Skip to content
Matchup DNAUpdated recently

ATP and WTA forecasts, updated twice a day

Who wins the next match, and how sure should we be?

Matchup DNA rates every pro tennis player from their full match history, gives each upcoming match a win probability, and locks every forecast before the first serve so anyone can check how it does.

781

players rated

360,000+

matches in the history

12

forecasts locked so far

8

upcoming forecasts

Featured forecasts

All matches
Bybit Stockholm Open · Round of 32
Alex Michelsen64.6%
Yannick Hanfmann35.4%
Why? See the reasoning
1 of 8 upcoming forecasts

How a forecast is made

Four steps, from a match result to a scored prediction.

Every result moves two ratings: the winner gains, the loser drops, and upsets move them more. The gap between two ratings gives each player's chance. Drag the sliders.

A wins 70%B wins 30%

Illustration of the Elo formula only. Real forecasts use more than Elo.

Matches →Every upcoming match with its forecast, grouped by tournament.Is it any good? →Calibration, held-out tests, and the honest part: the betting market is better (0.590 vs 0.613 log loss).Forecast Archive →Every forecast ever issued, timestamped and impossible to edit.How it works →The data, the ratings and the model, in five short steps.