Author: Albert Silver
Date: 07:47:41 07/04/01
Go up one level in this thread
On July 04, 2001 at 10:06:07, Wayne Lowrance wrote:
>On July 04, 2001 at 00:50:10, Bruce Moreland wrote:
>
>>On July 03, 2001 at 15:36:08, ERIQ wrote:
>>
>>>now I know their cheating
>>
>>I started a 20-game match between my Gerbil program and Gnuchess. Won the first
>>four games. Drew the fifth. Lost the match.
>>
>>"They're".
>>
>>bruce
>Bruce with respect to you I say your Gerbil and Gnuchess are not Fritz and Tiger
>Wayne
Suppose we are of equal rating and decide to play a 40 match game. Let's also
suppose that we play at our respective strengths and no one is having a better
day than the other, so that the equal ratings are correct. The predicted score
should be 20-20, right? Statistics say nothing about how we'll get to that
score, only that that is what we should expect. After the first 20 games I am up
20-0. Perhaps probability will say this is unlikely, but still, it can happen.
After 40 games the score is... 20-20. Did anyone cheat?
In any case, I saw the games and must say that I think DF benefitted a little
from its openings. Even the Marshall, where the programs are stating White is
much better is very misleading. White is playing with pieces against a queen
except that his pieces are having real trouble coordinating properly (in other
words they can't seem to attack anything and are on the defensive, and White's
big plus is really only in the pawn count. However, look at where those pawns
are. Very tough situation, so I think the big plus is really an evaluation issue
as White has real trouble there. In two other games DF came out better in the
opening as soon as they left book. Probably only a coincidence that those
particular openings and games weren't played earlier in the match.
Albert
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.