Author: pavel
Date: 08:10:40 09/14/00
Go up one level in this thread
>> >>this tourny is a good example of how important a big number of games are to come >>to *some* (or any) conclution about a set of programs strength. > >It is possible to come to some conclusion even without big number of games. > >You can see that Phalanx was lucky when it came for a good position out of book >in some games if you look not only in the results but also in the games. > >Uri it was lucky that Phalanx came out of book in good positions on some games, and that is why we need more games, to minimize the error margin. yes it is possible to come in to some concluions with small number of games, when the elo differance between 2 programs is reasonable (80-90 or higher). but if 2 programs are very close in strength (20-30), then its really hard to tell anything with small data. thanks Pavel
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.