Home › Accuracy
F1 Fantasy Prediction Accuracy (2026)
A public track record for BoxBoxF1Fantasy predictions across completed 2026 rounds. This page shows the misses as well as the hits.
Round-by-round accuracy
| Round | Race | Driver pts MAE | Constructor pts MAE | Finish pos MAE | 90% CI |
|---|---|---|---|---|---|
| R1 | Australian GP | 10.7 | 19.6 | 4.1 | 91% |
| R2 | Chinese GP | 16.3 | 28.8 | 4.7 | 73% |
| R3 | Japanese GP | 6.6 | 10.6 | 1.8 | 100% |
| R6 | Miami GP | 15.0 | 26.6 | 3.2 | 73% |
| R7 | Canadian GP | 13.8 | 19.1 | 3.9 | 77% |
| R8 | Monaco GP | 14.3 | 18.5 | 4.8 | 36% |
| R9 | Spanish GP | 11.8 | 15.2 | 3.9 | 68% |
| R10 | Austrian GP | 7.7 | 11.0 | 1.9 | 100% |
| R11 | British GP | 11.9 | 18.8 | 4.3 | 95% |
Biggest driver-point misses
These are useful because they show where the model was most wrong, usually because of DNFs, penalties, strategy swings, weather or surprise race pace.
| Round | Driver | Pred. | Actual | Abs. miss |
|---|---|---|---|---|
| R8 | Charles Leclerc | 35.1 | -13 | 48.1 |
| R2 | Lando Norris | 35.8 | -10 | 45.8 |
| R6 | Nico Hulkenberg | 12.3 | -29 | 41.3 |
| R2 | Oscar Piastri | 33.0 | -7 | 40.0 |
| R11 | Max Verstappen | 34.7 | -3 | 37.7 |
| R7 | George Russell | 39.5 | 4 | 35.5 |
| R6 | Liam Lawson | 23.4 | -12 | 35.4 |
| R9 | Kimi Antonelli | 28.9 | -4 | 32.9 |
How to read this
Fantasy-point accuracy is harder than finishing-position accuracy because fantasy scoring also includes qualifying, overtakes, positions gained, fastest lap, Driver of the Day, sprint scoring, constructors, pit stops and DNF penalties. Confidence interval coverage shows whether the uncertainty bands are calibrated, not whether every single pick was close.
FAQ
How accurate are BoxBoxF1Fantasy predictions?
Across the completed rounds currently published here, driver fantasy-point MAE is 12.0 points and race-position MAE is 3.6 positions. This page updates when completed-round actuals are exported.
What does MAE mean?
MAE means mean absolute error: the average size of the miss, ignoring whether the prediction was too high or too low. Lower is better.
Why publish the misses?
Publishing misses keeps the model honest. F1 Fantasy is noisy, and weather, DNFs, safety cars, penalties and strategy can all create large errors.