Author: Andrew Dados
Date: 11:32:47 06/11/01
Go up one level in this thread
On June 11, 2001 at 14:08:52, Gian-Carlo Pascutto wrote: >On June 11, 2001 at 13:50:58, Andrew Dados wrote: > >>You may try to treat one chess game as two connected binomial events - this >>model introduces draw scores nicely. >> >>However it also reduces 'played games' in your table by 2, so in fact 7-3 >>becomes a valid 'required score'. >> >>This due to fact that now one game score carries more information (as it >>should) then win-loss. > >Uhh something is wrong here. Using Peters utility, which, as far as I >can tell, provides correct results and takes the draws correctly into >account, I confirmed that the numbers in the table are correct as long >as all games are decisive. > >If some games were drawn the reliability seems to go up in all cases, >so the numbers in the table can be considered 'safe' values. > >As far as I can tell your method is false because it assumes a game >carries now twice as much information, which is false. Well you need 2 two-state results to combine it into one 3-state result. So I am pretty sure the assumption above is true. > >-- >GCP
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.