Super Tournament III Qualifier (after 76 of 76 rounds, but..

[Graham Banks]

Geschrieben von:/Posted by: Graham Banks at 17 September 2004 09:34:26:

Athlon XP2000
256mb hash each (or closest allowed by each program)
3,4,5 piece tablebases
All using Fritz Powerbook tournament book (maximum variety, minimum 5 games for a move to be played, no learning)
40 moves in 40 minutes repeating.
Ponder off
Tournament run under Deep Fritz 8 GUI

The top three finishers aside from CM9000 Judge II and CM 10th Edition (which have already qualified) will progress to the final stage which will be played at 40/120 with the powerbook settings optimised. At this stage that means List 512, Pro Deo 1.0 and SOS 4 will be playing in the final. However……..there are approximately 100 games that I will now be rerunning due to unsatisfactory openings caused by the powerbook settings used. I don’t expect this to have much effect on final placings, but time will tell!

STANDINGS
57.5 - CM9000 Judge II
48.5 - CM 10th Edition (default)
46.0 - List 512
45.5 - Pro Deo 1.0
44.5 - SOS 4
43.0 - Gandalf 5.1
42.5 - Delfi 4.5
40.0 - Thinker 4.6c
40.0 - SmarThink 0.17a
39.0 - Deep Sjeng 1.6
37.0 - Pharaon 2.62
36.5 - Yace 0.99.87
35.5 - Gothmog 1.0b10
34.5 - Anaconda 1.6.2
34.0 - Tao 5.6
30.0 - Fruit 1.5
29.5 - Little Goliath Revival
28.0 - Comet B68
25.0 - Crafty 19.15
23.5 - Slow Chess 2.93a

An updated full crosstable and a download file of all games will be on Kurt’s website as soon as the games being rerun have finished. http://www.utzingerk.com/banks/st_03_qual.htm
Comments welcome as always.
Graham.

[Roger Brown]

Geschrieben von:/Posted by: Roger Brown at 17 September 2004 11:28:53:
Als Antwort auf:/In reply to: Super Tournament III Qualifier (after 76 of 76 rounds, but......) geschrieben von:/posted by: Graham Banks at 17 September 2004 09:34:26:

The top three finishers aside from CM9000 Judge II and CM 10th Edition (which have already qualified) will progress to the final stage which will be played at 40/120 with the powerbook settings optimised. At this stage that means List 512, Pro Deo 1.0 and SOS 4 will be playing in the final.

Hello Graham,
Your data will be useful particularly in debates as to the performance of Winboard engines on other guis at long timecontrols.
As you practically demanded comment, here are are a few...

(a) Could you post the precise settings for CM9000 Judge II? Both for use in CM9000 and, more importantly, for use outside of it? As in Winboard?
(b) Have you conducted any experiments with CM 10000. Could you post the precise settings for any good candidates? Both for use in CM 10000 and, more importantly, for use outside of it? As in Winboard....

(c) What were the precise settings that you used for ProDeo? Posting those would be helpful too. Did ProDeo hang on your machine during this contest?
(d) List has a couple of famous bugs. How were games decided where the bug kicked in? Or did you not encounter any problems?
However……..there are approximately 100 games that I will now be rerunning due to unsatisfactory openings caused by the powerbook settings used. I don’t expect this to have much effect on final placings, but time will tell!

I followed some of this in the CCC. I guess Crafty was one of the engines which suffered from the unsatisfactory openings. Which othere engines will you be replaying games for?
How do you plan too deal with that issue in future incidentally? The powerbook setting problem is what I am asking about.
Deep Sjeng's performance is definitely disappointing. I believe that engine capable of a much better showing in this field.
The above having been said, Thinker continues to impress.
As I said before, thanks for the free games and for your time and efforts. We may not comment after each posting or update but believe me, we are out there, watching....
(Evil laughter rises then fades......)
Later.

[Graham Banks]

Geschrieben von:/Posted by: Graham Banks at 17 September 2004 12:03:22:
Als Antwort auf:/In reply to: Re: Super Tournament III Qualifier (after 76 of 76 rounds, but......) geschrieben von:/posted by: Roger Brown at 17 September 2004 11:28:53:

The top three finishers aside from CM9000 Judge II and CM 10th Edition (which have already qualified) will progress to the final stage which will be played at 40/120 with the powerbook settings optimised. At this stage that means List 512, Pro Deo 1.0 and SOS 4 will be playing in the final.

Hello Graham,
Your data will be useful particularly in debates as to the performance of Winboard engines on other guis at long timecontrols.
As you practically demanded comment, here are are a few...

(a) Could you post the precise settings for CM9000 Judge II? Both for use in CM9000 and, more importantly, for use outside of it? As in Winboard?
(b) Have you conducted any experiments with CM 10000. Could you post the precise settings for any good candidates? Both for use in CM 10000 and, more importantly, for use outside of it? As in Winboard....

(c) What were the precise settings that you used for ProDeo? Posting those would be helpful too. Did ProDeo hang on your machine during this contest?
(d) List has a couple of famous bugs. How were games decided where the bug kicked in? Or did you not encounter any problems?
However……..there are approximately 100 games that I will now be rerunning due to unsatisfactory openings caused by the powerbook settings used. I don’t expect this to have much effect on final placings, but time will tell!

I followed some of this in the CCC. I guess Crafty was one of the engines which suffered from the unsatisfactory openings. Which othere engines will you be replaying games for?
How do you plan too deal with that issue in future incidentally? The powerbook setting problem is what I am asking about.
Deep Sjeng's performance is definitely disappointing. I believe that engine capable of a much better showing in this field.
The above having been said, Thinker continues to impress.
As I said before, thanks for the free games and for your time and efforts. We may not comment after each posting or update but believe me, we are out there, watching....
(Evil laughter rises then fades......)
Later.

Kurt did send me afile with the Judge II settings for all GUIs but I'll have to hunt for it unless Kurt is good enough to post it again here.
To be honest I haven't really experimented with CM 10th Edition as yet, but my Creep settings will probably be used in the final (default positional settings, selectivity 14, material values as for Judge II).
No problems at all with Pro Deo 1.0 for which I used Ed's original settings. For the final I'll use his latest settings which he has already posted.
No problems with List 512 either.
Unsatisfactory openings affected all engines. Crafty was actually one of those least affected! Hardest hit were Thinker, Pharaon, Yace and Comet.
I will optimise the powerbook settings for the final and this should solve the problems with unsatisfactory openings.
Thanks for taking an interest Roger. It's always a pleasure to get feedback, especially when it's constructive as opposed to out and out whingeing or criticism!
Graham.

[George]

Geschrieben von:/Posted by: George at 17 September 2004 13:48:57:
Als Antwort auf:/In reply to: Re: Super Tournament III Qualifier (after 76 of 76 rounds, but......) geschrieben von:/posted by: Graham Banks at 17 September 2004 12:03:22:

As you practically demanded comment, here are are a few...

(a) Could you post the precise settings for CM9000 Judge II? Both for use in CM9000 and, more importantly, for use outside of it? As in Winboard?
(c) What were the precise settings that you used for ProDeo? Posting those would be helpful too. Did ProDeo hang on your machine during this contest?

Kurt did send me afile with the Judge II settings for all GUIs but I'll have to hunt for it unless Kurt is good enough to post it again here.
No problems at all with Pro Deo 1.0 for which I used Ed's original settings.
For the final I'll use his latest settings which he has already posted.
Thanks for taking an interest Roger. It's always a pleasure to get feedback, especially when it's constructive as opposed to out and out whingeing or criticism!

CM9_Judge2 [August 2004]
******************************************************
CM9_Judge2.CMP
Created by Graham Bank, New Zealand
[Setting under CM9-GUI]
*******************************************************
A/D (Attack/Defender) [ -0]
S/P (Strength of Play) [ 100]
M/P (Material/Position) [ 0]
Rnd (Randomness) [ 0]
MSD (Max Search Depth) [ 99]
SS (Selective Search) [ 12]
C/D (Contempt for Draw) [ 0]
C/C (Control of Center) [ 105]
MOB (Mobility) [ 110]
K/S (King Safety) [ 175]
P/P (Passed Pawns) [ 100]
P/W (Pawn Weakness) [ 125]
OWN OPPONENT
Queen [10.3] Queen [10.3]
Rook [ 5.4] Rook [ 5.4]
Bishop [ 3.3] Bishop [ 3.3]
Knight [ 3.3] Knight [ 3.3]
Pawn [ 1.0] Pawn [ 1.0]

*************************
CM9_Judge2 textfile
[for use outside CM9-GUI]
*************************
;---- CM9_Judge2.txt
; A personality created by Graham Banks, NZ
cm_parm opp=100 opn=110 opb=110 opr=108 opq=114
cm_parm myp=100 myn=110 myb=110 myr=108 myq=114
cm_parm cc=105 mob=110 ks=175 pp=100 pw=125
cm_parm cfd=0 sop=100 avd=0 rnd=0 sel=12 md=99
;---- File end

***********************************
CM9_Judge2 InitString=
[for use outside CM9-GUI]
***********************************
[ENGINE]
Name=The King 3.23 Judge2
Author=Johan de Koning, Netherlands
Filename=Wb2Uci.exe
[OPTIONS]
Program=TheKing.exe
InitString=cm_parm default\ncm_parm opk=236642 tts=67108864 opp=100 opn=110
opb=110 opr=108 opq=114 myp=100 myn=110 myb=110 myr=108 myq=114 mycc=105
mymob=110 myks=175 mypp=100 mypw=125 opcc=105 opmob=110 opks=175 oppp=100
oppw=125 avd=0 cfd=0 sop=100 rnd=0 sel=12
HashOnCommandline=false
InitTime=2
Ponder=false
Logfile=false
Edit=setboard
UseSAN=false
Priority=normal
AddHintMove=false
ShowThinkingMove=false
Protocol=1

********************************
CM9_Judge2 Quark.log
[output for use outside CM9-GUI]
********************************
Input to Quark: xboard
Input to Quark: cm_parm opk=183392
can't interpret move
Input to Quark: post
Input to Quark: new
Input to Quark: level 0 10 0
Input to Quark: cm_parm default
can't interpret move
Input to Quark: cm_parm opp=100 opn=110 opb=110 opr=108 opq=114
can't interpret move
Input to Quark: cm_parm myp=100 myn=110 myb=110 myr=108 myq=114
can't interpret move
Input to Quark: cm_parm mycc=105 mymob=110 myks=175 mypp=100 mypw=125
can't interpret move
Input to Quark: cm_parm opcc=105 opmob=110 opks=175 oppp=100 oppw=125
can't interpret move
Input to Quark: cm_parm cfd=0 sop=100 avd=0 rnd=0 sel=12 md=99
can't interpret move
Input to Quark: cm_parm tts=4194304

What are these settings? Can you post them?

Well criticism(with a good intention) is always better, since it makes people think deeper. We all learn from our mistakes and if people don't criticise us even if they are wrong, we would not improve our thinking. (While the criticise on CCC about worthless tournament due to the bad openings was not correct and fair, i still believe it had some true points. These true points are what you/we should keep and improve our knowledge. In this case the improvement was your willing to play again these 100 games. At first it seems wrong, since the openings are random and all engines have the same probability to have a bad opening and an unfair treatment so everything should be OK, but actually it is
right to play again those games, since some engines perhaps are designed to play only usual(actually not bad) opening lines.

[Graham Banks]

Geschrieben von:/Posted by: Graham Banks at 17 September 2004 21:01:00:
Als Antwort auf:/In reply to: Re: Super Tournament III Qualifier (after 76 of 76 rounds, but......) geschrieben von:/posted by: George at 17 September 2004 13:48:57:

As you practically demanded comment, here are are a few...

(a) Could you post the precise settings for CM9000 Judge II? Both for use in CM9000 and, more importantly, for use outside of it? As in Winboard?
(c) What were the precise settings that you used for ProDeo? Posting those would be helpful too. Did ProDeo hang on your machine during this contest?

Kurt did send me afile with the Judge II settings for all GUIs but I'll have to hunt for it unless Kurt is good enough to post it again here.
No problems at all with Pro Deo 1.0 for which I used Ed's original settings.
For the final I'll use his latest settings which he has already posted.
Thanks for taking an interest Roger. It's always a pleasure to get feedback, especially when it's constructive as opposed to out and out whingeing or criticism!

CM9_Judge2 [August 2004]
******************************************************
CM9_Judge2.CMP
Created by Graham Bank, New Zealand
[Setting under CM9-GUI]
*******************************************************
A/D (Attack/Defender) [ -0]
S/P (Strength of Play) [ 100]
M/P (Material/Position) [ 0]
Rnd (Randomness) [ 0]
MSD (Max Search Depth) [ 99]
SS (Selective Search) [ 12]
C/D (Contempt for Draw) [ 0]
C/C (Control of Center) [ 105]
MOB (Mobility) [ 110]
K/S (King Safety) [ 175]
P/P (Passed Pawns) [ 100]
P/W (Pawn Weakness) [ 125]
OWN OPPONENT
Queen [10.3] Queen [10.3]
Rook [ 5.4] Rook [ 5.4]
Bishop [ 3.3] Bishop [ 3.3]
Knight [ 3.3] Knight [ 3.3]
Pawn [ 1.0] Pawn [ 1.0]
*************************
CM9_Judge2 textfile
[for use outside CM9-GUI]
*************************
;---- CM9_Judge2.txt
; A personality created by Graham Banks, NZ
cm_parm opp=100 opn=110 opb=110 opr=108 opq=114
cm_parm myp=100 myn=110 myb=110 myr=108 myq=114
cm_parm cc=105 mob=110 ks=175 pp=100 pw=125
cm_parm cfd=0 sop=100 avd=0 rnd=0 sel=12 md=99
;---- File end
***********************************
CM9_Judge2 InitString=
[for use outside CM9-GUI]
***********************************
[ENGINE]
Name=The King 3.23 Judge2
Author=Johan de Koning, Netherlands
Filename=Wb2Uci.exe
[OPTIONS]
Program=TheKing.exe
InitString=cm_parm default\ncm_parm opk=236642 tts=67108864 opp=100 opn=110
opb=110 opr=108 opq=114 myp=100 myn=110 myb=110 myr=108 myq=114 mycc=105
mymob=110 myks=175 mypp=100 mypw=125 opcc=105 opmob=110 opks=175 oppp=100
oppw=125 avd=0 cfd=0 sop=100 rnd=0 sel=12
HashOnCommandline=false
InitTime=2
Ponder=false
Logfile=false
Edit=setboard
UseSAN=false
Priority=normal
AddHintMove=false
ShowThinkingMove=false
Protocol=1
********************************
CM9_Judge2 Quark.log
[output for use outside CM9-GUI]
********************************
Input to Quark: xboard
Input to Quark: cm_parm opk=183392
can't interpret move
Input to Quark: post
Input to Quark: new
Input to Quark: level 0 10 0
Input to Quark: cm_parm default
can't interpret move
Input to Quark: cm_parm opp=100 opn=110 opb=110 opr=108 opq=114
can't interpret move
Input to Quark: cm_parm myp=100 myn=110 myb=110 myr=108 myq=114
can't interpret move
Input to Quark: cm_parm mycc=105 mymob=110 myks=175 mypp=100 mypw=125
can't interpret move
Input to Quark: cm_parm opcc=105 opmob=110 opks=175 oppp=100 oppw=125
can't interpret move
Input to Quark: cm_parm cfd=0 sop=100 avd=0 rnd=0 sel=12 md=99
can't interpret move
Input to Quark: cm_parm tts=4194304

What are these settings? Can you post them?

Well criticism(with a good intention) is always better, since it makes people think deeper. We all learn from our mistakes and if people don't criticise us even if they are wrong, we would not improve our thinking. (While the criticise on CCC about worthless tournament due to the bad openings was not correct and fair, i still believe it had some true points. These true points are what you/we should keep and improve our knowledge. In this case the improvement was your willing to play again these 100 games. At first it seems wrong, since the openings are random and all engines have the same probability to have a bad opening and an unfair treatment so everything should be OK, but actually it is
right to play again those games, since some engines perhaps are designed to play only usual(actually not bad) opening lines.

Hi George,
thanks for posting the Judge II settings for Roger. It should be noted that my name is Graham Banks, not Graham Bank.
I accept that constructive criticism is good and welcome it. People drawing attention to the problem regarding some oddball openings were right to do so and that's why I'm rerunning some of the games. I want my tournament to have some credibility. What annoyed me was the member in CCC who never posted any feedback until he came out with a total rundown of my efforts in a rather nasty manner.
Regards, Graham.