BE GLOBAL · PLAY LOCAL
Football match predictions
Compare experimental score and outcome estimates using Elo, Poisson, Dixon–Coles and other result-based models.
Compare the models
Elo estimates relative strength from results. Poisson estimates scorelines from goal rates. Dixon–Coles weights recent scores and adjusts low-scoring outcomes where enough data exists. Glicko-2 and TrueSkill adaptations track strength and uncertainty. The balanced estimate combines these signals.
Expected goals and likely scores
The explorer shows model-derived goal expectations, likely scorelines and win, draw and loss probabilities. Monte Carlo samples from the balanced score distribution. Goal expectations from score models are different from shot-based xG.
Compare teams across borders
Pick club opponents independently by country, competition or club name. Then open the same matchup in the Simulator. Predictions are experimental and are not calibrated betting odds. Differences in data coverage and disconnected leagues can increase uncertainty.
Players and match statistics
Named squad contributors and recorded match statistics appear where available. Confirmed lineups, injuries and shot-level xG are not invented when missing.