Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: ChessMaster 6000 settings tests.

Author: Dan Kiski

Date: 05:36:03 12/08/98

Go up one level in this thread


On December 08, 1998 at 06:33:14, Harald Faber wrote:

>On December 08, 1998 at 05:41:54, Dan Kiski wrote:
>
>>>>>My Faber-settings ARE NOT GOOD for CM6000.
>>>>>They ARE GOOD for CM5X00 and probably CM4000.
>
>>I have played over 600 games using the faber\pilz settings on CM6000 against the
>>standard setting and other settings found on the web. In all my results that I
>>have posted here before the faber\pilz settings win clearly against all other
>>settings and also against other programs namely fritz 5.0, rebel 10 and CM5500.
>>
>>To say the faber/pilz settings don't work on CM6000 in my opinion just is not
>>correct. The settings beat all others endlessly.
>>
>>Computers used are two 233 mmx with 64 meg ram, CM 6000 settings at 16 meg hash.
>>Dan Kiski
>
>Ooops.
>Did you play 40/2h? If so then your investigations differ a lot from mine and
>facing so many games I will correct my statement. So, did you? With the original
>CM6k book?

On the faber\pilz settings I use the mentor.obk.
However all testing was done using the nunn opening positions as posted on
chessbase to give as equal a position as possible and to see only engine
strength.

I have played games against the standard settings using 40/2h, 1min move, 30
secs move, 15 secs move, 5 secs move.

Most of my investigation's (games) have been played against other settings than
the standard, since I found initially right off the bat that the faber\pilz
settings easily beat the standard settings, at all time controls.

After reading what I read here I was shocked to see that others get results
where the standard settings beat faber/pilz, therefore at this moment I have an
auto tourney running standard v faber/pilz at 40/2h tourney is set for 100
games. This tourney is with the mentor.obk as I have already seen that
faber/pilz will crush the CM using nunn.

All tourneys I played used the 100 game mark as standard and all games are
played using the nunn openings as posted at chessbase, excluding the current
tournament just started.

I also played the faber/pilz settings against itself with one using the
mentor.obk and the other using the cm6k.obk and found the mentor book to win a
100 game tourney by a score of 57-43.

As I stated I have also used the CM6000 faber\pilz mentor.obk v Fritz 5.0, v
Rebel 10, and CM 6000 faber\pilz v Fritz 5.0 and v Rebel 10 using the nunn
openings and found that neither Fritz or Rebel can win a match at 40/2h v CM6000
faber\pilz. These matches are on identical hardware 233 mmx both with 64 meg EDO
ram. I get a Fritz mark of 155 and 186,000 positions per sec with hash set at
46,000.

I am not at all in love with the CM 6000 and find that Fritz is the most user
friendly program and is my favourite, however as stated I have not found
anything that competes with the CM 6000 faber/pilz settings, including I should
add the CM 5500 faber\pilz v cm 6000 faber\pilz where I also find the 6000 wins
by normally a margin of around 59-41 at most time controls.

Dan Kiski.



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.