Computer Chess Club Archives


Search

Terms

Messages

Subject: LGGold good again

Author: Paul Richards

Date: 16:50:22 06/14/99

Go up one level in this thread


On June 14, 1999 at 16:39:38, Dann Corbit wrote:

>On June 14, 1999 at 15:38:14, Dann Corbit wrote:
>[snip]
>>Even after this position, the long term advantage is not clear:
>Oops, pasted from the wrong file:
>r1b2rk1/3pqpp1/p1nNp3/3nP2Q/5P2/1N6/1PPB2PP/5RK1 w - - acd 13; acn 100014453;
>acs 717; ce 26; pv Rf3 Nf6 exf6 Qxf6 Bc3 Qg6 Rg3 Qxh5 Rxg7+ Kh8 Rg5+ f6 Rxh5+
>Kg7 Kf2 Kg6;
>
>Anyway, chess programs don't see sacrifices at all.  The only thing they
>*will* see is a very deep combination.
>
>As far as extensions are concerned, I don't think you will find any
>program that solves a position like the one you propose except perhaps
>a mate solver, if there is a mate nearby.

This position is not that difficult, though there is no forced mate.
I tried three engines on a PII 400.  Fritz 5.32 finds the winning line
at 13 ply plus 30 or 40 selective.  Takes about 8 minutes.  Crafty
chooses the right move at 14 ply, though it only gives it 12 centipawns
initially.  Takes about 20 minutes, and I presume it will stay with that
and increase its evaluation.  LGGold 2.0 takes the cake again, finding
the winning move in 65 seconds, and displaying the winning line and
a strong eval about 15 seconds later.  It seems quite good at searching
tactical positions.



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.