Skip to content

Forecasting fantasy football with machine learning. Training a variety of models to predict player performance in the 2018/19 season - and then using that to build the ideal team for the next game week. Written in R.

Notifications You must be signed in to change notification settings

crkershaw/fantasy-football-forecaster

Repository files navigation

fantasy-football-forecaster

forecast team

You can read a full explanation of the scripts and the outputs here:

https://charliekershaw.com/fantasy-football-forecaster

These scripts scrape data on the Premier League Fantasy Football, aggregate it, then forecast how many points a player will score from their recent history. To do this it uses a variety of models, from linear regression to machine learning.

There are four main scripts:

  1. Data collection
  2. Data aggregation
  3. Model training
  4. Forecasting the ideal team (ideal team creation and graphing)

Note: all data is collected and modelled in 2019 - if trying to replicate this using these scripts, note that the FPL API and libraries may have changed since.

About

Forecasting fantasy football with machine learning. Training a variety of models to predict player performance in the 2018/19 season - and then using that to build the ideal team for the next game week. Written in R.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages