Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: CMQueen+ (CM6000.OBK) vs CMQueen+ ( Strong.obk Book) W21 L18 D11

Author: Jorge Pichard

Date: 08:44:54 04/22/00

Go up one level in this thread


On April 22, 2000 at 09:06:09, Michael Cummings wrote:

>On April 22, 2000 at 08:11:36, Jorge Pichard wrote:
>
>>On April 21, 2000 at 21:01:17, Michael Cummings wrote:
>
>
>>>I have not paid much notice to these games, but which book did you think was the
>>>strongest. CM6000.OBK is the strongest, the mentor.OBK or whatever is called is
>>>exactly the same, but it has a few notes that the help and the tutorial uses. So
>>>I have seen some people say the mentor is stronger than CM6000.OBK, thats funny
>>>and strange when it is the same book.
>>>
>>>I just hope you have used satisfactory settings, do not use any more than 2 meg
>>>hash for these short games, otherwise this will affect the games if they are too
>>>big.
>>>
>>>When you play about 200 more games, then I will look more closely at your
>>>results, I have played enough games battling chessmaster against chessmaster
>>>personalities to know 50 games means very little. Cause when you play those 50
>>>games again, the result can vary dramatically.
>>>
>>>You want to know another result that is also funny, over 20 20minutes games, the
>>>Karpov personality beat the Chessmaster by winning 14 games ?????
>>
>>I do not care much for the original Chessmaster personality, since it lost very
>>bad to my CMQueen+ personality setting Using the CM6000.OBK book, long before I
>>started to post my finding on this message board, therefore, only then knowing
>>that I discover this strong setting I Started to switch from one book to another
>>just to find out that the original CM6000.OBK (default book) is indeed the
>>strongest.
>>
>>PS: If you still remember how to set the CMQueen+ setting, Please do me a favor
>>and pit it against the Karpov personality, recommend to use a time setting of a
>>minimum of 1 hr/per side, pit it as many times as you are satisfied or in the
>>case when one setting is winning by a large margin, then  it becomes useless to
>>continue playing beyong 50 games.
>
>When you play personalities to see which is stronger, you should do it this way,
>otherwise your results can vary greatly, and most of the time, especially a
>short time settings, up to 15% of the games could be doubles of games previously
>played.
>
>So if you were playing your CMQ+ vs CM6000, have CM6K play white, finish the
>game, then have your CMQ+ play white from the end of the opening CM6K played in
>the previous game.
>
>Testing any other way is not acceptable (to me that is). My way your results
>reflect purely on your personality settings and nothing else.

Your way of testing seems very logical to me, therefore, I will provide you with
the CMQueen+ setting and I hope that you honestly test it against Karpov
personality or any of your choice. In case you never read the setting for CMQ+
here is the setting.

Personality    CMQueen+

Opening Book   CM6000


Style            Positional          Material

AD   5           M/P   5             Ow Q   9.9    (Please not not change these
SOP  100         COC   105           Op Q   9.7     controversial value for the
                                                    Queens or rooks)
R    0           M     120           Ow R   5.0
MSD  20          KS    100           Op R   4.9

SS   8           PP    110           Ow B   3.2
COD  1.0         PW    100           Op B   3.2

TD = ON                              Ow K   3.1
DT = ON                              Op K   3.1

                                     Ow P   1.0
                                     Op P   1.0











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.