Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Emulating Human Chess Using Chess-Playing Programs

Author: Bob Durrett

Date: 13:28:13 01/04/04

Go up one level in this thread


On January 04, 2004 at 12:54:11, Odd Gunnar Malin wrote:

>On January 03, 2004 at 19:16:25, Bob Durrett wrote:
>
<snip>
>>
>>I believe that it is essential that the game scores of weaker players not be
>>mixed up with game scores of stronger players because the CHARACTER and rates of
>>occurrences of the errors will be heavily dependent on the ratings of the
>>players.
>>
>
>I'm not sure what your point is here. A database should include all games, as a
>designer of the database you don't know what the user want.

I think I understand your point.  The developer of the database will wish for
his product to be useful for a wide range of potential users.

My idea was merely to model [or faithfully emulate the play of] the human as
faithfully as possible.  Maybe that's not a good idea, I don't know.

The large database could be such that the estimated playing strength of the
humans was noted, perhaps in the pgn header.  Similarly for time limits.  Then
the analysis findings could be parameterized with rating and speed being
parameters.  Doing so would make it possible to see how the statistics changed
with change of either or both of these parameters.

For the purpose of study of the types of [and statistics of] the errors, maybe
it would be best to throw out games where the performance levels [of the humans]
could not be estimated prior to statistical analysis of the games.  Note that
some chess-playing programs estimate the strength of the human based on it's
records of past that human's past performance against that chess-playing
program.

Bob D.

>The only offer you
>can give is an easy way for the user to select his wanted games.
>For the rating issue there is a little problem with unrated players because the
>tag for 'unrated' is an addition to the pgn-spec. and isn't handled by the main
>gameviewers ([WhiteElo "-"]). I just tested it and it didn't seems to do any
>harm neither so I will use it in the database below. In addition I will of
>course write it in plain text in the event/remark field as before.

<snip>
>
>Odd Gunnar



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.