Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Rebel 8 - ELO level realism

Author: José de Jesús García Ruvalcaba

Date: 15:25:21 09/17/98

Go up one level in this thread


On September 17, 1998 at 18:05:15, John Coffey wrote:

>
>>that a rating of zero, under any of the lists currently published, is far weaker
>>than anybody would ever need.
>
>As to what I mean as 0:  At that level it would be a random move generator.
>You can get weaker than that but there is no point.
>
It would be interesting to rate a random move generator, by the way.

>As to a program being weaker than anybody would need:
>
>1.  Some programs don't even come close.   The lowest "rating"
>on fritz 5 when running on a Petium II 400mhz is 1350, but that
>rating often beats me at speed chess and I am strong A player.
>The only other way to weaken the program is to change the way
>it evaluates the position, or tweak the "blunder range" value.
>
It is obvious that their rating function is unaccurate. (I can only tell that by
your statement, as I do not have Fritz 5, or even a PC).

>2.  What if you were teaching chess to a small child?  I want to
>start teaching chess to my 4 year old nephew, who might not be
>much better than a random move generator.  Even most beginners,
>child or adult, need something pretty weak to practice against.
>
Small children are usually very intelligent, and if they are interested they
quickly become far stronger than a random player. But you are absolutely right,
and a long chain of defeats can discourage anybody, child or adult.

>John Coffey
José.



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.