Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Strength of the engine in chess programs

Author: Uri Blass

Date: 09:22:40 05/20/02

Go up one level in this thread


On May 20, 2002 at 10:15:44, Rolf Tueschen wrote:

>Excuse me if I ask a question already answered a hundred times or more in this
>forum. Could someone explain, perhaps in understandable numbers of percentages,
>how important the strength of the engine is in chess programs, are there
>differences between commercial and amateur programs?
>
>Let me demonstrate a little thought experiment. If I would gauge (in 2002) the
>actually most known chess programs against say 1000 human chess players (first
>step) to get some insight into the Elo numbers, I would expect that the top
>programs would at best get Elo performances of 2200 - 2350, if I let the engines
>play without books and implemented book-like tricks.

I expect the best programs with no book to do better than 2200-2350 if the
humans also do not prepare special lines against the machines.

I think that book is overestimated
If you do a sweningen tournament between the top program on A1200 and no book
against the same programs on p200 with book I expect the program with no book
and better hardware to win(you need to do only 2 games between every 2 programs
in order to prevent repeating the same win again).

I say the following:

1)rating of top programs with no book on A1200 is bigger than rating of program
with book on p200(unless you do a match of many games)

2)rating of programs on p200 with book is better than 2200-2350.


I even believe that for the ssdf list it may be better to have 3 times faster
hardware and always change the first move and play with no book later.

The number of possibilities in the first move is 20 so the program need to lose
21 games in the same color in order to repeat the same loss twice.

There may be some stupid blunders like 1.e4 b5 or 1.e4 f6 or 1.e4 f5 but it is
possible to order this possibilities as number 18,19,20 so the program needs to
lose 17 games against 1.e4 in order to do it and this is not going to happen in
one match of 40 games.

If the tester delete all learning after every match I expect the programs with 3
times faster hardware but no book to win.

Maybe I am exagarating here but I am almost sure that at least A1200 with this
idea is better than P200 with no book.

Uri



This page took 0.01 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.