Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Chessbase adapter

Author: ujecrh

Date: 10:40:39 06/07/00

Go up one level in this thread


On June 07, 2000 at 09:51:34, Eric Trudel wrote:

>GoodDay,
>I think there was some talk about if those chessbase adapter's make engines
>weaker.
>I played 55 games each 10min long, with crafty 17.10 from chessbases site, and
>crafty 17.10 from the FTP site. I don't want to make any conclusions cause I'll
>probaly get jumped on :) Here they are +21 =32 -2 in favor of the native
>chessbase engine which of course uses NO adapter.  So again it was 37.0-18.0 in
>favor of no adapter. Now I know they're aren't enough games and all games were
>only played on a Intel Pentium 3, 500mhz. And yes the games were only 10min
>long. But still 37.0-18.0 is a very wide gap... I think...
>Bye-Bye.

There is no surprise for me in this result (even if the difference may well be
reduced a bit with more games played). This issue has been raised several times.

What happens is the winboard adapter sends a 'new' command after each move it
plays (followed by a 'force' and the replay of all moves so far). I guess that,
at least, most programs lose all benefit of pondering and restart their search
from scratch on the position.

Ujecrh



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.