Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How accurate are the Winboard Programs Rating ?

Author: Jorge Pichard

Date: 06:32:35 04/13/00

Go up one level in this thread


On April 13, 2000 at 09:25:49, Jorge Pichard wrote:

>I have seen different Websites and almost all have LG2000 2.5 rated second after
>Crafty Fallowed by AnMon 5.6 as the third rated Winboard program. I agreed so
>far with those rating, but last week I started to test the newer version of
>Gromit 3.0 Versus Lg2000 2.5 and against AnMon 5.06.  Either the testers were
>using a different processors than I was producing different result, or the
>settings and opening books were differents, I used an AMD K6 II 500 Mhz and the
>result for Gromit 3.0 were much higher than either Lg2000 2.5 and AnMon 5.06.
>The time for the first test was set to game in 15 minutes per side, using
>Winboard version 4.07 and the same computer with 128 Mb of RAM. Gromit Score
>28.5 games out of 50 against Lg2000 2.5 and 31 to 19 against AnMon 5.06. Now I
>was very upset simply because my favorite program Lg2000 2.5 lost to Gromit 3.0,
>but when I tested against Anmont 5.06, I finally realized that Gromit was way
>underated. Then I tested them once more, increasing the time control to game
>in 30 minutes per side, and again Gromit 3.0 won both matches, this time the test was shorter
>only 30 games and Gromit score 16.5 out of 30 games against Lg2000 2.5 and
>against AnMon 5.06 Gromit Score 18 to 12. Now what is wrong with this picture?
>Gromit 3.0 clearly beat the crap out of the two leading programs in the Winboard
>rating list. Why does Gromit has such a low rating on all Winboard Rating List?



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.