Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: program for analyze.....

Author: Uri Blass

Date: 22:51:53 11/01/00

Go up one level in this thread


On November 01, 2000 at 22:08:41, Roy Brunjes wrote:

>On November 01, 2000 at 22:00:01, pavel wrote:
>
>>HI,
>>   I would like to know your preferance of program to analyze games or
>>positions. I personally use crafty and junior6.
>>
>>I also have all versions of fritz, sos(yt), gandalf, wbnimzo2000b, lg2000, and
>>most other freeware engines..
>>
>>thanks
>>Pavel
>
>Pavel,
>
>You've hit upon my single biggest use of chess software.  I prefer Rebel Century
>(3.0 just shipping now and a great product for analysis).  I also use Junior 6
>though I trust Rebel's eval a bit more because it is steadier over time.
>Junior's jumps around a bit more than Rebel's, though not nearly as much as
>Fritz's.

You did not say the version of fritz and I am interested to know if you mean to
Fritz6a.

I read that Fritz6a does hardly evaluations at the root when previous versions
of Fritz were preproccesors.
>
>Gambit Tiger is a lot of fun to play, but it's eval is all over the map.  Rebel
>Tiger's eval is a bit more consistent than Gambit Tiger's eval.
>
>It will be interesting to see what others prefer for their analysis.
>
>Roy

Programs will have a jump in the evaluation if they see a tactical idea.
There is a problem if the evaluation jumps because of preproccesing.
I saw this problem mainly with old versions of Fritz and with Genius.

I remember that I analyzed a game of myself with Genius3 and Genius3 said that I
had 0.5 pawn advantage and suggested both sides to castle.
Genius3 did not change its mind about the moves even after a long time and the
evaluation almost did not change.
After playing the expected moves of castling genius3 saw that I had no advantage
and not because of deeper search.

I hate this behaviour because I cannot learn about the mistakes in the game.
Did I have an advantage?
If I had an advantage what was the mistake that I did?
Genius3 did not answer these questions.

I want to generate a tree in order to analyze interesting positions and I want
to use the scores of the leaves after search in order to decide about my moves.

This way of analysis can cause problems with preproccesors because the score in
one line when the leaf position is an endgame is not comparable with the score
in a second line when the leaf position is in the middle game.

Uri



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.