Author: Dann Corbit
Date: 13:11:21 08/30/04
Go up one level in this thread
On August 30, 2004 at 15:49:31, David H. McClain wrote: >On August 30, 2004 at 15:00:43, Dann Corbit wrote: > >> >>Don't forget the Cadaques experiment. At one point it was a total blowout for >>Junior, and then Fritz knocked Junior silly. A small sequence can have a lot of >>error in many ways. It is also possible that the points of the opponent were a >>freak accident. So it might be 500 Elo stronger. Until 30 games have been >>played, the statisical significance is insignificant. >> >>Chances are good that Hydra is very strong. But we do not have enough data to >>really know it. > >Dann, > >In time it would be interesting to obtain all the games Hydra played on the >Chessbase server with the myriad of different hardware and programs it has >played against and let the experts take a good look at what it has accomplished. > >To date, the results of Hydra on the Playchess server are quite impressive by >percentage. However, it certainly is not unbeatable. > >It would also be interesting for Hydra to play on ICC for an ever more complete >overall evaluation. Many, including me, would be curious to simply play against >it and take our lumps. DHM Let's not forget that their online games may be tests, and with a tiny fraction of their available processor nodes. They may be fooling around with various parameters to discover what effect that they have on strength. Unless the contest is under somewhat formal conditions, I think the results are dubious.
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.