Marginal Distributions: Goals
The number of goals each side scores — the marginal distributions that the scoreline grid below combines into a joint distribution.
What Drives This Forecast
The model's inputs in four groups, each shown as how far it tilts the expected goals towards one side. Together they give the expected-goals ratio behind the probabilities above.
Liverpool expected to score 1.49× as many goals (1.63 v 1.09) → 49.3% / 25.8% / 24.9%.
What the model weighed, not causes: its inputs overlap (Elo, long-run form and this season's goals all measure strength), so the split between bars is the model's partial credit. Whiskers show the 90% range of the model's uncertainty about each weight.
Scoreline Probabilities
Each cell is the probability of that exact full-time score. Hover a cell for its 90% credible interval. Colour shows the result — Win: Liverpool, draw, Win: Manchester United — shaded by likelihood.
| Liverpool | Manchester United | ||||||
|---|---|---|---|---|---|---|---|
| Goals 0 | Goals 1 | Goals 2 | Goals 3 | Goals 4 | Goals 5 | Goals 6 | |
| Goals 0 | 7.2 | 6.6 | 3.9 | 1.4 | 0.4 | 0.1 | 0.0 |
| Goals 1 | 10.1 | 12.3 | 6.4 | 2.3 | 0.6 | 0.1 | 0.0 |
| Goals 2 | 8.7 | 9.5 | 5.2 | 1.9 | 0.5 | 0.1 | 0.0 |
| Goals 3 | 4.7 | 5.2 | 2.8 | 1.0 | 0.3 | 0.1 | 0.0 |
| Goals 4 | 1.9 | 2.1 | 1.2 | 0.4 | 0.1 | 0.0 | |
| Goals 5 | 0.6 | 0.7 | 0.4 | 0.1 | 0.0 | ||
| Goals 6 | 0.2 | 0.2 | 0.1 | 0.0 | 0.0 | ||
| Goals 7 | 0.0 | 0.0 | 0.0 | ||||
Grid covers scores up to 5–5; cells need not sum to 100% (rare higher scores fall outside the grid). The outlined cell is the most-likely scoreline.