Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Hard test position for chess programs( Steve Ham-Nimzo7.32)

Author: Ed Schröder

Date: 00:02:49 06/18/00

Go up one level in this thread


On June 17, 2000 at 18:10:24, Jeff Lischer wrote:

>On June 17, 2000 at 12:44:21, blass uri wrote:
>
>>>Here's what I get with an overnight analysis using Rebel Century 1.2
>>>(Anti-GM=Strong, PC: PIII 600, 60 MB hash table):
>>
>>I remember that Ed claims that it is not a good idea to use anti-GM=strong with
>>analysis.
>>
>>I found that Rebel can avoid c3 also without anti-GM.
>>
>>Uri
>
>I haven't heard that recommendation from Ed. I remember that he has said
>Anti-GM=Smart is not good for analysis. This is because the Smart setting is a
>function of the game history (i.e. previous moves) and will give erratic results
>with analysis.
>
>I thought Anti-GM=Off versus Strong was more a matter of personal taste. Early
>on, I do recall him saying he personally didn't like Rebel's playing style as
>much with Anti-GM turned on, but I don't know what he thinks about it these
>days.
>
>What I've found in comparing the analysis results of the two, is that when they
>are different, sometimes go to the same move if given more plies to think. When
>you found that Rebel avoided c3, at what ply did it finally leave c3 for good?

I will try to be complete...

anti-GM=SMART only makes sense playing games since this setting (also) relies
on the game history. Using the SMART setting in positions sometimes will
produce other moves because of the lack of the game history.

The ACTIVE and STRONG setting of anti-GM are different. They don't need the
game history and they will always produce the same move. In general especially
anti-GM=STRONG has a better feeling for tactics. If you are testing tactical
positions I would recommend this setting.

My personal taste for playing games goes to the default setting anti-GM=OFF
except for playing humans below 2500. The OFF setting ensures Rebel to play
its original style as implemented in the evaluation function.

The SMART option tries to solve or avoid some typical weak points of computers
such as a) avoid strategic positions b) make sure to have the initiative and
c) strive for tactical positions.

Personally I don't like any off these options as they (sometimes) have a huge
effect on Rebel's playing style and positional understanding making moves I
dislike. But it is all a matter of taste and I know that other people have
the STRONG option for instance set to ON all the time.

Playing GM's I always have SMART=ON as the benefits of SMART are superior to
its failures playing these giants of chess.

Ed




This page took 0.01 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.