FIDE have published a new article: FIDE seeks public discussion about proposed rating changes. They say:
In the past decade, certain innovations have caused rating deflation, a concern that has been raised by professional players and mathematicians and did not go unnoticed by FIDE. Players' ratings are spread out too widely, and the situation is deteriorating with each passing year. The problem arises due to various factors, including the minimal rating being set too low at 1000 points, changes to the famous 400-point rule and the way the initial ratings are calculated.
FIDE Qualification Commission and mathematician Jeff Sonas put significant effort into analysing the situation. As a result of this work, they propose two corrective measures.
The first measure, which they called compression, aims to adjust for deflation. All players rated below 2000, which constitutes the bottom 85% of all rated players, shall be given a one-time rating increase on the January 2024 rating list. The additional points awarded will gradually decrease from 400 (for those rated 1000) to 0 for those rated 2000. For instance, a player with a rating of 1600 will receive an additional 160 points.
The second measure, calculation improvements, focuses on improving the rating regulations to prevent future deflation. This includes raising the minimum rating from 1000 to 1400, restoring the 400-point rule to its earlier state so that it can apply multiple times in an event, and adjusting the formula for calculating initial ratings to incorporate two additional draws against hypothetical 1800-rated opponents.
Understanding the significance of resolving the issue of rating deflation, we are sharing these proposals and inviting the chess community to engage in a public discussion. You can read the full text of Jeff Sonas' suggestions here.
I kind of get the reasons for the rating calculation changes:
- Reintroduce the 400 point rule - the strong players (e.g. Magnus) have complained that they don't get enough rating points when they beat much lower rated players.
- Change the calculation of initial rating by introducing 2 hypothetical draws against 1800 players. It is correct that a player's first rating can be widely different from their real playing strength / performance.
For instance, if a player about 2000 strength gets their first rating in a tournament where they beat 5 players all rated 1000 then their first rating will wildly underestimate their true rating. Other federations also have separate calculations for first ratings. For instance, the English Chess Federation iterates. That is to say they perform the calculation and get a first estimate of the players rating. They then repeat the calculations as if the player already has this estimate as their rating. This gives a new rating which they again use in a new iteration. Eventually the iteration process reaches a limit and this is used as the players first official rating. For me that makes more sense than adding 2 hypothetical draws.
The proposed rating compression from January 2024 applies to ratings under 2000. Ratings below 2000 will be adjusted according to the following formula:
New rating = old rating + (2000 - old rating ) * 0.4
So anybody with a rating of 1000 will get a new rating of 1400 which will become the new minimum. Anybody with a rating of 1500 will get a new rating of 1700, etc.
I don't get this. To me this just looks like a vanity thing to make lower rated players feel better, particularly older players (like me) who are not necessarily worth the extra rating points. Can anybody explain to me how this change is supposed to help?