Author: Robert Hyatt
Date: 22:22:12 06/16/01
Go up one level in this thread
On June 16, 2001 at 11:19:15, Gian-Carlo Pascutto wrote: >On June 16, 2001 at 11:11:01, Mark Young wrote: > >>>>The problem with discussing this issue is some do not understand what ratings >>>>are, and how they work. As when Bob stated all players in the ratings pool must >>>>play each other a equal number of times for the rating system to be accurate. >>>>This is just incorrect. Thing start to become more ridiculous from there. >>> >>>It's an overstatement. >> >>No Dann the statment is just wrong. If you have two established players in the >>ratings pool that have never played each other before the system can and does >>predict accurately 1. Who the stronger player will be. 2. The dominace of the >>stronger one over the other. > >It will give an approximation for 1 and 2 in the general case. > >In no way is it guaranteed that the ratings system will provide an >_accurate_ prediction for either. > >There are way to many factors involved to allow this for specific cases. >On the average it will be right yes, but if you look at the individual >cases it will be off a lot too. > >-- >GCP It's a hopeless argument. Until you have seen two players meet and see a rocking upset, and then look to see why this happened, it won't ever be clear what Elo is all about. If I earn my rating by _only_ playing in the local chess club, and you earn yours playing somewhere else. The predictive powers of our two ratings are not very good. Normally this should not be a serious problem, due to lots of cross-playing between clubs at things like the US Open, and state-wide tournaments. But it _does_ happen. And it is certainly mathematically correct for it to happen. However, I think we can all save our breath...
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.