A comprehensive, code-heavy walkthrough for building a football betting model in R. Covers the full pipeline: fetching historical match data via worldfootballR or CSV, feature engineering, fitting a Poisson goals model with attack/defense parameterization, applying the Dixon-Coles low-score correction, converting scoreline
Table of contents
SetupGet match dataFeature engineeringModel 1: Poisson goals (baseline)Model 2: Dixon–Coles adjustment (improves low scores)From scorelines to 1X2 probabilitiesOdds, implied probabilities & valueBacktest: flat stake vs KellyCalibration diagnosticsProduction: weekly pipelineFAQSort: