
Twelve Football Community
Welcome to our exclusive members' community! Here you have access to monthly courses curated by David Sumpter, renowned mathematics professor and world-leading data football scientist.
Football data with David Sumpter
-
Masterclasses
-
1. Using large language models for scouting
David goes through the methods for using Large Language models to talk about data. Ágúst breaks down scouting metrics.
-
2. Modelling player movement
David explains the mechanistic approach to modelling player movement and Pegah goes into detail with her approach to building an off-ball expected threat model.
-
3. Predicting transfer values between leagues
Building models of how players skills change as they age, what happens when they move league and how their value changes.
-
4. How we built Twelve GPT Match
David is joined by Ida and Matthias, who explain how we decided on the metrics to build into our match reports and the concept of AI agents for talking about football data.
-
5. Visualisation and using data in the media
Johan Karlsson talks to David about the design behind using data in football, focussing on communicating with different groups an. getting the processes right. And David gets to tell some anecdotes about using data on TV, including a bit about the Gary Neville metric.
-
6. History of football analysis
Welcome to our presentation on the History of Football Analysis, where we will explore the evolution of the game through the lens of data and analytic insights, highlighting how analysis has transformed football into the sport we know today.
-
7. Analysis workflows in modern football
In the fast-evolving world of football, analysis became a critical component of both tactical decision-making and long-term strategic planning. Modern analysis workflows integrate advanced technologies, from video analytics to machine learning, to provide insights on player performance, opposition analysis, and game dynamics.
-
8. Using data in set piece analysis
Jernej Flisar’s masterclass teaches how to analyze set pieces using data. He explains how data can help teams better understand and improve their performance during set-piece situations, such as corners or free kicks. By using statistics and patterns, teams can find weaknesses in their opponents and build stronger strategies to succeed during these critical moments in a game.
-
9. Manager recruitment
In this masterclass, Ágúst Pálmason Morthens, the Lead Data Scientist at Twelve Football, delves into best practices for manager recruitment in football. Drawing on his expertise in data science and football analytics, Ágúst explores how data-driven insights can revolutionize the recruitment process, enabling clubs to identify managerial candidates who align with their strategic goals.
-
10. Build your own language agent
-
11. Data skills for the modern game
-
-
Q&A and Guest Sessions
-
Q&A: The early days of Expected Goals with Michael Caley
-
Q&A: Liam Henshaw on live and data-driven scouting
Liam works on both the live scouting approach and uses data to help his decisions. He talks to David and answer’s the community’s questions about how to combine the two. And even speculates on whether he will one day be replaced by a computer!
-
Q&A with Suds Gopaladesikan
How to win Europa League with data! David talks to Suds about his work as Head of Football Intelligence at Atalanta.
-
Bonus material 1: Pegah updates on her explainable xT tool
At the weekly Twelve meeting, Pegah told us about her Streamlit tool for explaining passes. This can be interesting as a behind the scenes look at how we build models.
-
Bonus material 2: Nicole explains league comparisons
Looking at how player skills change as they move between leagues.
-
Q&A session - Set pieces with István Beregi
Our guest is István Beregi, an analyst for the Hungarian National Team with extensive experience in researching set-pieces.
He’ll be sharing insights from the top level of international football.
Don’t miss it!
-
Q&A session - Technical Recruitment & Analysis
-
-
Essentials
-
Twelve Day 2024
-
Creating match reports using words and data
Creating match reports using words and data combines narrative insights with statistical analysis to provide a comprehensive overview of a game’s key events and performance metrics.
-
Seeing beyond the stats - explainable AI in football coaching and strategy
AI in football coaching and strategy helps coaches and analysts interpret complex data, providing clear insights into player performance and tactics, making data-driven decisions more transparent and actionable.
-
Player recruitment trends in the Bundesliga
Player recruitment trends in the Bundesliga have recently focused on nurturing young talent, attracting promising prospects from around the world, and balancing financial sustainability with competitive ambition, with an increasing emphasis on data-driven scouting and player development.
-
Personalisation of Music
Personalization of music leverages algorithms and user data to tailor listening experiences, offering recommendations, playlists, and content that align with individual tastes, moods, and listening habits, creating a unique and immersive music journey for each listener.
-
Upcoming events
Theme for October is Trends in Modern Football
Thursday October 10th 20:00 (CET): Community feedback. A chance for our members to present their work for David Sumpter, David Laszlo and other Twelve data scientists. You will get feedback and suggestions from us.
Thursday October 17th 20:00 (CET): Masterclass on manager recruitment, with Ágúst Pálmason Morthens (Lead Data Scientist, Twelve Football).
Thursday October 31st 20:00 (CET): Q&A session with Max Hahn, Head of Technical Recruitment & Analysis, West Ham.
Theme for November and December is Create your Own Football Bot
Be the first to access Twelve’s Open Source release of the code used to create a bot which talks about football data. During November, David Sumpter will walk you through the approach to using Twelve GPT technology.
Monday November 18th 16:00 (CET): Football Bot Code Walkthrough. David Sumpter presents the Twelve GPT Open Source code.
Monday December 2nd 16:00 (CET): Help and feedback on the code. David Sumpter and Ida Höglund Persson provide feedback and ideas.
Monday December 16th 16:00 (CET): Community members present their bots. A chance for you to show us what you have done!
New themes will be announced for December after consultation with our members.
Research Paper Presentations
Twelve football community will be holding 1 hour bi-weekly meetings where members will have the opportunity to review and present recent research papers
Click the button for more information and to secure a spot to present.