Football intelligence. Made simple.
Discover upcoming matches, explore statistical predictions, compare available bookmaker prices and review model performance — all in one place.
Every match has a story.
Explore the numbers.
Study 23 football markets, inspect model probabilities and fair odds, compare recorded bookmaker prices, and follow how predictions perform over time.
Intelligence at a glance
Cached dataset · live statistics appear after your first sync
On the horizon / Next 7 days
23 markets. One statistical baseline.
Regularised Poisson scoring estimates power the public scanner across 23 markets. Review sample-quality flags, model-implied fair odds and walk-forward research. These estimates are not guaranteed or independently calibrated.
Every league. One home.
Discover supported countries, competitions and seasons across the world.
Countries
Leagues & competitions
Goals Scanner.
All seven Over/Under markets. Research probabilities, fair prices and data limitations before a decision.
How are these numbers calculated?Understand Expected, Probability, Fair Odds, Book Odds and Model EV
Home estimated goals + away estimated goals. Our baseline measures prior completed matches in this league, discounts older results (180-day half-life), estimates each team's scoring and conceding strength at its venue, adjusts for the opponent, and shrinks small samples toward the league average (7 weighted matches). Not shot-based xG.
The probability of the selected goals market (for example, Over 2.5 means 3 or more goals), calculated from a Poisson distribution using Expected as the mean. It changes when you choose another market.
1 ÷ model probability. Example: 48.4% → 1 ÷ 0.484 = 2.07. This is a model-derived price, not a bookmaker quotation.
The best stored bookmaker decimal odds for that fixture and market, where available. Quotes may be stale: confirm the current price and bookmaker before relying on it.
(Probability × Book odds − 1) × 100%. A positive number means the bookmaker price is higher than our model's estimated fair price; it is not evidence of proven profitability. A dash means there is no usable quote.
Research-only baseline. These estimates are uncalibrated and do not account for every injury, lineup or match circumstance. Click Explore to see each team's expected-goals split and historical sample sizes.
| FIXTURE | MARKET | COMPETITION | DATA QUALITY | Match details | |||||
|---|---|---|---|---|---|---|---|---|---|
| Choose a competition and run the scan. | |||||||||
Research Lab.
Walk-forward performance metrics. No future match data is used in earlier forecasts.
Every model deserves a reality check.
Select a league with imported historical seasons and run a walk-forward test to examine the Brier score and prediction calibration for each market.