Man City v Brighton

The bar is the model's probability; the caret on the same scale is the market's, with the margin removed. The distance between them is the disagreement.

Model fair prices, computed as 1 / probability from a Dixon-Coles score matrix. No bookmaker margin is included, and no bookmaker is quoting these. A real price will be shorter than every number here.

Computed arithmetic, not generated text. Each driver is a stored quantity: the Elo difference replayed from results, points per game over the last five matches, rolling expected goals, and each absent player's share of team expected goals.

Removing a player scales that side's expected goals by their share of recent attacking output, then rebuilds the score matrix. Every market moves together.

Written by gpt-5.6-terra from the numbers above, which it cannot add to or override.

The model assigns Man City a 49.4% win probability, with Brighton at 21.3% and the draw at 29.3%. Expected goals are 1.43 for City and 0.86 for Brighton. It puts over 2.5 goals at 40.2% and both teams to score at 44.7%. The market is more favourable to City, pricing their win at 58.5%, while assigning only 21.4% to the draw and 20.1% to Brighton. Both sides have won 4 of their last 5 matches and have 2.4 points per game. City’s recent xG figures are 11.7 for and 6.8 against; Brighton’s are 10.1 and 5.1. Across 10 head-to-head meetings, City have 5 wins, Brighton 2, and there have been 3 draws. No named absentees are listed. There are 14 further Brighton absentees whose attacking-output shares could not be measured, so their effect is unknown. The scorecard says the model does not beat the market.

Shown with the model's own record, because that is what decides how much these are worth.

Dixon-Coles bivariate Poisson fitted to Premier League results with exponential time decay. Statistical model, no language model involved.