About Chess Stats
Chess Stats is a free set of tools for tournament players, parents and organisers. It answers the questions people ask before and after an over-the-board event: how do I match up against this player, who is entered in my section, how is the tournament likely to go, and what will my rating be afterwards?
What it does
- Individual Stats — a US Chess member's ratings, every rated event they have played, in-person events by state, their best wins, and Chess.com or Lichess accounts that may be theirs.
- USCF Compare Players — two members side by side: their head-to-head record, and how each scored against the opponents they have both played.
- USCF Find Tournaments — upcoming and past US Chess events.
- Tournament Predictor — load an entry list and simulate the whole event: a predicted cross-table, each player's chance of each finishing place, and the best, typical and worst rating change.
- USCF Rating Estimator and FIDE Estimator — your new rating after an event, using each federation's published formula, and any US Chess norms earned.
- Chess.com and Lichess Analysis — results by opening, colour, time control and opponent strength for any public account, and a comparison of two accounts.
Where the numbers come from
US Chess ratings, events and game results come from the official US Chess ratings database. Your browser cannot ask that database directly, so the US Chess tabs go through this site's ratings server, which fetches the public record and passes it on.
Both US Chess figures are shown: the published rating on the monthly list, which tournaments pair on, and the live rating after the player's latest rated event, which the Tournament Predictor uses for its estimates. FIDE ratings come from FIDE's monthly rating list, as Lichess publishes it, using the FIDE ID on the player's US Chess record.
Chess.com and Lichess data comes from each site's public data service. Your browser asks it directly for the username you type; the request does not pass through Chess Stats, and no login to either site is ever asked for or used.
Tournament entry lists can be loaded from Austin Chess, Coppell Chess Club, OnlineRegistration.cc (which many organisers use) or a public Google Sheet. Only what a prediction needs is kept from those lists: each player's name, US Chess ID, rating, section and requested byes. Schools, teams, grades, contact details and payment information are dropped before anything reaches your browser.
How the predictor works
Each game is scored from the two players' ratings using the expected-score curve that ratings are built on: a 200-point favourite expects about three points in four. The predictor pairs the event round by round the way a Swiss tournament director would, and builds the predicted cross-table by giving every game its most likely result. It then replays the whole event 1,200 times with every game decided at random at those odds; the finishing-place chances and the best, typical and worst rating changes come from the spread of those runs. The predictor guide has the details.
Rating changes use the published US Chess formula, including the bonus for beating expectation, and the FIDE estimator uses FIDE's own tables. The US Chess formula is checked against the worked examples in the US Chess rating document every time this site is built.
What it cannot tell you
A rating is an estimate of strength, not a measurement of it. Juniors improve faster than their ratings can follow, players have good and bad days, and an entry list changes up to the first round. Treat a prediction as the likely range, not the result. Chess Stats is not affiliated with US Chess, FIDE, Chess.com or Lichess, and their own published figures are always the official ones.
Who makes it
Chess Stats is built by Sumupid Productions LLC, which also runs Absurd Tools, where Chess Stats started. Ideas, corrections and bug reports are welcome on the contact page.