Inside soccer’s data renaissance

2026-06-11 · Source: MIT Technology Review · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics, Sports Analytics · Depth: Advanced, extended

Summary

Jesse Davis's Sports Analytics Lab at KU Leuven is at the forefront of soccer's data revolution, applying advanced machine learning to uncover tactical insights. A notable finding, presented in a 2024 paper titled "Boot it", demonstrates the strategic value of intentionally kicking the ball out of bounds on the opponent's side from the middle third of the pitch. This seemingly counterintuitive move, simulated using tree ensemble models on a dataset of over 1.4 million passes and 60,000 throw-ins (partly from the 2022 World Cup), can position a team within 10 actions of scoring. The lab's work extends to developing open-source analytics tools like VAEP and xG models, which see thousands of monthly downloads. Furthermore, Davis's team is actively researching methods to standardize in-game data using transformer neural networks, aiming to automate the currently manual and time-consuming process of tactical annotation, which can take up to six hours per game. Their research significantly impacts professional clubs across Europe, including Royal Sporting Club Anderlecht and Club Brugge KV.

Key takeaway

For soccer analysts and data scientists evaluating tactical decisions, consider integrating counter-intuitive strategies like the "Boot It" approach into your models. Your analysis should quantify the trade-off between possession and positional advantage, especially in the defensive third. Implement machine learning frameworks to simulate and compare alternative actions, such as long clearances followed by aggressive pressing. This can reveal hidden value, potentially increasing expected goals per season, and inform more dynamic, data-driven coaching strategies.

Key insights

Machine learning reveals counter-intuitive soccer strategies, like "Boot It," by analyzing complex game data.

Principles

• Data analysis can uncover hidden tactical advantages.
• Sacrificing possession can create positional advantage.
• Standardizing data improves analytical efficiency.

Method

A framework uses an XGBoost model to compare observed backward passes with simulated long clearances, evaluating the probability of scoring minus conceding within 10 actions based on game state features.

In practice

• Use "Boot It" strategy in defensive third to gain advantage.
• Apply pressure after long clearances to maximize benefit.
• Employ open-source VAEP and xG models for game analysis.

Topics

• Sports Analytics
• Machine Learning
• Soccer Tactics
• Data Standardization
• XGBoost
• Open-Source Tools
• Transformer Models

Best for: NLP Engineer, AI Scientist, Data Scientist, Research Scientist

Related on AIssential

• Can Machine Learning Predict the World Cup? — Predicting football outcomes requires robust feature engineering, as complex models offer minimal gains over simpler regression approaches.
• Athletes’ performance and injury management in sports training using association rules and data mining techniques — Association rule mining reveals interpretable patterns in sports data, linking training factors to performance and injury outcomes.
• Robustness of Refugee-Matching Gains to Off-Policy Evaluation Choices — Algorithmic refugee matching gains are robust across diverse off-policy evaluation methods, confirming original findings.
• Predicting IPL Match Outcomes Using Powerplay Scores and Machine Learning — Predicting IPL match winners post-second innings powerplay is feasible using a machine learning model focused on early game metrics.
• A Universal Dense Football Event Representation Based on TabTransformer — A TabTransformer learns dense, semantic representations of heterogeneous football event data, improving downstream task performance.
• Luis Serrano + Josh Starmer Q&A Livestream!!! — Persistent effort and foundational understanding are key to navigating the fast-evolving AI landscape.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by MIT Technology Review.