Chess Tools

Tools for US Chess (USCF) players — compare two players head-to-head across every rating system, dig into one player’s stats and rating history, and search for upcoming tournaments, all straight from the official US Chess database.

About these chess tools

A rating is a measurement rather than a property, and the Elo system says something precise about what it measures. A 200-point gap predicts that the stronger player scores about 76 per cent — not 76 per cent of games won, but 76 per cent of available points, with draws counting half. A 400-point gap predicts about 91 per cent. The formula behind this is a logistic curve, which has a consequence worth sitting with: no gap, however large, predicts certainty. A 600-point underdog is still expected to take around five per cent of the points, so an upset is not evidence that the ratings were wrong. It is the system working as designed. The second thing worth knowing is that ratings from different pools are not comparable. A national federation rating, an online rapid rating and an online blitz rating are three separate scales that happen to use similar numbers, and they can differ by hundreds of points for the same player without anything being amiss, because each is calibrated only against the population it was earned in. Comparing them directly is the commonest mistake people make with chess numbers. The third is that rating change is driven entirely by the gap, not by how well you played. Beating someone 300 points below you gains almost nothing and losing to them costs a great deal, because the system measures surprise rather than quality. That asymmetry is why strong players avoid weak fields, and why a rating climbs fastest when a player is consistently underrated for their current strength — the number always lags the player, in both directions.

← Browse all 1001 tools