Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How long to find White's winning move?

Author: Will Singleton

Date: 20:18:00 01/03/02

Go up one level in this thread


On January 03, 2002 at 15:20:41, Peter Rosendahl wrote:

>On January 03, 2002 at 14:45:20, Roy Eassa wrote:
>
>>On January 03, 2002 at 14:33:00, Peter Rosendahl wrote:
>>
>>>I have a question about this position: is it really clear that b4
>>>is the (only) best move? Motivated by my stupid engine,
>>>which couldn't find b4 but stayed with Rg5, I tried first Rg5 and
>>>after that b4 with Fritz 6 on the same position and got the
>>>scores +4.00 / +4.09 after about 10 minutes (Athlon 900MHz).
>>
>>
>>I think you're right, Rg5 is probably also a win.
>
>Thank you for the quick reply. After half an hour of analysis with Fritz 6
>the score for Rg4 drop a little to 3.59, but it seems both moves are wins.
>
>>
>>I should have asked how long until a program selects b4 with a winning score,
>>but I didn't want to give the first move away to anybody eyeballing the position
>>(because b4 is such an odd-looking move and is very difficult for must humans to
>>find).
>
>The position is very interesting to analyze (and my question wasn't meant as
>a critique). Thanks for sharing!

My program also sticks with Rg5.  I'm running it on my laptop, so the depths
aren't great, but I did your experiment (above), and found that the score after
b4 was around -3.5, and after Rg5 was only around -1.5 (for 10-11 ply).  Based
on that, my prog should have selected b4.  So, I think maybe we have a similar
bug.  I'll investigate hash, null-move, etc.

Will




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.