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 | 917 | 0.669 | 61% |
| Base rate | 917 | 0.681 | 58% |
| Recent form | 917 | 0.677 | 58% |
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) | 22 Jan 2026 | 250 | 0.665 | 0.668 | 0.4% |
| Polymarket | 27 Nov 2025 | 132 | 0.613 | 0.616 | 0.3% |
| Kalshi | 17 Jan 2026 | 243 | 0.659 | 0.660 | 0.2% |
The table is the SHL's: three points for a win in regulation, two for a win in overtime or a shootout, one for losing in one.
The playoff odds follow the current format: the top six go straight to the quarter-finals, seventh to tenth play best-of-three play-ins, then best-of-seven rounds re-seeded so the best team left meets the worst.
The team that finishes last goes down, or into a relegation series with HockeyAllsvenskan; the page shows the chance of finishing last but doesn't simulate the qualifiers.
Goaltending is not in the rating, and live scores aren't shown for the SHL yet.
Questions and answers
How accurate are PodiumOracle's SHL predictions?
On 917 games from held-back seasons (2024–25 to 2026–27), which the model never saw while it was tuned, the side it favoured won 61% of the time. Its log loss was 0.669, against 0.681 for the base rate and 0.677 for recent form. Against the betting markets on the same finished games, its log loss was 0.665 against the bookmakers' 0.668 over 250 games, 0.613 against Polymarket's 0.616 over 132 games and 0.659 against Kalshi's 0.660 over 243 games (lower is better).