Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Rebel 9's time usage in handicap modes - realistic?

Author: Robert Hyatt

Date: 12:30:17 08/02/98

Go up one level in this thread


On August 02, 1998 at 12:25:52, Robert Pawlak wrote:

>On August 02, 1998 at 09:36:20, Robert Hyatt wrote:
>
>>>>3.  factor in some sort of random number so that positions that are bad
>>>>will randomly look less bad or actually look good.  Or, if the random
>>>>addition is large enough, but infrequent enough, the program might begin
>>>>to make tactical blunders (ICC has some *bach programs that do (or did))
>>>>this, and they would make an occasional blunder like a weaker human).
>
>I think I remember playing one of these, and having bach overlook a mate in 1!
>(is this possible, or is my memory confused) Yeah, this is what I want :)
>
>But seriously, from a marketing standpoint, I think that good handicapping could
>be an effective selling point for some of the up-and-coming software.
>
>It is probably too much to ask Ed which approach he uses - probably a trade
>secret.
>
>Bob P.

This is a difficult problem.  You want a program to play worse, but a 1500
player does't hang his queen very often, nor does he overlook mate in 1 very
often.  That's what I don't like about the handicapping ideas I have tried so
far...  program still plays like s 2200 player except for an occasional mistake
that looks like an 800.

I think you have to tone down the search + qsearch as Don suggested, but I think
you also *must* tone down the positional scores.  IE does a 1200 player really
know much about distant passed pawns, outposts, backward pawns, etc?  Crafty at
2 plies still plays non-1500 chess, although it might overlook something
tactically that a 1500 player would probably not...



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.