Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Evaluation Should Be Winning Probability - Not Pawns

Author: Ricardo Gibert

Date: 10:44:41 01/16/02

Go up one level in this thread


On January 16, 2002 at 13:26:24, Roy Eassa wrote:

>On January 16, 2002 at 07:41:28, Graham Laight wrote:
>
>>It has occurred to me that it is wrong to evaluate a position in terms of
>>relative pawns (the "de facto" standard - whereby an evaluation of 2 means that
>>you're approximately the equivalent of 2 pawns ahead).
>>
>>This means that many aspects of evaluation have to be squeezed into a dimension
>>which is not appropriate at all.
>>
>>A better way would be to evaluate "winning probability". If a position was a
>>draw, the value would be 0.50 (or 50%). If the player should win 3 out of 4
>>times, the eval should be 75%. If the player must win from here, then the
>>evaluation should be 100%.
>
>
>I had the exact same thought some time ago.  But I figured the "ingredients" in
>determining the value are the same either way.  There should be a simple way to
>convert from one to the other.  E.g., a table something like this might do the
>trick:
>
>Pawns ahead   % Chance to Win
>-----------   ---------------
>0.0           50
>0.1           51
>0.2           52
>0.3           53
>0.4           54
>0.5           55
>0.6           56
>0.7           57
>0.8           58
>0.9           59
>1.0           60
>1.5           66
>2.0           77
>3.0           88
>5.0           99
>
>Then the program could just show whichever value the user prefers.  (Of course,
>the probabilities I used above are off the cuff, but they seem reasonable to me
>at first glance.)

IM Kaufman did some statistical work and came up with 1 pawn worth to be about
75% if I remember correctly. I think he made some arbitrary assumption about
compensation, though, so I'd take it with a grain of salt.

Of course, I understand that you are merely providing an example. So my comment
is just FYI.



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.