Sport, and football especially, is a one of the most popular recreation and entertainment industry in the world. For example, millions are playing football on a daily basis, while the television coverage of the leading European football leagues reaches hundreds of millions transforming these matches into worldwide events. By nature, sport is a multi-disciplinary area that involves areas such as sport science, medicine, biomechanics, nutrition science, etc. Recently, computer science is starting to have a significant impact on sports with the help of ever-increasing data on sports. These datasets allow a thorough, quantitative analysis of every facet of sports: from evaluating the individual performance of the players through the players’ involvement in the team efforts to the automatic analysis of the strategy of the teams.
The goal of our sport analytics project is to develop algorithms, methods, and procedures to quantify and predict the performance and strategies of players, teams. We believe our research efforts will have an impact on the whole sport industry including amateurs, professional teams, and the media.
Qatar, e.g., by hosting the FIFA World Cup in 2022, made a long-term commitment to sports, and its local and global development. As such, it is a perfect location to establish a global hub for sport analytics.
The benefit of a data-driven, automated analysis of football is twofold. First, football analytics is able to provide objective and quantitative evaluation of the performance of players and teams. This could support the decision making of the technical teams of football clubs. Second, analytics is able to scale up and out, i.e., a team can automatically analyze any number of matches in a short period of time without any geographical restrictions (contrary to the limited time an analyst/scouts have).
There are three main areas where football analytics could have an impact on a club: a) analyzing and preparing for the next opponent, b) improving the skills of the club’s players with customized training programs based on the players’ match-day performance, c) identifying potential new players who would fit into the style and strategy of the team.
qScout is a quantitative football player scouting system developed by QCRI. It provides advanced, cutting-edge metrics to evaluate the performance of the players, to estimate how a player would fit in a team, and to predict the future potential of football players.
Recurring Pass Patterns
Speed and Movements of Players
- L. Gyarmati, R. Stanojevic: QPass: a Merit-based Evaluation of Soccer Passes, In. Proc. of KDD Workshop on Large-Scale Sports Analytics, San Francisco, USA, August 2016. link
- L. Gyarmati, M. Hefeeda, Analyzing In-Game Movements of Soccer Players on Scale, In Proc. of MIT Sloan Sports Analytics Conference, Boston, USA, 2016. link
- L. Gyarmati, X. Anguera: Automatic Extraction of the Passing Strategies of Soccer Teams, In. Proc. of KDD Workshop on Large-Scale Sports Analytics, Sydney, Australia, August 2015. link
- L. Gyarmati, M. Hefeeda, Estimating the Maximal Speed of Soccer Players on Scale, In Proc. of Machine Learning and Data Mining for Sports Analytics Workshop, Porto, Portugal, September 2015. link
- Laszlo Gyarmati, Haewoon Kwak, and Pablo Rodriguez. Searching for a Unique Style in Soccer. In Proceedings of the ACM International Conference on Knowledge Discovery and Data Mining (KDD) workshop on Large-Scale Sports Analytics, August 2014. link
- Laszlo Gyarmati: Quantitative Player Scouting: from Datasets to Insights, FC Barcelona Sport Technology Symposium link
- Laszlo Gyarmati: Quantitative player selection: Which player fits into the team strategy? link
- MIT Technology Review: How Network Theory Is Revealing Previously Unknown Patterns in Sports link -The Economist: A different ball game link
- BBC: FC Barcelona: The science behind their success link
- New Scientist: AI football manager knows how different teams play the game link
- Lexpress: Le big data explique comment l’équipe de foot de Barcelone gagne ses matchs link
- PhysicsToday, News Picks : Network analysis reveals difference in soccer styles link
- NyTeknik: FC Barcelonas spelstil unik link
- Slate: Le FC Barcelone a bien un style de jeu unique en Europe: ces graphiques le prouvent link
- BigData: Fútbol y Data Mining link
- TechGraphs: Scientists Use Passing Data to Explain Barcelona’s Prowess link
- El Confidencial: La teoría de redes desvela los secretos del tiki-taka link