Comet B68 against two of its predecessors

Archive of the old Parsimony forum. Some messages couldn't be restored. Limitations: Search for authors does not work, Parsimony specific formats do not work, threaded view does not work properly. Posting is disabled.

Comet B68 against two of its predecessors

Postby Robert Allgeuer » 10 Mar 2004, 10:08

Geschrieben von: / Posted by: Robert Allgeuer at 10 March 2004 10:08:50:

ponder off, 5min+2sec, all 3,4 and 5 men EGTB, Winboard, WBTM 0.60, Elostat 1.1b, Athlon 1.1GHz, Windows 2000
B44-2 I think is one of the strongest Comets, B62-3 the latest one I had before installing B68.
Robert


Comet B68 playing 20 games against each other engine. Score: 17.5 / 40 (43%)
 
Rank|No  |Name       |                    |Pts             |
----|----|-----------|--------------------|----------------|
  1.|  2.|Comet B44-2|1==011==0=0111110==0|  11.5 / 20  57%|
  2.|  3.|Comet B62-3|1==101=0=101=0110===|  11.0 / 20  55%|
Robert Allgeuer
 

Re: Comet B68 against two of its predecessors

Postby Ulrich Tuerke » 10 Mar 2004, 13:38

Geschrieben von: / Posted by: Ulrich Tuerke at 10 March 2004 13:38:12:
Als Antwort auf: / In reply to: Comet B68 against two of its predecessors geschrieben von: / posted by: Robert Allgeuer at 10 March 2004 10:08:50:
ponder off, 5min+2sec, all 3,4 and 5 men EGTB, Winboard, WBTM 0.60, Elostat 1.1b, Athlon 1.1GHz, Windows 2000
B44-2 I think is one of the strongest Comets, B62-3 the latest one I had before installing B68.
Robert


I agree.
However, it's very difficult to judge. There have been 2 testers (one of them Clemens who occasinally posts her and another guy from CSS forum) who had very intensively tested various Comet versions.
It's not easy to draw conclusions from their results even though they had generated a lot of statistics. Some versions seem to do exceptionally good against toppers but bad against other Comet versions, and so on. I wonder whether these conclusions may just be the result of an attempt to interprete statistical fluctuations.
There hadn't been really a version which had considerably better results than all of the others.
BTW, Clemens also came to the conclusion that the B4x versions had been quite good, which is a bit of a riddle to me anyway.
According to my own tests the B.67 / B.68 is at least equal, perhaps even a tiny bit improved.
In summary, the results are quite sobering.
Uli

>Comet B68 playing 20 games against each other engine. Score: 17.5 / 40 (43%)
> 
>Rank|No  |Name       |                    |Pts             |
>----|----|-----------|--------------------|----------------|
>  1.|  2.|Comet B44-2|1==011==0=0111110==0|  11.5 / 20  57%|
>  2.|  3.|Comet B62-3|1==101=0=101=0110===|  11.0 / 20  55%|
>
Ulrich Tuerke
 

Re: Comet B68 against two of its predecessors

Postby Robert Allgeuer » 10 Mar 2004, 15:34

Geschrieben von: / Posted by: Robert Allgeuer at 10 March 2004 15:34:07:
Als Antwort auf: / In reply to: Re: Comet B68 against two of its predecessors geschrieben von: / posted by: Ulrich Tuerke at 10 March 2004 13:38:12:
ponder off, 5min+2sec, all 3,4 and 5 men EGTB, Winboard, WBTM 0.60, Elostat 1.1b, Athlon 1.1GHz, Windows 2000
B44-2 I think is one of the strongest Comets, B62-3 the latest one I had before installing B68.
Robert


I agree.
However, it's very difficult to judge. There have been 2 testers (one of them Clemens who occasinally posts her and another guy from CSS forum) who had very intensively tested various Comet versions.
It's not easy to draw conclusions from their results even though they had generated a lot of statistics. Some versions seem to do exceptionally good against toppers but bad against other Comet versions, and so on. I wonder whether these conclusions may just be the result of an attempt to interprete statistical fluctuations.
There hadn't been really a version which had considerably better results than all of the others.
BTW, Clemens also came to the conclusion that the B4x versions had been quite good, which is a bit of a riddle to me anyway.
According to my own tests the B.67 / B.68 is at least equal, perhaps even a tiny bit improved.
In summary, the results are quite sobering.
Uli
>>Comet B68 playing 20 games against each other engine. Score: 17.5 / 40 (43%)
>> 
>>Rank|No  |Name       |                    |Pts             |
>>----|----|-----------|--------------------|----------------|
>>  1.|  2.|Comet B44-2|1==011==0=0111110==0|  11.5 / 20  57%|
>>  2.|  3.|Comet B62-3|1==101=0=101=0110===|  11.0 / 20  55%|
>>
Hi,
I have actually also played quite some games against lots of different opponents with B44-2, B62-3 and B60 and looked at some other Comets less intensively for my YABRL rating list (newest list posted some messages down). YABRL is Blitz of course, but there B44-2 is 20 points better than B62-3 and 40 points or so (did not check it now exactly) stronger than B60. This is after 900 games each, I think. With these few games played for B68 I do not draw any conclusions, probably with the exception that it is unlikely that B68 is _signifcantly_ better than the older versions.
Robert
Robert Allgeuer
 


Return to Archive (Old Parsimony Forum)

Who is online

Users browsing this forum: No registered users and 26 guests