Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: CM Personality 40/120 Tournament Finished

Author: John Merlino

Date: 15:49:16 08/14/01

Go up one level in this thread


On August 14, 2001 at 17:52:55, Chessfun wrote:

>On August 14, 2001 at 17:11:46, John Merlino wrote:
>
>>On August 14, 2001 at 14:28:38, Chessfun wrote:
>>
>>>On August 14, 2001 at 14:14:38, Sune Larsson wrote:
>>>
>>>>On August 14, 2001 at 13:09:49, Chessfun wrote:
>>>>
>>>>>On August 13, 2001 at 15:05:02, Kurt Utzinger wrote:
>>>>>
>>>>>>Thank you for playing such a lot of games and this comparison. But what I am
>>>>>>much more interested in is to have games of CM_Fun against other top programs
>>>>>>with a sufficient number of games on a playing level 40/40' or higher.
>>>>>
>>>>>IMO it makes no difference if the engines are all kings.
>>>>>As the parameters are totally different and Fun has such different material
>>>>>values It seems to me that beating a CM engine or personality is the same as
>>>>>beating another commercial engine.
>>>>>
>>>>>I have started playing with the exact same parameters and against
>>>>>the same engines as you played. I will also add Junior 7 and Deep Fritz.
>>>>>I will play 10 games against each engine.
>>>>>
>>>>>Engine-Engine Matches under CB-GUI
>>>>>ponder=off / time control 40/40' / 32 MB hash each
>>>>>Playing on two Thunderbird 945 mhz.
>>>>>
>>>>>Current;
>>>>>CM_Fun v Deep Fritz W1 L0 =0
>>>>>Game two looks also like Fun will win at move 60. +1.50
>>>>>Deep seemed to miss something at move 54.+0.56 55.-0.91
>>>>>I will post this game later, then complete all matches.
>>>>>
>>>>>CM_Fun v Hiarcs 7.32 W2 L0 =0
>>>>>
>>>>>[Client2Server]
>>>>>; TheKing 3.12 CM_Fun with Nz7book.ctg
>>>>>xboard := xboard\npost\nnew\ncm_parm default\ncm_parm opk=392439 tts=33554432
>>>>>opp=100 opn=100 opb=100 opr=100 opq=100 myp=114 myn=138 myb=138 myr=125 myq=138
>>>>>cc=100 mob=105 ks=134 pp=105 pw=125 cfd=0 sop=100 avd=5 rnd=0 sel=12 md=99\n
>>>>>
>>>>>Sarah.
>>>>
>>>> Do you really have different values for own and opp pieces?
>>>> If the above values are correct - why?
>>>>
>>>> Sune
>>>
>>>Actually no as during set up the creation of the log file
>>>created both op and my as the same. I typed in manually 100
>>>here as I posted it. This is the same as Kurt's and I simply wanted
>>>to avoid the confusion.
>>>
>>>Sarah.
>>
>>According to the personality you gave me, the piece values are:
>>
>>Queen = 12, Rook = 6, Bishop = 4, Knight = 4, Pawn = 1.1
>>
>>And they are identical for own and opp. Is this not correct? Am I confused? Or
>>is it possible that the confusion is coming from the values that are displayed
>>in the CM GUI might not be the same values that are actually sent to The King
>>engine?
>>
>>jm
>
>Hi John,
>       That looks right (see below) translating them for use in CB gui
>gives different values, why this is so I have no clue. But it's the same
>for all personalities. I assume it's some multiple of the default.
>
>; TheKing 3.12 CM_Fun with Nz7book.ctg
>xboard := xboard\npost\nnew\ncm_parm default\ncm_parm opk=392439 tts=33554432
>opp=114 opn=138 opb=138 opr=125 opq=138 myp=114 myn=138 myb=138 myr=125 myq=138
>cc=100 mob=105 ks=134 pp=105 pw=129 cfd=0 sop=100 avd=5 rnd=0 sel=12 md=99\n
>
>CMFun.CMP    Base Rating:  2724    CM8000.OBK
>
>A/D  (Attack/Defender)    [   5]    M/P  (Material/Position)  [  -5]
>S/P  (Strength of Play)   [ 100]    C/C  (Control of Center)  [ 105]
>Rand (Randomness)         [   0]    MOB  (Mobility)           [ 110]
>MSD  (Max Search Depth)   [  99]    K/S  (King Safety)        [ 140]
>SS   (Selective Search)   [  12]    P/P  (Passed Pawns)       [ 110]
>C/D  (Contempt for Draw)  [   0]    P/W  (Pawn Weakness)      [ 135]
>
>OWN              OPPONENT
>Queen   [120]    Queen   [120]
>Rook    [ 60]    Rook    [ 60]
>Bishop  [ 40]    Bishop  [ 40]
>Knight  [ 40]    Knight  [ 40]
>Pawn    [ 11]    Pawn    [ 11]
>
>Transposition Table  [16 MB]    Ponder On
>Player Type:  Personality       Gender: Female
>____________________________________________________________________
>--------------------------------------------------------------------
>
>Sarah.

Ok, your above cm_parm values make more sense. The reason that some of the
values are (or can be) different from what you see within the CM GUI is that the
Material/Position value is actually a MODIFIER for several other values, rather
than a value that is sent straight to the engine.

jm



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.