Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Yace 0.99.56 still the strongest amateur engine!? (Crafty 9.)

Author: Aaron Tay

Date: 18:40:03 09/05/02

Go up one level in this thread


On September 05, 2002 at 16:55:15, Bas Hamstra wrote:

>On September 05, 2002 at 16:31:46, Uri Blass wrote:
>
>>On September 05, 2002 at 16:15:36, Bas Hamstra wrote:
>>
>>>On September 05, 2002 at 15:20:55, Uri Blass wrote:
>>>
>>>>On September 05, 2002 at 14:39:09, Roy Eassa wrote:
>>>>
>>>>>On September 05, 2002 at 07:03:21, Joachim Rang wrote:
>>>>>
>>>>>>On September 05, 2002 at 04:20:24, Jouni Uski wrote:
>>>>>>
>>>>>>>Hmm... Yace still the strongest amateur!? Hasn't here been any progress in last
>>>>>>>year really in the top?
>>>>>>>
>>>>>>>Jouni
>>>>>>
>>>>>>
>>>>>>Aristarch 4.4 will beat Yace ;-)
>>>>>>
>>>>>>I', not sure, if Aristarch is the stongest Amateur (I doubt it), but at least
>>>>>>against Yace it's a bit stronger. I did several Matches between Yace and
>>>>>>Aristarch 4.4 and all ended with a slightly advantage for Ari.
>>>>>
>>>>>
>>>>>"List" is not included in the above, either, and there's some evidence that it's
>>>>>the strongest amateur program.
>>>>
>>>>I do not think that engine that does not know to ponder and does not have book
>>>>is the strongest amateur program.
>>>>
>>>>Pondering and book is part of the program.
>>>>
>>>>Uri
>>>
>>>Why not? What counts is the result you get. BTW my guess is Ruffian is on top.
>>>At FICS it was in the (standard) top-20 at a lousy P2-300. That's amazing. I
>>>would estimate Crafty would have at best 2100 at that hardware. Ruffian is over
>>>2400.
>>>
>>>Bas.
>>
>>The point is that the result of list in Kurt unziker's tournament was achieved
>>by a book that it not list's book.
>>
>>I have nothing against the tournament and I read Kurt's post but the
>>consequence
>>that there is evidence that list is the strongest amteur program
>>seems to be wrong.
>>
>>Mogens larsen test List against the first division programs in fair
>>condition(every program use it's book and 2 games)
>>that means that list is using no book and list scored sligthly less than 50%.

>I didn't know that, but that's still good if you think about it. The others are
>pondering, List it not. Even if zero pondermoves are guessed right, the others
>use 100% hardware when on move, List only 50%. Guessing 50% of the moves right
>will only make it worse than a factor 2 time disadvantage.

Er... I'm pretty sure Mogens Larsen tested with ponder off for both.After all
that is what "fair conditions" mean. (Books not withstanding)

I have done a few test games with Ponder off for both, the results are about
50%-60% depending on which engines are tested and time controls..

To be able to score 50% while the other guy is pondering and you are not is far
too impressive to pull off, difficult (impossible?) even for top commericials..


>And it still scored
>slightly less than 50%. You really start to wonder why he does not implement
>that quick win, because if he does he is back in business, I think.

Maybe because the score would remain about the same..

Aaron








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.