How Tournament Elo works

Tournament Elo measures results in eligible tournament sets. It is completely separate from normal 1v1 and 2v2 Sets Elo.

Winning against a higher-rated opponent produces a larger expected gain. Score margin does not affect Elo. Byes, disqualifications, forfeits and incomplete/unreliable results do not count.

Confidence reflects rated set count, not fame, imported match totals or activity. PROVISIONAL: fewer than 10 sets. LIMITED: 10–29. ESTABLISHED: 30+.

Activity grants no free Elo. Inactive players keep their rating. The active window is 120 days since the last rated set.

Historical results may be imported after review. Adding older events replays results chronologically, so ratings can change while the public historical dataset is being assembled. This is an estimate of competitive strength, not a perfectly objective measure.

Rules v1: starting rating 1000; provisional / limited / established K = 64 / 48 / 32. Expected win probability = 1 / (1 + 10^((opponent − own) / 400)). Change = personal K × (result − expected), rounded to six decimals. Different personal K factors need not produce zero-sum changes.

All-Time is a transparent achievement view: wins, podiums, then peak Tournament Elo. Current ratings are not decayed for All-Time purposes.

Qualification rules are configured per series. Before organizer finalization, positions are provisional. Finalized qualifiers and locked seeds do not change automatically with Elo.