FIDE rating conference 2010

General discussions about ratings.
Sean Hewitt

FIDE rating conference 2010

Post by Sean Hewitt » Thu Jun 10, 2010 12:36 pm

Interesting (to a statto like me) discussion of FIDE rating "inflation"

http://www.chessbase.com/newsdetail.asp?newsid=6401

Kevin Thurlow
Posts: 5821
Joined: Wed Apr 30, 2008 12:28 pm

FIDE Ratings

Post by Kevin Thurlow » Thu Jun 10, 2010 1:01 pm

See http://www.chessbase.com/newsdetail.asp?newsid=6401

An interesting discussion on inflation in FIDE ratings. What is not mentioned is that in November 1986, every woman (except one) was awarded an extra 100 points for political reasons, which probably contributed to the rating inflation. A look at the ratings graph seems to suggest this was the case.
"Kevin was the arbiter and was very patient. " Nick Grey

Roger de Coverly
Posts: 21301
Joined: Tue Apr 15, 2008 2:51 pm

Re: FIDE Ratings

Post by Roger de Coverly » Thu Jun 10, 2010 1:27 pm

I don't really buy the Sonas argument that the top player in any era should be given a 2800 rating. I think somewhere you have to accept the argument that the top players have got that bit stronger than their predecessors particularly in openings and particularly over the last 20 years with the help of database research and chess engines. It's not easy to measure objectively - subjectively you can do it by looking at past move choices and past annotations. Old theory books sometimes say "x is bad because .." when today it's the current main line.

User avatar
Carl Hibbard
Posts: 6028
Joined: Fri Dec 08, 2006 8:05 pm
Location: Evesham

Re: FIDE rating conference 2010

Post by Carl Hibbard » Thu Jun 10, 2010 2:18 pm

A topic merge there, sorry to Alan who might have been reading it at the time!
Cheers
Carl Hibbard

Ola Winfridsson
Posts: 324
Joined: Fri Aug 14, 2009 12:26 pm

Re: FIDE Ratings

Post by Ola Winfridsson » Sun Jun 13, 2010 10:15 pm

Roger de Coverly wrote:I don't really buy the Sonas argument that the top player in any era should be given a 2800 rating. I think somewhere you have to accept the argument that the top players have got that bit stronger than their predecessors particularly in openings and particularly over the last 20 years with the help of database research and chess engines. It's not easy to measure objectively - subjectively you can do it by looking at past move choices and past annotations. Old theory books sometimes say "x is bad because .." when today it's the current main line.
Basically, I'd say that this report shows Sonas' approach to be too flawed to be practically useful. Isn't the fixed system he advocates unable to handle situations like the fall of the Berlin Wall and a rapid influx of strong players, or any other rapid increase in the number of players (such as the Fischer-Spassky effect)? He also seems to have a problem with fluctuations in form, which strikes me as very odd. Any grading or rating system can only ever reflect past performance, which, as any stock broker will tell you, is no guarantee of future performance! Only a reasonably good guide.