Author: Terry McCracken
Date: 07:02:36 04/02/02
Go up one level in this thread
On April 02, 2002 at 01:52:06, J. Wesley Cleveland wrote: >On April 01, 2002 at 19:31:26, Terry McCracken wrote: > >>On April 01, 2002 at 18:42:20, Miguel A. Ballicora wrote: >> >>>On April 01, 2002 at 17:56:12, Art Basham wrote: >>> >>>>Here white plays Bxg6..!! >>>>(taken from the Yazgac test suite...) >>>> >>>>[D]r1b5/p2k1r1p/3P2pP/1ppR4/2P2p2/2P5/P1B4P/4R1K1 w - - >>>> >>>>I do not know if a computor program can find this one or not..:-) >>> >>>Many can, but the problem is that it is not the only solution. >>>IIRC crafty chooses the prosaic cxb5 that also wins (with the idea of Ba4 and >>>b6+). The pressure is too much for black to take. >>> >>>Regards, >>>Miguel >> >>True but it's about finding the "Best Move" and _all_ programmes look at cxb5 >>first and is a good move, but not the best, the programme needs to see there is >>better, then choose it. That is what we humans "Try" to do!;) >> >>Terry >It is not clear that Bxg6 is better. The evaluation you gave in a different >message shows white up 1.36 pawns. Crafty thinks that cb5 is +2 at the same >depth and by ply 17 is up 3 pawns. It's clearly better, and if you give a machine more time to work on it, it will hit higher evals. than +2 pawns. I've solved this position about 20 years ago, and it was clear in my mind to snap the pawn on g6. It was clear, clean win. Bxg6 is best, give a strong programme both moves cxb5 and Bxg6 and find out if given ample time, which wins more quickly. As a human, I examined Bxg6 as I play combinations, and it was evident to me that if I send the pawn home in a few moves, I'll win quickly. Terry
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.