FAQ

The percentages are modelled estimates of how likely each outcome or scoring scenario is before kick-off. They are designed to describe relative likelihood, not certainty, so a 54% angle is still a forecast that can miss in a single match. The practical use is comparison: which result, goal line, or scoring pattern is most supported by the current data profile.

Top Pick is the model's clearest supported angle for the selected fixture at that moment. It can come from the 1X2 market, O/U 2.5, or BTS, depending on where the strongest signal sits. It is best read as the model's preferred lens on the match, not as a guarantee that other markets are irrelevant.

Model Confidence is a summary signal that reflects how clearly the fixture separates into a stronger forecast shape. Higher confidence usually appears when the result or scoring profile has more separation and the surrounding context is more stable. Lower confidence does not mean the model has failed; it usually means the match is more balanced or harder to separate cleanly.

The xG cards describe the projected scoring environment rather than a promised final score. Home xG, Away xG, and Total xG help explain whether the match profile looks tight, balanced, or more open. These numbers are most useful when read together with Over/Under and BTS probabilities instead of in isolation.

Football contains substantial match-level variance. Red cards, finishing swings, set-piece variance, game-state changes, and late tactical shifts can all push a match away from its central projection. MatchProb is built to describe the most likely range of outcomes, not to remove uncertainty from the sport.

Forecasts can move when the underlying season data changes, when new matches are added, or when the scoring and form context around the teams shifts. That means a fixture checked today may not show exactly the same balance it showed weeks earlier. This is normal in a live statistical system and should be interpreted as updated context, not inconsistency for its own sake.

The historical Results section shows how the model has behaved on previously recorded fixtures within the selected scope. It is useful for understanding hit rates, calibration, and how the model tends to perform in different markets. It should be read as performance evidence, not as proof that the next fixture will behave the same way.

Yes. MatchProb works best as a structured pre-match reference alongside confirmed team news, player availability, tactical setup, schedule pressure, and broader match context. The strongest use is to compare the model view with current football information rather than treating the dashboard as a stand-alone verdict.

League and team availability depends on what data is currently installed, active, and processed in MatchProb. If a competition is not present, it usually means the required dataset is not currently available in the live scope rather than the site ignoring it arbitrarily. New leagues will be added gradually, and the long-term aim is to expand coverage across as many competitions as possible.

No. MatchProb is an informational football analytics product, not a betting service or a certainty engine. The intended use is structured interpretation of match probabilities and scoring context in a clearer format than raw numbers alone.

Use the output as one analytical input alongside your own football knowledge, team news review, and match-context judgement. Do not treat a single percentage or Top Pick as a guaranteed outcome. The most disciplined approach is to look for consistency across the result, scoring, confidence, and historical sections before drawing a conclusion.