Ninety90Machine vs the bookmakers
How good are bookmaker odds, really, and can a free model keep up? On 898 matches of 2025/26, 90Machine beat Bet365's closing odds, the bookmaker average and Pinnacle. Across all 1752 matches with odds it was level with the bookmaker average, and over six seasons it is slightly behind Pinnacle's close.
The numbers
898 matches of 2025/26, five big leagues, graded on the result.
| Forecast | Probability given to the result | Log loss | Brier |
|---|---|---|---|
| 90Machine Best | 38.22% | 0.9618 | 0.5720 |
| Bookmaker average, opening | 38.19% | 0.9626 | 0.5724 |
| Pinnacle, opening | 38.18% | 0.9628 | 0.5725 |
| Bookmaker average, closing | 38.17% | 0.9631 | 0.5729 |
| Bet365, closing | 38.16% | 0.9634 | 0.5731 |
| Pinnacle, closing | 38.13% | 0.9641 | 0.5736 |
All 1752 matches
| Forecast | Probability given to the result | Log loss | Brier |
|---|---|---|---|
| Bookmaker average, closing | 37.63% | 0.9774 | 0.5820 |
| 90Machine | 37.62% | 0.9776 | 0.5819 |
| Bet365, closing | 37.61% | 0.9779 | 0.5823 |
| Bookmaker average, opening | 37.59% | 0.9784 | 0.5825 |
Bookmaker odds include a margin; we remove it before grading, so every source is judged on its real probabilities.
How it is measured
Each match is graded by how much probability a forecast gave to what actually happened: home win, draw or away win. That is log loss, the standard measure for forecasts, and it punishes confident misses. Brier score, the average squared error, tells the same story. Higher "probability given to the result" is better; lower log loss and Brier are better.
Nothing is graded on matches the model was fitted on. Every forecast here came from a model trained only on earlier seasons, the same way it runs live. Since late October 2026 the 2025/26 season has also been used to decide which versions ship, so it is no longer an untouched test; the six-season figure is the fairer long-run view. The 2025/26 tables were measured with 90Machine 1.7; 1.7.1 changed only its calibration.
90Machine 1.7.1. Updated 4 November 2026. Forecasts are information, not betting advice.
Questions
Is 90Machine better than the betting odds?
Not overall. Over six seasons 90Machine is close to, but slightly behind, the sharpest closing odds. On the 2025/26 season it finished ahead by a small margin (it gave the actual result 38.22% on average against 38.13% for Pinnacle's closing line). Combined with the market, Ninety's best estimate is level with Pinnacle's close. Across all 1752 2025/26 matches with bookmaker odds it was level with the bookmaker average and ahead of Bet365.
How is accuracy measured?
With log loss: how much probability a forecast gave to what actually happened, punishing confident misses. It is the standard way to compare forecasts. Brier score (squared error) tells the same story. Each season is forecast by a model trained only on earlier seasons, so nothing is graded on matches it was fitted on.