Author: John Merlino
Date: 20:49:41 01/25/04
Go up one level in this thread
On January 25, 2004 at 18:15:03, Mike Hood wrote:
>This isn't a constructed position. It cropped up in a small engine tournament
>that I was running today. White was Fritz 5.32, Black was Junior 7. I have all
>the 5-piece EGTBs on my PC, but Fritz doesn't use them.
>
>[D] 8/8/7Q/8/3K2p1/q5p1/6kP/8 w - - 0 75
>
>It looks bad for White, although if I had this position on the board I might try
>for a draw by playing Qd2+. In the tournament Fritz 5.32 was playing White, and
>played hxg3, which it evaluated at -1.87. A typical blunder for a program that
>doesn't use EGTB probing. Junior 7 immediately recognized that the resulting
>position was a Mate in 65.
>
>Other non-probing engines make the same mistake. List 5.04 plays hxg3,
>evaluating it at -2.80. Goliath Light also plays hxg3 (-2.00). Are there any
>non-probing engines that avoid this move?
1.Qd2 might seem to be a good way to fight for the draw, but it quickly loses to
1...Kh1 (or Kg1), followed by:
2.Qe1+ (or 2.Qd1+) Kxh2 3.{any move} g2
or
2.hxg3 Qd6+! 3.Ke3 Qxd2
So, in fact, I'm pretty sure that 1.hxg3 is the best move, as it might lose the
slowest! And, not only that, if the program playing Black also does not have
EGTBs, then it might play 1...Qxg3 which is a tablebase draw! After 1.hxg3,
Black must play 1...Kxg3 to get a forced mate.
jm
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.