The state of chess ratings

What this series measures

Once a month, every rating pool this site tracks is captured and compared with the month before: the Lichess distributions, the FIDE lists, and the top of both online leaderboards. The edition for that month is what changed, written out.

The point is that rating pools drift, and almost nobody notices. A distribution that slides fifteen points over a year moves every percentile on this site with it, and a reader who was told they were in the top 12% last spring deserves to know whether that number moved because they did or because the pool did.

Each edition names the date its data was pulled and carries the same tables the rest of the site uses, so a claim in one can be checked against the tool in the other.

How to read a month

Most months are quiet, and the editions say so plainly rather than manufacturing a story. A month is classified by what the data actually did — one pool moving alone, two pools moving in opposite directions, a broad drift, or nothing much — and the paragraph written for that shape is the one you get.

That is deliberate. A template with a number dropped into it says the same thing about a two-point drift and a sixty-point collapse, and twelve of those a year would be a content farm with better manners.

So a short edition is a real result. It means the month was uneventful, and the alternative — writing four hundred words about nothing — would make the interesting months harder to spot rather than easier.

Who moved

Alongside the pools, each edition lists the players whose ratings moved most over the month at the top of the game. These are titled players only: a public leaderboard is a place people expect to be named, and an untitled account holder never asked to appear in a monthly write-up.

A large monthly swing near the top usually means a tournament rather than a transformation, and the editions try to say which where it is knowable.