Rating, Covid and Junior Deflation

General discussions about ratings.
Post Reply
Joseph Conlon
Posts: 339
Joined: Thu Jun 06, 2019 4:18 pm

Rating, Covid and Junior Deflation

Post by Joseph Conlon » Fri Jul 30, 2021 10:54 am

In the old Clarke system, juniors were treated as a fresh player every year. The disadvantage was that the grades could be out of date at any one time; the advantage was that, when games were actually graded, the estimate used for a junior's strength was not that inaccurate.

We now have a situation with 4 figure ratings where junior improvement is dealt with by a k-factor of 40 in months of above expectation performance and a k-factor of 20 for months of below expectation performance.

Is this sufficient? My observation of junior events in the last few months is that currently (with the range of events taking place) is that the rating system is currently deflationary on established players. The reason is that, with Covid, many players (especially juniors) have improved a lot with the result that there are plenty of players whose last grade may be 0F (or 20E or some such) who are now playing at much higher levels. When these player enter events with ratings of 700, 800 or 900, they are vastly underrated and so have a net deflationary effect on everyone else (even with the larger k-factors), and it takes quite a number of games to move someone from 800 (say) to a 'true' playing strength of 1300.

In the long term, I suppose the idea is that this is resolved through equilibrium through the pool of junior-adult games and the different k-factors allow net improvement even with the junior-junior pool. However it seems that the long term may be the very long term, and that for quite a while to come junior ratings (especially rapidplay ones) may have a systematic bias towards deflation.

Roger de Coverly
Posts: 21291
Joined: Tue Apr 15, 2008 2:51 pm

Re: Rating, Covid and Junior Deflation

Post by Roger de Coverly » Fri Jul 30, 2021 11:34 am

Joseph Conlon wrote:
Fri Jul 30, 2021 10:54 am
However it seems that the long term may be the very long term, and that for quite a while to come junior ratings (especially rapidplay ones) may have a systematic bias towards deflation.
They've put in a trick whereby new players are assumed to score a draw against a dummy player, That has the practical effect of applying a minimum rating to new players whether deserved or not. Such is the turnover of junior players that many currently active underrated players will disappear fairly quickly.

Publication of the end July FIDE and ECF lists must be imminent. With most players at the recent Leamington 4NCL weekend in the Open at least having both ratings with the ECF one higher, it will be of interest to see how the performances diverge.

Post Reply