HOW IT WORKS

For every match across our covered leagues, we combine squad market value, recent form, and momentum into a statistical model. This is not a gut feeling, it is a repeatable, data-driven process.

Each match gets a confidence score based on how clearly the data points in one direction. Matches with weak or mixed signals score low, matches with strong statistical agreement score high.

We do not publish predictions for every match. We select only the top 30% by confidence, across every category, in every league. If the data is not clear, we do not pretend it is.

Every published prediction is tracked, every week, hits and misses both. Our performance page shows the real record once enough predictions have been made to report honestly.

At the start of every season, we skip the first five weeks. Without enough recent form data, a real prediction is not possible yet, so we do not pretend otherwise.