Kickoff Lens

Data sources and methodology

Where Kickoff Lens match data comes from, how pages are built and what the numbers mean.

Sources

Fixtures, results, kick-off times, venues and tables for the 2026-27 season come from ESPN's publicly available scoreboard and standings data. Where possible we check results against a second public dataset (openfootball). The World Cup 2026 archive is built from the tournament data we published during the event, checked against public results.

How pages are updated

An automated job collects the latest data once a day, rebuilds the affected pages in all six languages and publishes them. Each page shows when its data was last updated. Times are in UTC.

What the numbers mean

Limits

We do not publish lineups or in-match statistics for club matches. Source data can be delayed or corrected; pages update on the next daily run. Previews describe the numbers; they are not predictions or betting advice.

How the match estimate works

Each upcoming match shows a statistical estimate of the chances of a home win, a draw and an away win. It is calculated only from this season's results in the same competition, using matches played before kick-off:

These inputs are combined into a score for each team. The gap between the two scores sets the split between a home and an away win; the draw share is larger when teams are close or have played few matches, and smaller when the gap is wide. Teams with few matches are pulled toward an average rating. The three percentages always add up to 100%.

Limits

The estimate does not know about injuries, suspensions, line-ups, transfers, managers, weather or motivation, and early in the season it rests on very few matches. Football is low-scoring and unpredictable: an outcome rated 50% fails half the time. This is a statistical description, not a prediction to rely on and not betting advice.

Accuracy

Since tracking began, the most likely outcome has been right in 123 of 235 matches (52%). As a retrospective check on earlier 2026-27 matches, using only data available before each kick-off, the model picked the right outcome in 181 of 374 (48%). Always picking the home team would score about 40–45% in these leagues.