Poisson correct score
Enter each team's expected goals (xG) and the Poisson model returns the most likely scorelines, plus win/draw, over 2.5 and both-teams-to-score probabilities. The same maths the pros use for score prediction.
| Score | Probability |
|---|
What is a Poisson correct score model?
Poisson is the standard probability distribution for counting independent events in a fixed window, and goals in a football match fit that shape well. Given an expected goal rate λ, the chance of exactly k goals is P(k; λ) = λk e−λ ⁄ k!.
Expected goals (xG) is your estimate of how many goals a side should score, based on the quality of chances rather than what happened to go in. We run Poisson separately for each team's xG, then combine the two distributions to get the probability of every scoreline, and from those, every market.
Why it's the benchmark: from two numbers it produces a full, internally consistent set of probabilities: correct scores, 1X2, over/under, BTTS, all from the same model. That's why it underpins most public score predictors.
The caveat: basic Poisson assumes the two teams' goals are independent and that scoring rate is constant. Real matches have red cards, game-state effects and correlation, so a raw model slightly under-rates draws and 0-0s. Treat it as a strong baseline, not gospel.
Built by the team behind a publicly verified football model, every pick logged before kick-off, 72% hit rate, losses included.
The daily sheet: every leg, with its chance of landing
Every leg on the card ranked by how likely it is to land, the accas built from that list, the facts behind each leg, and the calls the desk refused, with the reason. Locked before kick-off, graded in public, wins and losses alike.