What RadarOnze analyses
RadarOnze is a proprietary football analysis platform. Its database connects competitions, seasons, clubs, fixtures, line-ups, events, team statistics, players and availability to organise a single reading of each context.
Pages use only records available in the database. When a competition, fixture or metric has no valid sample, the interface states “Insufficient data” instead of filling the gap with an estimate.
Data source and processing
Structured data is obtained from API-Football and persisted by RadarOnze. Controlled routines reconcile competition, season, club, player and fixture identifiers before information is presented publicly.
Processing keeps each competition and season separate. Completed fixtures feed historical samples; future fixtures remain in the calendar; status, line-up and statistical updates replace records only when the new payload is valid.
Averages, frequencies and scopes
Averages divide a recorded total by the number of valid matches in the scope. Frequencies show both the number of occurrences and the full sample size — for example, 7 of 10 matches, or 70%.
Five-, ten- and twenty-match, season, home and away scopes are independent. A 70% frequency across the latest ten matches describes seven occurrences in that sample; it does not guarantee a 70% chance in the next fixture.
Probabilities and models
RadarOnze probabilities are produced by versioned statistical models, calibrated and assessed separately for each competition. They combine only predefined variables and information available before each fixture kicks off.
An estimated probability is a model output; a historical frequency is an observed count. The measures may be close, but answer different questions. Odds, fair odds, edge and EV, when shown, are materialised through deterministic RadarOnze calculations — not generated copy.
Limitations and responsible use
Football contains randomness, line-up changes, physical context, refereeing decisions and rare events that no model captures completely. Small samples, partial coverage and competition changes also reduce the confidence of a reading.
Past results do not guarantee future performance. Statistics, probabilities and analysis are informational and educational and do not constitute betting advice or a promise of returns.
Updates and review
Scheduled routines update calendars, results and the applicable statistical sets. RadarOnze preserves the last valid record when a source fails and does not interpret incomplete responses as deletions or sporting results.
Models and rules keep identifiable versions. Coverage or status corrections are applied centrally and checked before publication, while histories and calculations remain separated by competition.