METHODOLOGY
How match estimates work
The current prototype uses a Poisson score model. It estimates each team’s expected goals from the home team’s scoring rate, away team’s conceding rate, away team’s scoring rate, home team’s conceding rate, and a small home advantage. It then calculates likely scorelines and sums their probabilities into match outcomes.
Markets shown
- Home win (1): home team scores more goals.
- Away win (2): away team scores more goals.
- 1X + BTTS Yes: home win or draw, with both teams scoring.
- X2 + BTTS Yes: away win or draw, with both teams scoring.
- 12 + BTTS Yes: either team wins, with both teams scoring; a draw does not qualify.
The combined markets are calculated directly from scoreline probabilities, so the conditions must all occur together. Percentages are model estimates and may not add to 100 across these markets because they overlap.
Limits
This simple prototype does not yet account for lineups, injuries, schedule congestion, red cards, or reliable league-strength adjustments. Results depend on the accuracy and freshness of the input data. Forecasts are estimates, not guarantees.