Sign In

Communications of the ACM

ACM TechNews

AI Football Manager Knows How Different Teams Play the Game

View as: Print Mobile App Share:
Atletico Madrid used few predictable passing patterns in the 2013/14 seasonand won the league that year.

Researchers used a detailed computer analysis of the 2013-2014 Spanish football league to learn about different teams' patterns of play.

Credit: Adrian Dennis/AFP/Getty

A team of researchers from the Qatar Computing Research Institute has used a detailed computer analysis of the 2013-2014 Spanish football league to better understand the various teams' different patterns of play.

The analysis involved gathering several data points, include players' pitch coordinates, the distance between them, and the time of a given pass from videos of the matches, in which more than 100,000 individual passes were made. Using this information, the system was able to identify hundreds of passing patterns used by the teams. For example, the algorithm found Barcelona and Real Madrid each had more than 150 different passing patterns, but the overall league winner for that season, Atletico Madrid, had only 31 different patterns.

The hope is that this sort of system could help coaches better understand their opponents' strategies.

There are already several companies offering such services to a number of different sports teams. A Finnish company called SportsIQ is developing a real-time monitoring and analysis system for hockey that involves wearable devices and a connected hockey puck. Predictive play services have already been developed for and are being used by teams in basketball and U.S. football.

From New Scientist
View Full Article


Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account