Backtested These matches were played before this site existed, so the model did not publish a forecast at the time. Each one is priced from a fit over matches played strictly before its own kickoff — never the match itself, never anything after it. That is an honest test, but it is not the same as a prediction made in advance, and it is not counted in the live record.
Matches priced
41
RPS
0.1985
baseline 0.1224
Correct calls
58.5%
baseline 57.1%
Log loss
1.1843
baseline 0.6829
| Date | Home | Score | Away | Model said | On the result | RPS |
|---|---|---|---|---|---|---|
| 20 Aug 26 | Rayo Vallecano | 1–1 | Alaves |
33%
33%
33%
|
33.3% ✗ | 0.111 |
| 19 Aug 26 | Atletico Madrid | 2–0 | Malaga |
33%
33%
33%
|
33.3% ✗ | 0.278 |
| 17 Aug 26 | Deportivo La Coruna | 1–1 | Elche |
33%
33%
33%
|
33.3% ✗ | 0.111 |
| 16 Aug 26 | Espanyol | 3–0 | Levante |
33%
33%
33%
|
33.3% ✗ | 0.278 |
| 16 Aug 26 | Racing Santander | 2–2 | Villarreal |
33%
33%
33%
|
33.4% ✓ | 0.111 |
| 15 Aug 26 | Sevilla | 2–1 | Rayo Vallecano |
33%
33%
33%
|
33.4% ✓ | 0.277 |
| 15 Aug 26 | Alaves | 3–0 | Getafe |
33%
33%
33%
|
33.3% ✗ | 0.278 |
No matches match those filters.
RPS scores a single match: 0 is a perfect call, 1 is being certain of the opposite. Around 0.2 on average is the level a bookmaker operates at.