Resources

This page provides a selection of my favorite soccer analytics resources, including open datasets, Python libraries, career advice blog posts, books and other learning resources.

Datasets

The StatsBomb Open Data repository contains StatsBomb event data and 360 data for a number of domestic leagues and international tournaments, including the UEFA Men’s EURO 2024, the CONMEBOL Men’s Copa América 2024, the CAF Men’s African Cup of Nations 2023, FIFA Men’s FIFA World Cup 2022, the UEFA Women’s EURO 2022, the 2015/2016 season in the top divisions of England, Spain, Germany, Italy and France, all Lionel Messi club matches, and Bayer Leverkusen’s invincible Bundesliga title win.
The Wyscout Match Event Dataset contains Wyscout event data for the FIFA Men’s FIFA World Cup 2018, the UEFA Men’s EURO 2016, and the 2017/2018 season in the top divisions of England, Spain, Germany, Italy and France. The dataset is accompanied by a research paper that provides further details on the content of the dataset.
The SkillCorner Open Data repository contains SkillCorner broadcast tracking data for nine matches between the champions and runners-up in the 2019/2020 season in the top divisions of England, Spain, Germany, Italy and France.
The Metrica Sports Sample Data contains Metrica Sports broadcast tracking data and corresponding event data for three anonymized matches.

Open-source Python libraries such as kloppy, socceraction, floodlight and statsbombpy provide functionality to load and process these open datasets.

PySport maintains an extensive overview of open-source Python projects for soccer. The following list includes my favorite Python libraries.

kloppy to load event and tracking data from multiple data providers, and to transform the data into a standardized representation that facilitates subsequent analysis.
socceraction to load event data from multiple data providers, to transform the data into the SPADL or Atomic-SPADL unified representation, and to value the individual actions performed by soccer players using VAEP, Atomic-VAEP or Expected Threat (xT).
floodlight to load event and tracking data from multiple data providers, and to compute geometric and kinematic metrics such as the stretch index and metabolic power metrics.
statsbombpy to load StatsBomb event data into a pandas DataFrame representation.
soccerdata to gather data from websites such as Club Elo, FBref, SoFIFA and WhoScored.
scraperfc to gather data from websites such as SofaScore, Transfermarkt and Understat.
soccer_xg to train and analyze expected-goals models.
penaltyblog to estimate team abilities and predict match outcomes using statistical models.
mplsoccer to produce different types of soccer visualizations.
plottable to produce visually appealing data tables.
football-data-analytics to perform different types of player, team and match analyses.

David Sumpter discussed potential pathways into soccer analytics with Javier Fernández, Sudarshan Gopaladesikan, Pascal Bauer and Fran Peralta in the Friend of Tracking episode Advice for Anyone Who Wants to Become a Football Data Scientist in March 2020.
Sam Gregory published a Getting Into Sports Analytics blog post in November 2017 and a follow-up Getting Into Sports Analytics 2.0 blog post in January 2020.
Benoit Pimpaud published a mini blog series in 2022: A Career in Football Analytics, The What, A Career in Football Analytics, The How and A Career in Football Analytics, The Reality.
Liam Henshaw published a How to Get Started in Data and the Football Industry blog post in April 2022 and a follow-up How to Become a Football Analyst blog post in October 2024.

The following two books arguably have helped shape my thoughts on soccer analytics the most.