Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Test suites - can they reliably predict ELO?

Author: Enrique Irazoqui

Date: 06:49:07 12/13/99

Go up one level in this thread


On December 13, 1999 at 00:53:08, Tina Long wrote:

>Hi Enrique,

Hi Tina,

>When I was young I made my own test set from problems in the ICCA Journal in the
>1980's, I developed it over many years, but the problems were to simple and
>Genius2 onwards was solving half my testset in 0.00 secs.  I turfed it & found
>the job too daunting to start again.

It can be very frustrating, isn't it. :(

>Regarding your results F532 & F6 both on 2658.
>
>Did they get exactly the same answers right and wrong?
>
>I guess the answer is no.

You are right. F6 did better in some positions, worse in others.

>So, did the differences (F532 right F6 wrong & vice versa) tell you anthing
>interesting about the development from F532 to F6.
>
>I talk tactical vs positional, endgame vs middlegame etc.

It's a tactical test, so I can't see non-tactical patterns with it. Besides, for
this sort of thing I much prefer to watch the games and I didn't see many played
by F6. I am not being very helpful. Hmmm...

Enrique

>Thanks,
>Tina



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.