Author: Kolss
Date: 14:47:36 07/26/02
Go up one level in this thread
On July 26, 2002 at 16:55:27, Slater Wold wrote: >On July 26, 2002 at 14:53:33, Dieter Buerssner wrote: > >>On July 26, 2002 at 00:54:10, Slater Wold wrote: >> >>>8/pR4pk/1b2p3/2p3p1/N1p5/7P/PP1r2P1/6K1 b - - 0 31 >>> >>>The only move to make here is Rxb2. The win is only about 10 moves away, and >>>while some programs I tested find Rxb2, none have an impressive score. And >>>that's *if* they find it. >> >>What would be an impressive score? >> >>>8/pR4pk/1b5p/2p5/N1p2P2/6P1/PP1r2P1/6K1 b - - 0 29 >>> >>>Same position. Only slightly deeper. No program I tested can find Rxb2. >> >>I think, programs that have a rather high value for the Q will find it in >>reasonable time. Yace on a rather slow computer needs about 1 minute. Yace uses >>10 for the Q. I did not test, but I guess with a value of 9, it would be much >>harder. I get a score of about one half pawn advantage for black, which seems >>stable. Is this won for black? I cannot see it ... >> >>Regards, >>Dieter > >Both postions are confirmed wins for black after Rxb2. > >One is deep and one is deeper. In fact, the first one seems more difficult - at least for Ikarus. It solves the first position one ply later, after 134 seconds, as compared to 41 seconds for the second one (both K6/2-475MHz, 8MB Hash). I suppose that this is because Ikarus thinks black is better anyway in the first position, while it slightly favors white in the second one and therefore switches to Rxb2 rather quickly. I suppose Dieter is right with his hypothesis about queen values. Regards, Munjong.
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.