xG Calculator: Correct Score Matrix & Match Odds
Enter the expected goals for each team. The Poisson model turns them into home, draw and away chances, BTTS, over/under goals and a 0-5 correct score matrix, with fair decimal odds for every outcome. Works with xG from any source.
Free xG calculator: enter each team's expected goals to get win, draw, BTTS, over/under and correct score probabilities with fair odds (Poisson model).
Frequently asked questions
How does this xG calculator work?
It treats each team's goals as a Poisson distribution with the expected goals you enter as the average. Multiplying the two distributions gives the probability of every scoreline, and adding scorelines together gives home win, draw, away win, BTTS and over/under probabilities.
What is the correct score prediction formula?
P(home scores h and away scores a) = P(h | home xG) × P(a | away xG), where P(k | λ) = e^-λ × λ^k / k!. For example, with 1.6 and 1.1 expected goals, 1-1 has a probability of about 11.8%, which is fair odds of about 8.45.
Where do I get expected goals numbers for a match?
Use team xG averages from a free source such as Understat, FotMob or SofaScore, or a bookmaker's team goals lines. StatPair does not provide xG. If you only have goals data, the helper on this page estimates expected goals from attack strength, opponent defence strength and the league average.
What are fair odds?
Fair odds are 1 divided by the probability, with no bookmaker margin. If the calculator gives an outcome 40% (fair odds 2.50) and a bookmaker offers 2.80, the price is higher than your estimate of fair value. It is only value if your expected goals inputs are accurate.
How accurate is the Poisson model for football?
It is a good first estimate but not perfect. It assumes the two teams' goals are independent, which slightly underestimates low-scoring draws such as 0-0 and 1-1. Models such as Dixon-Coles adjust for this. Your inputs matter more than the formula: poor expected goals estimates give poor probabilities.
Is this an xGOT calculator?
No. xGOT (expected goals on target) is a post-shot measure of where shots ended up, so it describes past matches. For pre-match probabilities, use pre-shot xG averages as inputs. Our xG vs xGOT guide explains the difference.