Open-source statistical model (Elo + Dixon-Coles + Monte Carlo) for the 2026 FIFA World Cup — now running year-round on club football at https://theopenmodel.com
Models for estimating football (soccer) team-strength
2026 FIFA World Cup prediction research system — 7-component fusion (Dixon-Coles, Elo, Pi-Rating, Weibull, Market, Stacking, Conformal), auditable walk-forward backtesting, bilingual reports
竞彩足球智能预测 Skill：Dixon-Coles、赔率价值、赛前情报、回测与离线 HTML 报告
My aim to develop a model that predicts the scores of football matches. My code (python) implements various machine learning algorithms to analyze team and player statistics, as well as historical match data to make informed predictions.
A quant forecasting system + live API bot for Jump Trading's Probability Cup 2026 — 89th/4,013 (top ~2%). Brier-optimal probabilities, de-vigged odds, Dixon-Coles, calibration loop, and an honest post-mortem.
A stats model (pkg) for the NHL
⚽ Falsifiable, self-calibrating real-time probability engine for World Cup 2026 · Dixon-Coles double-Poisson + Bayesian credible intervals + in-play live W/D/L · personal educational open-source (not betting advice)
Stats-first football betting research: Dixon-Coles + lineup-aware models, walk-forward backtested for real edge vs. bookmaker closing odds.
Held-out FIFA World Cup prediction model: Dixon-Coles goals, market-value prior, Elo baseline, daily-locked picks, Monte Carlo title odds.
Calibrated football match prediction & soccer betting model — Dixon-Coles, Poisson, Elo and a LightGBM ensemble with leakage-safe features, walk-forward backtesting, and a paper-money web UI.
Pipeline de Machine Learning para previsão de resultados, gols e mercados de apostas do Campeonato Brasileiro Série A. XGBoost + Poisson Bivariado + Calibração Isotônica.
2026 World Cup predictor: Dixon-Coles + XGBoost + Neural Net + Hierarchical Bayesian (Stan/brms), stacked via meta-learner. R API (plumber) + single-file HTML dashboard with live market odds. Walk-forward, temporal split, isotonic calibration, stacking beats simple average.
World Cup predictions based on a Bayesian Dixon–Coles model, integrating Elo priors, dual GAS dynamic strength estimation, and a gradient-boosting ensemble implemented via log pooling. The system covers 49,000 Kaggle international matches and 336 teams, and is served via FastAPI.
Football probabilities sealed and hashed before kickoff, then scored against Pinnacle's closing line. The live record is public — including when it's bad. Runs proofodds.com
Reproducible backtest of the Elo + Poisson model behind footytips.io — all 22 World Cups (1930-2022), 964 matches, 56.6% W/D/L hit rate.
Premier League title-race predictor + FPL recommender. Dixon-Coles + Monte Carlo + ILP. TUI / web / CLI.
2026 FIFA World Cup match predictor — Dixon-Coles + Elo model with live backtesting. 基于Dixon-Coles双泊松与动态Elo的世界杯逐场胜平负预测模型,含三届回测验证与实时滚动预测看板。
An AI that prices World Cup matches better than the market — and bets real money on the gap. +202%.
To associate your repository with the dixon-coles topic, visit your repo's landing page and select "manage topics."