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 | 327 | 0.878 | 61% |
| Base rate | 327 | 1.051 | 46% |
| Recent form | 327 | 0.924 | 60% |
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) | 23 Jan 2026 | 123 | 0.872 | 0.831 | −4.9% |
| Polymarket | 4 Sep 2026 | 63 | 0.914 | 0.856 | −6.9% |
| Kalshi | 13 Feb 2026 | 106 | 0.869 | 0.814 | −6.7% |
A draw is its own outcome: the model has a band around an even contest in which a draw is the likeliest single result, fit on the tuning seasons.
The gap between the top clubs and the rest is wide in the WSL, so ratings spread further than in the men's game.
There are no bookmakers' odds in the history, so there is no market comparison.
Whether the bottom side goes down directly or into a play-off depends on the season's rules; the odds are for finishing bottom.
Questions and answers
How accurate are PodiumOracle's Women's Super League predictions?
On 327 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 0.878, against 1.051 for the base rate and 0.924 for recent form. Against the betting markets on the same finished games, its log loss was 0.872 against the bookmakers' 0.831 over 123 games, 0.914 against Polymarket's 0.856 over 63 games and 0.869 against Kalshi's 0.814 over 106 games (lower is better).