Author: Uri Blass
Date: 07:02:41 01/27/01
Go up one level in this thread
On January 27, 2001 at 09:08:15, Mogens Larsen wrote: >On January 27, 2001 at 07:58:03, Uri Blass wrote: > >>The score does not prove that something is wrong but the score leads people to >>suspect that something is wrong. > >Yes. > >>It is logical to ask questions when the information is more unexpected. > >The problem is that people jump to conclusions instead of asking questions. > >>I agree that it is also possible that something is wrong with 23-17 result but >>the probability that something is wrong after you know thet the result is >>31.5-9.5 is bigger. > >No. > >Mogens. Yes. I will give an extreme example. Suppose that you get in the ssdf list Fritz3(p90)-Fritz6(K6-2 450) 40-0 I think that in this case it is clear that something is wrong with the result. The 31.5-9.5 result is not so extreme and does not prove that something is wrong but you know that the probability that something is wrong is bigger relative to the case of 23-17 for deep fritz that is closer to the expected result of Rebel based on the ssdf rating(Deep Fritz does not have ssdf rating but we can guess based on other results that it is not clearly better than Fritz6). 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.