Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: WAC #252 - Bg4

Author: Steffen Jakob

Date: 02:42:42 01/04/02

Go up one level in this thread


Hi Uri!

On January 04, 2002 at 05:19:01, Uri Blass wrote:

>On January 04, 2002 at 03:41:34, Steffen Jakob wrote:
>
>>An another one: WAC #252:
>>
>>[D]1rb1r1k1/p1p2ppp/5n2/2pP4/5P2/2QB4/qNP3PP/2KRB2R b - -
>>
>>Here the key move is 1... Re2 which is surely the best move if you define "best
>>move" as the move which wins and leads to the shortest mate (it's a mate in 5).
>>Hossa (version of today) needs 19.73 seconds to play 1... Re2. Before he prefers
>>1... Bg4 which also seems to win. So if you only look for winning move then this
>>position should be counted as solved in short depths, too. How do others handle
>>such cases?
>
>My program needs 8.282 seconds on p800
>to play Re2
>
>I am surprised that my program find Re2 faster than your program
>because I did almost nothing about improving the search rules
>and my search rules for today are similiar to tscp search rules

That's probably your advantage here :-) I see Re2 in ply 7 and the mate in ply
10. I am almost never surprised if any program finds something faster than my
engine :-(

>My program likes Bg4 at depth 5 but at depth 6 changes it's mind
>twice when the first time is changing it's mind from Bg4 to
>Nxd5 when the score goes up from 1.1 to 1.3 and
>the second time it changes it's mind from Nxd5 to Re2
>Score goes up from 1.3 to 4.86 when depth 7
>is enough to see mate in 5.

My score for Bg4 (first time at ply 5) is about +3.7.

Greetings,
Steffen.



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.