Author: Manuel Rodriguez Blanco
Date: 07:02:38 06/09/98
Go up one level in this thread
On June 09, 1998 at 00:55:16, Christophe Theron wrote: >On June 08, 1998 at 18:42:07, Ed Schröder wrote: > >>>Posted by Fernando Villegas on June 08, 1998 at 14:54:25: >> >>>In Reply to: CCL results 08-06-98 posted by Ed Schröder on June 08, 1998 at >>>13:05:30: >> >>>On June 08, 1998 at 13:05:30, Ed Schröder wrote: >> >>>Hi Ed: >>>This one just to congratulate you because of your honesty and fair play. >>>I know more than a programmer that would have looked for a pretext to >>>end a tournament if his baby was not winning. Rebel has been >>>systematically in third place and even so you go on. Not that a >>>tournament with so few games is decisive, we all know that here, but >>>anyway for a lot less some people just run. That speak very well of you >>>commercial attitude and is a reassurance about the quality of your >>>products, not the contrary. >>>Fernando >> >>Thanks for all compliments. One thing I like to do on CCL is to play >>an NPS tournament in the tradition of the Crafty-Rebel event. This time >>with more games and not such a huge time difference, say factor 10. >> >>To some extend it will reveal something what you may expect from faster >>hardware the coming years and if speed is decisive. Is it? :) >> >>Say you play the NPS tournament, Program_X gets 4:30 per move, >>Program_Y gets 45:00 per move, both programs with permanent brain off. >>All six programs will play one game against each other. The program >>with 4:30 per move has always white otherwise the NPS tournament will >>take too long. >> >>The paring then would be: >> >>G5 - M7 N98 - H6 S2 - R9 >>M7 - G5 H6 - N98 R9 - S2 >>G5 - N98 R9 - H6 S2 - M7 >>N98 - G5 H6 - R9 M7 - S2 >>G5 - H6 S2 - N98 R9 - M7 >>H6 - G5 N98 - S2 M7 - R9 >>G5 - R9 S2 - H6 M7 - N98 >>R9 - G5 H6 - S2 N98 - M7 >>G5 - S2 R9 - N98 M7 - H6 >>S2 - G5 N98 - R9 H6 - M7 >> >>Say one game takes 60 hours = 2.5 days. 30 games will take 75 days. >>Using 2 autoplayer pairs the total NPS tournament would take 30-40 >>days which looks pretty acceptable to me. >> >>Interesting enough? >> >>Alternatives? >> >>Or better play the normal (10 game) 40/2:00 matches? >> >>- Ed - > >Sorry Ed, I think this is a complete waste of time. > >But maybe this has to be done one time, just to verify that we learn >nothing from such matches, and then definitely throw away such ideas. > >Please keep on doing the good work on normal 40/2:00. > > > Christophe I most differ from Christophe, the NPS match not only is very interesting, also this match will be the GREAT value for chess programers. Yes this match will advance the behavior in term of strength of the actually chess programs in future hardware, so the programer will be more orient in what proportion of effort (and what it pay) they most direct to a)speed programs or b)knowledge programs. I know that all programers have a good idea of this proportion, but this is an hypothesis, the match will be a practical proof. (only in term of near future hardware) Maybe when we finish the match some programers change a litter his perspective. Manuel
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.