Defending KQ v. KQP

After a long battle, we reached this position in my game last night:

6Q1/8/7K/7P/8/8/6k1/5q2
Black to move

Playing Black, I wasn’t sure whether this was a theoretical draw or a win for White, but I thought I had a reasonable chance to draw. (I looked it up this morning using this endgame tablebase web site, and it is a theoretical draw.)

I hadn’t studied this type of endgame before, but fortunately I had a few minutes on the clock to think about what to do. Obviously, I can’t let White force a queen exchange. If White is checking me, then he can force a queen exchange most typically via a fork, a skewer, or a pin. I decided that when White was checking, I would strive to keep my king off the same file, rank, diagonal, and color as my queen. In the position above, for example, I played 53…Kh2.

When Black is checking White, the primary danger is allowing White to get out of check in a way that forks Black’s king and queen. For example, after 57…Qd8+ 58.Qf6:

3q4/8/5Q2/6KP/8/8/7k/8
Black to move

58…Qd2+ would be a disaster because of 59.Qf4+ Qxf4 60.Kxf4 1-0.

With these ideas in mind, I managed to play 10 moves without error and without allowing the pawn to advance. It was getting close to midnight, though, and I wasn’t at all sure that I could play the 40 more error-free moves necessary to get to a draw.

You just never know what’s going to happen in a chess game, though. My opponent, who must have been getting pretty tired as well, walked his king into getting checkmated at move 62.

This entry was posted in Chess. Bookmark the permalink. Both comments and trackbacks are currently closed.

2 Comments

  1. Posted March 22, 2007 at 3:38 pm | Permalink

    For a couple of patzers I thought it was a very interesting game. It had some opening drama, complex positional themes in the middle game and a 10 move calculation that resulted in the first diagram position. I thought it might be drawish but decided to test the waters. After 3 hours of careful play it only took 1 bad move to lose. I wondered during the game whether the KQ v. KQp endgame would’ve been ‘winnable’ if the pawn was on g-file instead of h.

  2. Steve
    Posted March 22, 2007 at 4:03 pm | Permalink

    That’s a good question, Fred. I would have guessed that if you moved the position over one file, it would be a win for White. The endgame tablebase I consulted, however, says it’s still a theoretical draw.