Loading trust dashboard
Preparing probability logs, settlement summaries, and match context.
This page explains how predictions are logged, how results are settled, why partial or no-data states exist, and why the product is intentionally positioned outside gambling workflows.
We publish settled performance transparently through our probability log system. Match outcomes, hit rate, and ROI are based on settled records only rather than manually curated examples.
Coverage depends on the competitions currently supported by our production data pipeline. You can browse the active league directory directly from the leagues page.
Probabilities are generated by our model pipeline before kickoff and stored as append-only public logs. The UI only reads these logs; it does not generate forecasts on the fly.
We prefer closer match windows so lineup context, injuries, and odds snapshots have a better chance of being available and auditable before kickoff.
No. Predictions are stored as append-only logs. Results are stored separately, which makes it possible to audit whether a prediction existed before kickoff.
Partial means a prediction exists, but one or more supporting datasets such as odds snapshots, lineups, or event feeds were not available at query time.
We intentionally return no-data states instead of fabricating values. If there are no settled predictions, ROI and hit rate remain null rather than being shown as misleading zeroes.
We rely on provider-backed football data feeds for fixtures, standings, odds, injuries, and live support data. Our own probability logs and settled results layer sit on top of those sources.
Different datasets update on different schedules. Match and result data refresh more frequently than slower-moving league or educational content modules.
No. This site is positioned as a football intelligence research product. It does not process wagers and it does not provide gambling functionality.