Algo Bet is a set of step-by-step tutorials for building football models in Python: from a CSV of results to Poisson, Elo and Dixon-Coles models, data pipelines and honest backtests.
Modules
- Python Tutorials: from your first script to a working model.
- Models: Poisson, Elo, Dixon-Coles and machine learning, explained and coded.
- Data Pipelines: getting, cleaning and storing match data.
- Backtesting: testing a model without fooling yourself.
The Poisson match predictor runs the same calculation as the first tutorial in your browser.
How the tutorials are made
All code is run before it is published, and the Python version is stated. Example data is labelled as invented; real data sources are linked and used according to their terms.
The honest part
A model that looks good on past data can still lose money. Every tutorial says where its method stops working, and none promises a profitable system.
Money
Algo Bet may show advertising. It has no affiliate or paid links: see our Disclosure. Readers must be 18 or over to bet; see Responsible Gambling.