Every figure below was computed the way the model works on game day: from games and information known before tip-off. Lines are the closing line from ESPN (DraftKings first). Regular season only.
Straight talk: the model has not beaten the closing line. Its average miss on the margin is about half a point worse than the market’s, and its picks have won about half of their bets against the spread. It is built to explain a game, not to promise an edge.
Season by season
| Season | Games | Winner: model | market | Margin miss: model | market | Total miss: model | market | ATS, edge 3+ | O/U, edge 3+ |
|---|---|---|---|---|---|---|---|---|---|
| 2022-23 | 1230 | 62.9% | — | 10.184 | 9.758 | 14.555 | 14.16 | 234-236 49.8% | 256-260 49.6% |
| 2023-24 | 1231 | 66% | — | 10.951 | 10.543 | 14.612 | 14.227 | 218-220 49.8% | 311-289 51.8% |
| 2024-25 | 1231 | 67.7% | — | 10.834 | 10.513 | 14.666 | 14.11 | 228-203 52.9% | 291-317 47.9% |
| 2025-26 | 1231 | 68.6% | — | 11.268 | 10.877 | 15.029 | 14.572 | 210-226 48.2% | 306-307 49.9% |
| 2026-27 so far | 92 | 63% | — | 12 | 11.163 | 15.395 | 15.293 | 17-20 45.9% | 22-21 51.2% |
1st half
The 1st-half model is fitted from the full-game model and past halves, using only games before each date. There is no 1st-half closing-line history before this season, so its betting record starts building now; until then it is graded on how close it gets to the half-time score.
| Season | Games | Winner | Margin miss | Half of full game | Total miss | Half of full game |
|---|---|---|---|---|---|---|
| 2022-23 | 1230 | 59.2% | 8.723 | 8.718 | 9.579 | 9.604 |
| 2023-24 | 1231 | 62% | 8.915 | 8.93 | 9.836 | 9.798 |
| 2024-25 | 1231 | 60.1% | 8.683 | 8.685 | 9.526 | 9.518 |
| 2025-26 | 1231 | 63.7% | 9.009 | 9.021 | 9.779 | 9.787 |
| 2026-27 | 92 | 60.7% | 9.4 | 9.411 | 9.646 | 9.64 |
How it works
- Each player gets a value, refit every few days from every earlier game, weighted toward recent ones, by how his team’s margin and total moved with his minutes.
- Before a game, the model estimates who plays: recent minutes, with players who missed their last game with an injury and those on the injury report discounted.
- That is blended with a team offence and defence rating, then adjusted for home court and back-to-backs.
- It tells you where it differs from the market (Model edge), in points. A big edge is a reason to look, not a bet.