Prediction accuracy: tennis
What is checked: who wins the match. An honest probability means outcomes at 70% happen about 70% of the time.
Check on history
The players’ ratings update after every match and the estimate is made before it; the final formula of each year was trained on the years before it only. The check covers every completed ATP and WTA match of the last two years. That is how 9,894 matches from 13 October 2024 to 12 October 2026 were checked.
| Model estimate | Estimates | Average | Happened |
|---|---|---|---|
| 0–10% | 327 | 6.8% | 4.6% |
| 10–20% | 1,094 | 15.7% | 17.6% |
| 20–30% | 2,066 | 25.4% | 25.8% |
| 30–40% | 2,935 | 35.2% | 34.5% |
| 40–50% | 3,472 | 45.1% | 45.0% |
| 50–60% | 3,472 | 54.9% | 55.0% |
| 60–70% | 2,935 | 64.8% | 65.5% |
| 70–80% | 2,066 | 74.6% | 74.2% |
| 80–90% | 1,094 | 84.3% | 82.4% |
| 90–100% | 327 | 93.2% | 95.4% |
How to read probabilities and calibration →
Only the model's favourite
| Confidence from | Share of matches | Estimate | Came true |
|---|---|---|---|
| 60% | 65% | 72.7% | 72.7% |
| 70% | 35% | 79.4% | 78.8% |
| 80% | 14% | 86.3% | 85.4% |
| 90% | 3% | 93.2% | 95.4% |
Overall the model's most likely outcome came true in 67% of matches, with the favourite estimated at 66% on average. When the two are close, the model neither overstates nor understates its confidence.
By competition
«Honest»: the gap between the favourite’s estimate and how often it won is within the random spread for that many matches.
| Competition | Matches | Favourite came true | Favourite’s estimate | Probabilities |
|---|---|---|---|---|
| ATP | 5,053 | 66.0% | 66.1% | honest |
| WTA | 4,841 | 67.0% | 66.9% | honest |
Since launch
We store the last estimate before each match and never change it afterwards. Since 29 September 2026 the most likely outcome came true in 230 of 358 matches with an expected share of 66%.
| Competition | Matches | Came true | Expected |
|---|---|---|---|
| ATP | 192 | 129 | 66% |
| WTA | 166 | 101 | 66% |
Latest 12 matches
- Ostapenko J.Pegula J.14 October · WTA · Wuhan28722The most likely outcome did not happen: 22:1Final score 2:1. Model's score: 0:2.20.5
- Volynets K.Guo H.14 October · WTA · Wuhan64361The most likely outcome did not happen: 11:2Final score 1:2. Model's score: 2:0.21.5
- Tjen J.Snigur D.14 October · WTA · Wuhan30702The most likely outcome did not happen: 22:0Final score 2:0. Model's score: 0:2.20.5
- Ruud C.Auger-Aliassime F.14 October · ATP · Shanghai42582The most likely outcome came true: 20:0Final score 0:0. Model's score: 0:2.22.5
- Paolini J.Bartunkova N.14 October · WTA · Wuhan40602The most likely outcome did not happen: 22:0Final score 2:0. Model's score: 0:2.21.5
- Putintseva Y.Shnaider D.14 October · WTA · Wuhan23772The most likely outcome came true: 20:2Final score 0:2. Model's score: 0:2, exactly right.19.5
- Muchova K.McNally C.14 October · WTA · Wuhan78221The most likely outcome came true: 12:1Final score 2:1. Model's score: 2:0.19.5
- Jovic I.Yastremska D.14 October · WTA · Wuhan78221The most likely outcome came true: 12:0Final score 2:0. Model's score: 2:0, exactly right.19.5
- Van De Zandschulp B.Jodar R.14 October · ATP · Shanghai44562The most likely outcome came true: 21:2Final score 1:2. Model's score: 0:2.22.5
- Bouzkova M.Chwalinska M.14 October · WTA · Wuhan56441The most likely outcome did not happen: 11:2Final score 1:2. Model's score: 2:0.21.5
- Frech M.Swiatek I.14 October · WTA · Wuhan14862The most likely outcome came true: 20:2Final score 0:2. Model's score: 0:2, exactly right.18.5
- Svitolina E.Samsonova L.14 October · WTA · Wuhan70301The most likely outcome came true: 12:1Final score 2:1. Model's score: 2:0.20.5