CEGT 40/40 download and rating updated

[Heinz van Kempen]

Hi all ,

ratings and downloads for CEGT 40/40 (2 Ghz) are updated.

Pure and CM rating list will be available again next Monday. The games Wilhelm Hudetz is currently playing with the Mr. Spock setting might have some influence on the latter.

http://www.husvankempen.de/nunn/downloads.htm

http://www.husvankempen.de/nunn/rating.htm

New completed tournament from Master Class league tomorrow.

Best Regards
Heinz

[Robert Allgeuer]

Hi Heinz,
I notice Thinker 5.0b in the list. Is this a public release (would not see it on the Thinker website, though) or a beta version?

Thanks
Robert

[Graham Banks]

Hi Heinz,
I notice Thinker 5.0b in the list. Is this a public release (would not see it on the Thinker website, though) or a beta version?

Thanks
Robert

Hi Robert,

the Thinker 5.0b version that you see in the CEGT list is the same beta version also being used by Leo in WBEC Ridderkerk. It was only released to a few testers.
Lance has made significant changes to this version and is no longer interested in this particular version's results or games. For that reason it is not available to share.

Regards, Graham.

[Tord Romstad]

Hi Heinz,

It is funny that the three Scandinavian entries (Gandalf, Ruffian, Glaurung) occupy three consecutive places on the best version list. In a few months, I hope the Norwegian program will no longer be the weakest of the three.

By the way, is there a good reason why Kiwi 0.6d and Kiwi 0.6b are both included on the best version list?

Tord

[Heinz van Kempen]

Hi Tord ,

Hi Heinz,

It is funny that the three Scandinavian entries (Gandalf, Ruffian, Glaurung) occupy three consecutive places on the best version list. In a few months, I hope the Norwegian program will no longer be the weakest of the three.

yes, really nice to have those strong Scandinavian programs. For Ruffian it seems that we will not have a new version anymore. But we can already see games for Deep Gandalf 7. Do you think that we will have one day a Glaurung SMP version? I know you are not overly ambitious, but somehow this seems to be the future with dual cores and more offered with steadily dropping prices. Glaurung is performing excellent in some CEGT tournaments and matches, except in the league so far. Here we have only 120 games, so promotion is based mainly on luck and the SMP versions of course are having a big hardware advantage.

By the way, is there a good reason why Kiwi 0.6d and Kiwi 0.6b are both included on the best version list?

Tord

Sure, my always present stupidity and superficialness . As these entries are removed manually you can be sure that I will always forget to remove one with each update.

Best Regards
Heinz

[Tord Romstad]

Hi Tord ,

Hi Heinz,

It is funny that the three Scandinavian entries (Gandalf, Ruffian, Glaurung) occupy three consecutive places on the best version list. In a few months, I hope the Norwegian program will no longer be the weakest of the three.

yes, really nice to have those strong Scandinavian programs.

Actually I don't care much about nationality; it was just funny to see the three programs line up like that.

For Ruffian it seems that we will not have a new version anymore.

You are probably right. I wrote to LokaSoft a couple of weeks ago and asked whether they planned a Ruffian version for Intel Macs, and their reply was that they hadn't even been able to contact Per Ola since a long time.

But we can already see games for Deep Gandalf 7. Do you think that we will have one day a Glaurung SMP version?

Almost certainly. Since yesterday, I own a dual core computer myself, and I don't intend to let that second core be idle.

How soon I will have an SMP version is hard to say, though. Opinions on how hard it is to make an SMP chess program seem to vary a lot. Because Glaurung is a very small and simple program I have some hope that the task will be easier for me than for most other authors, but I am far from sure. It also doesn't look like I will have much time for programming in the near future.

I know you are not overly ambitious, but somehow this seems to be the future with dual cores and more offered with steadily dropping prices. Glaurung is performing excellent in some CEGT tournaments and matches, except in the league so far.

Are the conditions the same? I am slowly beginning to suspect that Glaurung 1.0.x scores better when playing with its own book than when playing with a generic book or from fixed positions.

Here we have only 120 games, so promotion is based mainly on luck and the SMP versions of course are having a big hardware advantage.

Yeah, I'll have to do something about that before the next season of the league.

By the way, is there a good reason why Kiwi 0.6d and Kiwi 0.6b are both included on the best version list?

Sure, my always present stupidity and superficialness . As these entries are removed manually you can be sure that I will always forget to remove one with each update.

Aha, I see. I didn't know that you did it all manually. The number of mistakes would have been much bigger if I had made the lists.

One more question: Why does the darker background behind some of the entries on the CEGT list mean? I can't see any obvious pattern.

Tord

[Heinz van Kempen]

Hi Tord ,

this bright or darker squares are just random. In the original html file without deletions they alternate.

Conditions within CEGT are always the same regarding adapted time controls and generic books. So there might be differences to other lists with different conditions, especially own books and I guess the authors will like the variety of tournaments presented in all the fora.

Best Regards
Heinz

[Dariusz Orzechowski]

One more question: Why does the darker background behind some of the entries on the CEGT list mean? I can't see any obvious pattern.

All even-numbered entries are on dark fields.

[Uri Blass]

Hi Heinz,

I know you are not overly ambitious, but somehow this seems to be the future with dual cores and more offered with steadily dropping prices. Glaurung is performing excellent in some CEGT tournaments and matches, except in the league so far.

Are the conditions the same? I am slowly beginning to suspect that Glaurung 1.0.x scores better when playing with its own book than when playing with a generic book or from fixed positions.

Tord

I wonder if you think that it has better book.
I remember that I read that you changed the book and the question is what improvements did you make.

My book for WBEC is based partly on WBEC games of strong programs(another part is correspondence games of strong players and another part is games of strong humans) and it may be possible that having a book that is based on WBEC games when only moves of the winners are included can be effective for WBEC(note that my book for WBEC include only moves of the winner with white and mainly moves of the winner with black but also draws with black).


I wonder if some programmers tried to build a book only based on WBEC games without limit on the number of games and the only condition to include only game of the winners.

I suspect that such book may be effective for WBEC when programs may simply win games from book.

Note that I limited my book to 56 plies and a move is included only if it was played in at least 2 games so book wins of movei can be only short wins but if somebody includes only WBEC games of top programs in his book and include all moves that were played in at least one game then his program has a chance to win games directly from book in the same way that it happened in the past.

Uri