Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Real Text Chess game analysis

Author: Peter Schäfer

Date: 02:32:36 02/01/05

Go up one level in this thread


On January 31, 2005 at 23:32:41, Michael Henderson wrote:

>>As a result of this line of play, you win a pawn.
>>Additionally, your mobility is slightly increased, and Black's
>>mobility is slightly decreased. Also, your pawn structure is
>>a little stronger, and Black's pawn structure is somewhat
>>weakened. Lastly, the pressure on your King is eliminated.
>
>The above paragraph is above is complete garbage for any chess player.

I think it's not /that/ bad. After all, it spots some motives in the position
(winning material, weakening pawn structure, etc.)
I'd say it could become pretty useful if it concentrated on the important
issues. For example, "eliminating pressure on your King" is insignificant when
you can just win a pawn.

I guess you could extract similar information from other engines, too, if they
printed more information about their evaluation function (not for all positions
in the search tree, of course, but only for the PV).
Maybe like this:

 PV is dxc6 Nh6 ...
 total score is +5.43
   this score is calculated from:
   material balance: +3.0
   white king safety:  +0.15
   black king safety:  -0.2
   mobility: +0.03

hope you get the idea...

Crafty already has some debug options to print this kind of stuff. If you could
enable it just for the PV it would be a useful start...

>the player nothing about what has happened or will happen in the game.  Relying
>on that garbage would probably worsen your game.

You must not expect too much. An engine can only search to a limited horizon,
and has (usually) little ideas about long term strategies.

--Peter





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.