How good are these forecasts?
Every forecast scored on seasons the model never saw while its settings were chosen.
Held-back seasons
| Forecast | Forecasts | Log loss | Favourite won |
|---|---|---|---|
| The model | 2421 | 0.670 | 59% |
| Base rate | 2421 | 0.693 | 52% |
| Recent form | 2421 | 0.684 | 55% |
Against the betting markets
Each market's price before the start, scored on the same finished games as the model. A negative figure means the market was more accurate.
| Market | Since | Priced | Model log loss | Market log loss | Model vs market |
|---|---|---|---|---|---|
| Bookmakers (Pinnacle, DraftKings, bet365) | 6 Mar 2026 | 112 | 0.654 | 0.676 | 3.3% |
The ECHL sits below the AHL: players move up and down between the two all season, so rosters are less settled than in the NHL.
The playoff odds follow the league's format: the top four in each division, 1 v 4 and 2 v 3, best of seven throughout, division winners meet in the conference final, then the Kelly Cup Finals.
Call-ups to the AHL and NHL are not in the model until they show up in results.
Goaltending is not in the rating.
Questions and answers
How accurate are PodiumOracle's ECHL predictions?
On 2,421 games from held-back seasons (2024–25 to 2026–27), which the model never saw while it was tuned, the side it favoured won 59% of the time. Its log loss was 0.670, against 0.693 for the base rate and 0.684 for recent form. Against the betting markets on the same finished games, its log loss was 0.654 against the bookmakers' 0.676 over 112 games (lower is better).