Author: blass uri
Date: 14:29:51 07/28/00
Go up one level in this thread
On July 28, 2000 at 17:10:18, Mogens Larsen wrote: >On July 28, 2000 at 16:23:27, Ratko V Tomic wrote: > >>Of course, this isn't meant to mean that the conventional rating is useless. My >>original point is that the folks ridiculing a human player's judgment after a >>few games as worthless are missaplying, out of ignorance ands/or malice, the >>uselessness of a simple-minded statistical modeling (such as the memoryless >>steady process) to the strengths of human modelling of the situations with high >>uncertainty (such as a very small sample of games). > >We don't disagree on the usefulness of human evaluation on a very small sample >of games. However, the statistical model doesn't give absolute values, only >estimate and uncertainty. Coincidentally, the same applies to human evaluation. > >The assumption that a huamn player, however capable, can estimate within a broad >margin of error the strength of a program with very little data, is not correct. I disagree. It is possible to check it by a simple experiment. Ask the human who claim that (s)he can estimate the program's strength by one game to do it by getting tha data about one game. The human can look at one game including the main lines of the program and analyze the game later with programs. After it calculate an estimate for the rating of the program based on hundreds of games. I guess that in more than 60% of the cases the human will not be wrong by more than 200 elo. If you give the human to watch 5 games then I believe that in more than 60% of the cases the mistake in the estimate will be less than 100 elo. If you try to estimate the same only by the results of the games you will get more errors. 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.