 Ke8 Kf6 is again just in time, as the f7 pawn is lost if it doesn't move, while 51 f8(Q)+ Ke6 and White has no better than to give up his queen, since ... Qd7# is threatened.

That said, I'm sure I have missed something.