Track record
Every week we log three numbers per player before kick-off: our prediction, the official game’s own prediction, and once the whistle goes, what actually happened. The log cannot be edited after the fact. It starts at week one and never resets.
| Week | Our average miss | Official average miss | Who was closer | Players |
|---|---|---|---|---|
| 1 | 2.11 | 2.11 | Us+0.0% | 308 |
| 2 | 1.88 | 1.90 | Us+0.9% | 308 |
| 3 | 2.00 | 2.43 | Us+17.8% | 307 |
| 4 | 2.15 | 2.29 | Us+6.5% | 307 |
| 5 | 2.06 | 2.23 | Us+7.8% | 301 |
Predictions are snapshotted hourly before each deadline, so “before kick-off” is provable rather than promised. “Average miss” is how far a prediction landed from the real score, averaged over every player. Lower is better.
How it works, the whole story
Thousands of simulations of every fixture, minutes modelled first, scored under this season’s rules including the new defensive points most eyes still underprice. The rest is our edge. This page is where we prove it rather than explain it.