Author: Vincent Diepeveen
Date: 14:02:18 04/27/01
Go up one level in this thread
On April 27, 2001 at 10:40:45, Uri Blass wrote: >On April 27, 2001 at 08:56:51, Robert Hyatt wrote: > >>On April 26, 2001 at 23:03:46, Vincent Diepeveen wrote: >> >>>On April 25, 2001 at 21:52:51, Robert Hyatt wrote: >>> >>>>On April 25, 2001 at 20:02:27, Vincent Diepeveen wrote: >>>> >>>>> >>>>>I'm nodding bigtime here. SE is good to solve testsets, but not near >>>>>what recaptures and check extensions give. >>>>>As far as we know Deep Blue only played 10 practice games of around 30 0 >>>>>level against Rebel 8. >>>> >>>> >>>>Actually we know a lot more than that. We know they played 40 games >>>>as someone _else_ reported here after attending a lecture by one of >>>>the team last year. The 40 game match ended something like 38-2 but >>>>the original poster can give the exact quote. We also know it played >>>>hundreds of games vs GM Joel Benjamin (DB Jr). And we know it played >>>>dozens of exhibition matches against GMs at various conferences. I >>>>watched it play Byrne 6 games at a supercomputing conference exhibit hall. >>> >>>40 games is what i get for diep within a day of testing with Jan. >>> >>>40 games of 3 hours a game or more. >> >>Then you have longer days there than I have here. I only get 24 hours per >>day here in the US. > >In order to get 40 games of 3 hours a game per day you do not need more than 24 >hours if you use enough machines at the same time. > >I guess that Jan has at least 12 different computers(10 different computers are >enough only for 5 games at the same time and it means that the average length of >a game needs to be at most 3 hours in order to get 40 games per day) Jan has 18 computers, from which usually at least 12 play at the same time. Average game length is not so long as chessbase auto232 players quickly abort games when they fear they lose and fail low, despite that i had put in a 50 minutes timeout to start next game. So then they abort game after like 25 moves, but i learn then that this opening is cool to play 'em. Also some manage to get chessbase to use it as slave, i had big problems to manage that. Could only get'em to work as master. So quickly a new game gets started then, despite 50 minutes timeout... So quite some games are under 30 moves and unfinished (so called aborted). Other games are longer, but of course no game is longer as 60 moves. If one of the sides get mated then of course result can get stored and next game can start. I forgot whether chessbase auto232 player directly starts game then, idem for shredder and diep. No idea. I guess diep needs 50 minutes time out then in the few cases it is master. So a game on average hardly gets above 6 hours for the fast machines. Some slower machines like celeron500 play at 5 to 6 hours a game though, so that takes longer there. Those produce like 3 games a day (day=24 hours). But the fast machines, and Jan has a lot of them nowadays, they produce quickly games. Usually i get from Jan a bunch of games after a run of 3 to 5 days. Because of the many aborted games i hardly count the total number of games. Usually a combination produced rapid games so those are inside the testset too, though i hardly check them out. But with nowadays 12 machines playing and in past up to 18 machines playing the number of games produced is *huge*. Way more as diep plays games on FICS + ICC and most important is that i also have the scores of the opponent and other interesting info which you do not have from ICC/FICS. Best regards, Vincent >Uri
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.