Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: A Topprogram needs hours to find 30.a5!

Author: Thomas Lagershausen

Date: 13:27:38 01/15/01

Go up one level in this thread


On January 15, 2001 at 07:05:54, Mogens Larsen wrote:

>On January 15, 2001 at 06:22:35, Thomas Lagershausen wrote:
>
>>>[D]1r2k1r1/3p1p2/2bR1Bp1/p1K1P2p/5P1P/1P6/3R4/8 w - - 0 32
>>
>>It is now clear that two moves earlier the move 30.a5 is the winner.
>>Black has no good alternative defense.
>>For example 31...Ra6 32.Ra2 Ra7 33.b4 a4 34.b5 Ra5 35.Rxc6 dxc6 36.Rd2 Ra8
>>37.bxc6 and winning. Or 31...Rc8 and 32.Kb6! wins clear.
>>So the real question is:"Wins a program with the white pieces after the 29th
>>move by black? We know 30.a5! is a winner.
>
>I gave Gandalf a chance on my Athlon 800 with 56Mb hash.
>
>Analysis by Gandalf 4.32g:
>
>30.Kc3 b5
>  =  (0.09)   Depth: 2   00:00:00
>30.Kc3 b5 31.axb5 Bxb5
>  =  (-0.01)   Depth: 3   00:00:00
>30.Rd1
>  =  (0.01)   Depth: 3   00:00:00
>30.Rd1
>  =  (0.01)   Depth: 4   00:00:00
>30.Rd1
>  =  (0.01)   Depth: 5   00:00:00  2kN
>30.Kc3
>  =  (0.03)   Depth: 5   00:00:00  2kN
>30.Kc3 b5 31.axb5 Bxb5 32.R2d5 Rc8
>  =  (0.11)   Depth: 5   00:00:00  2kN
>30.Kc3 Rb8 31.Kc2 Be4+ 32.Kb2 Rb7 33.Rd1
>  =  (0.11)   Depth: 6   00:00:00  25kN
>30.Kc3 Rf8 31.Kb2 Rg8 32.Kc1 b5 33.axb5 Bxb5
>  =  (0.08)   Depth: 7   00:00:00  79kN
>30.Kc3 Rf8 31.Kd3 Rg8 32.Kc4 Rc8 33.Rd1 Ra8
>  =  (0.16)   Depth: 8   00:00:01  232kN
>30.Kc3 Rf8 31.Kd3 Rg8 32.Kc4 Rb8 33.Kc3 Ra8 34.Kc4
>  =  (0.16)   Depth: 9   00:00:01  323kN
>30.Kc3 Rf8 31.Kc2 Rg8 32.Kb2 Rf8 33.Kc3 b5 34.axb5
>  =  (0.14)   Depth: 10   00:00:04  874kN
>30.Kc3 Rf8 31.Kc2 Rg8 32.Kb2 Rf8 33.Kc3 Rg8 34.Kc4
>  =  (0.14)   Depth: 11   00:00:06  1362kN
>30.Kc3 Rf8 31.Kc2 Rg8 32.Kb2 Rf8 33.Kc3 Rg8 34.Kc4
>  =  (0.14)   Depth: 12   00:00:08  2016kN
>30.Kc3 Rf8 31.Kc2 Rg8 32.Kb2 Rf8 33.Kc3 Rg8 34.Kc4 Rc8
>  =  (0.14)   Depth: 13   00:00:15  3621kN
>30.Kc3 Rf8 31.Rc2 Rg8 32.Kd3 Ra5 33.Ke3 Rf8 34.Rcd2 Ra8
>  =  (0.12)   Depth: 14   00:01:03  14559kN
>30.a5
>  =  (0.14)   Depth: 14   00:02:02  14559kN
>30.a5 bxa5 31.Kxc5 Ra6 32.Ra2 Ra7 33.b4 a4 34.Rad2 Be4
>  =  (0.24)   Depth: 14   00:02:21  14559kN
>30.a5 bxa5 31.Kxc5 Rb8 32.Rxc6 dxc6 33.Kxc6 Kf8 34.Kc7 Ra8
>  ²  (0.46)   Depth: 15   00:07:16  87829kN
>30.a5 bxa5 31.Kxc5 Ra6 32.Ra2 Ra8 33.Kc4 Be4 34.b4 a4
>  ²  (0.57)   Depth: 16   00:13:23  163041kN
>
>Regards,
>Mogens

Wow.Maybe Gandalf beets every other program in this position?

I had only time to testing Rebel Century 3.
Rebel needs 4 hours 28 minutes and 29 sec. to see 30.a5!
CPU: K6-2+ 600 Mhz ply 16 Variation 9 score only +0,24
Rebel shows only 30.a5 ba5: 31.Ra2?! Bb7
I think now we have a difficult position for computers.



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.