Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Nullmove implementation

Author: Uri Blass

Date: 09:51:27 05/20/02

Go up one level in this thread


On May 20, 2002 at 10:23:13, Benny Antonsson wrote:

>Alarm analyzis on a P3/533 MHz with 32 MB hash.
>
>[D]r1b1r1k1/p1p3pp/2p2n2/2bp4/5P2/3BBQPq/PPPK3P/R4N1R b - - 0 1
>
>With nullmoves:
>1       92      0       127      Bg4
>2       109     2       811      Bg4 Qf2
>3       99      7       4680     Bg4 Qf2 Bb6
>4       66      35      25457    Bg4 Qf2 Bb6 Kc1
>5       110     66      44565    Bg4 Qf2 Bb6 Kc1 Ne4 Bxb6 cxb6 Bxe4 Rxe4
>6       110     249     278029   Bg4 Qf2 Bb6 Kc1 Ne4 Bxb6 cxb6 Bxe4 Rxe4
>7       108     921     1237281  Bg4 Qf2 Bb6 Bxb6 cxb6 Ne3 c5 Nxg4 Nxg4
>8       97      3387    4760024  Bg4 Qf2 Bb6 Bxb6 cxb6 Ne3 c5 c3
>
>Without nullmoves:
>1       92      0       127      Bg4
>2       109     2       798      Bg4 Qf2
>3       99      7       4322     Bg4 Qf2 Bb6
>4       66      44      28050    Bg4 Qf2 Bb6 Kc1
>5       110     87      74448    Bg4 Qf2 Bb6 Kc1 Ne4 Bxb6 cxb6 Bxe4 Rxe4
>6       110     478     667798   Bg4 Qf2 Bb6 Kc1 Ne4 Bxb6 cxb6 Bxe4 Rxe4
>7       108     2491    3970606  Bg4 Qf2 Bb6 Bxb6 cxb6 Ne3 c5 Nxg4 Nxg4
>8       97      13183   20784340 Bg4 Qf2 Bb6 Bxb6 cxb6 Ne3 c5 c3
>
>In this position alam really benefits from nullmove pruning :)

It needs to search more time.

0.97 is not a convincing score for Bg4.

it needs to see at least +2.50

Movei also likes Bg4 at very small depthes but change it's mind to d4 only to
change it later again to Bg4.

Uri



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.