Skip to content

Introduction

pyrf is a quantitative analytics library for football betting. It provides a pipeline from raw statistical models through to market-implied pricing:

  1. Model match outcomes with statistical distributions
  2. Price betting markets using payoff surfaces
  3. Calibrate models from live market quotes
  4. Identify value by comparing model-implied odds to market odds

Architecture

The library is built around a grid-based computation model. A GoalGrid represents all possible scorelines (e.g., 0-0 through 10-10). Every distribution, payoff, and pricing calculation operates on this grid, making the maths transparent and composable.

GoalGrid → Distribution.pmf() → PayoffSurface → implied_odds() → Quote
                                                                     ↑
                                                        Market data via calibrate()

Modules

Module Purpose
pyrf.core Goal grids — the foundation data structure
pyrf.models Statistical distributions (Poisson, Skellam, Dixon-Coles)
pyrf.market Betting lines, odds, margins, and quotes
pyrf.payoffs Payoff surfaces for Asian Handicap and Total Goals
pyrf_data FootyIQ data client for historical match data (separate package)
pyrf.utilities Helper functions

See the Features section for detailed guides on each module.