Author: Dann Corbit
Date: 11:14:17 04/25/01
Go up one level in this thread
On April 25, 2001 at 13:16:09, Uri Blass wrote: >On April 25, 2001 at 13:02:22, Mogens Larsen wrote: > >>On April 25, 2001 at 12:45:10, Frank Wolf wrote: >> >>>Is there an explanation from the DF team? Has DJ just been lucky? Is there a >>>bug somewhere? My gut says something is wrong. >> >>There's nothing very unusual about winning five in a row in autoplayed >>engine-engine matches. >> >>Mogens. > >Assuming the probability to win with white is 40% and the probability to win >with black is 30% you get probability of 0.00432 to get 5-0 for Deep Junior. > >After the first games I guessed that it was only luck but now I guess that it is >not only luck so I expect Junior to increase the difference in the match. > >Junior sacrificed material in most of the games. > >Deep Fritz was very happy in part of them. > >In the first game Deep Fritz saw an advantage of 1.19 pawn at move 35. >In game 3 the score was even +1.31 for Deep Fritz at move 43. Stranger things have happened. I seem to recall an 11-0 start for one match in which the two engines were actually about equal. Maybe it was 8-0. At any rate, it won't end up 24-0, I can promise you. I do strongly suspect that Deep Junior really is superior, but it is only a feeling at this point. Perhaps Amir has been superior in his multi-threaded programming or even in his search or evaluation algorithms in general. At any case, Amir Ban is obviously one of the best chess programmers in the world. I'm not surprised at what I see so far.
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.