A Comet is shining above us!

[Gábor Szõts]

Geschrieben von: / Posted by: Gábor Szõts at 25 May 2000 08:09:47:
Hello fellows,
Last night I did the GS-2930 test for Comet B.22 (DOS version) and got the incredible result of 2632 Elo points!
For comparison: LG2000 2.7 scored 2614, while Crafty 17.10 scored 2581.
Computer: Celeron 433 MHz, hash tables: 32 MB
Comet solved 6 out of 14 problems, LG and Crafty 5 each. That means that for the latter two engines (and perhaps for Comet either) the BS-2830 test should also be performed in order to obtain more reliable values.
Cheers,
Gábor

[Dann Corbit]

Geschrieben von: / Posted by: Dann Corbit at 25 May 2000 09:24:14:
Als Antwort auf: / As an answer to: A Comet is shining above us! geschrieben von: / posted by: Gábor Szõts at 25 May 2000 08:09:47:

Hello fellows,
Last night I did the GS-2930 test for Comet B.22 (DOS version) and got the incredible result of 2632 Elo points!
For comparison: LG2000 2.7 scored 2614, while Crafty 17.10 scored 2581.
Computer: Celeron 433 MHz, hash tables: 32 MB
Comet solved 6 out of 14 problems, LG and Crafty 5 each. That means that for the latter two engines (and perhaps for Comet either) the BS-2830 test should also be performed in order to obtain more reliable values.

Quite an incredible result. Sometimes excellence in EPD test suites does not translate to excellence in actual play. On the other hand, sometimes it does!

I look forward to excellence from new iterations of Comet.


My FTP site

[Aaron]

Geschrieben von: / Posted by: Aaron at 25 May 2000 18:31:00:
Als Antwort auf: / As an answer to: Re: A Comet is shining above us! geschrieben von: / posted by: Dann Corbit at 25 May 2000 09:24:14:

Hello fellows,
Last night I did the GS-2930 test for Comet B.22 (DOS version) and got the >>incredible result of 2632 Elo points!
For comparison: LG2000 2.7 scored 2614, while Crafty 17.10 scored 2581.
Computer: Celeron 433 MHz, hash tables: 32 MB
Comet solved 6 out of 14 problems, LG and Crafty 5 each. That means that for >>the latter two engines (and perhaps for Comet either) the BS-2830 test should >>also be performed in order to obtain more reliable values.

Quite an incredible result. Sometimes excellence in EPD test suites does not >translate to excellence in actual play. On the other hand, sometimes it does!

I look forward to excellence from new iterations of Comet.

Just wondering what were the results of previous versions in the same test?
Is this a sudden improvement or has it being doing well at this test all this while? If so, i guess playing strenght doesn't correlate at all..

[Gábor Szõts]

Geschrieben von: / Posted by: Gábor Szõts at 26 May 2000 08:14:28:
Als Antwort auf: / As an answer to: Re: A Comet is shining above us! geschrieben von: / posted by: Aaron at 25 May 2000 18:31:00:

Hello fellows,
Last night I did the GS-2930 test for Comet B.22 (DOS version) and got the >>incredible result of 2632 Elo points!
Computer: Celeron 433 MHz, hash tables: 32 MB

Just wondering what were the results of previous versions in the same test?
Is this a sudden improvement or has it being doing well at this test all this while? If so, i guess playing strenght doesn't correlate at all..

The GS-2930 test is quite new, so I have no values for comparison. I intend to do the BS-2830 test at the weekend. For this I have the result of Comet B.11 which was 2469 on the same machine (if I remember well, I have to check this).
More about this on Monday...
On the other hand, in my opinion Comet has shown quite a progress recently, which is demonstrated e.g. by the results of the GS-5 tournament (see Frank's pages). So there has to be some correlation between test and tournament results after all.
Best wishes,
Gábor

[Gábor Szõts]

Geschrieben von: / Posted by: Gábor Szõts at 29 May 2000 07:58:53:
Als Antwort auf: / As an answer to: Re: A Comet is shining above us! geschrieben von: / posted by: Gábor Szõts at 26 May 2000 08:14:28:

Hello fellows,
Last night I did the GS-2930 test for Comet B.22 (DOS version) and got the >>incredible result of 2632 Elo points!
Computer: Celeron 433 MHz, hash tables: 32 MB

Just wondering what were the results of previous versions in the same test?
Is this a sudden improvement or has it being doing well at this test all this while? If so, i guess playing strenght doesn't correlate at all..

The GS-2930 test is quite new, so I have no values for comparison. I intend to do the BS-2830 test at the weekend. For this I have the result of Comet B.11 which was 2469 on the same machine (if I remember well, I have to check this).
More about this on Monday...
On the other hand, in my opinion Comet has shown quite a progress recently, which is demonstrated e.g. by the results of the GS-5 tournament (see Frank's pages). So there has to be some correlation between test and tournament results after all.
Best wishes,
Gábor

I obtained quite controversial results at the weekend. In BS-2830 Comet B.11 scored 2574, while Comet B.22 only 2506. I have doubts about the reliability of such tests, but I decided to invest some more time on the topic. I'll be back.
Gábor