EURO 2024: How player tracking works technically

Contents

"The ball is round and a game lasts 90 minutes," said Sepp Herberger in 1954. Nowadays, the game lasts at least 100 minutes because it is constantly interrupted. At least the ball is still round, but apart from that it no longer has much in common with the rustic playing equipment of Herberger's time: the European Championship ball "Football Love" is a high-tech toy with a sensor core, while the stadium has mutated into a gigantic 3D scanner. Together, the two collect vast amounts of match data that directly influence the match and referee decisions and enable real-time evaluations and visualizations for spectators. The system records the positions of players and the ball and prepares the image material relevant to the decision graphically: with automatically generated position lines and markings on the relevant parts of the players' bodies. At the same time, the data flows into a pool from which associations and clubs draw the information for their match and performance analyses. It also forms the basis for lucrative games such as FIFA.

The combination of the sensor ball ("Al Rihla") and camera technology celebrated its premiere at the 2022 World Cup in Qatar: its main purpose was to speed up the work of the video assistant referee (VAR) in detecting offside. Because the ball records exactly when a player has touched it, the assistant referees can also find the disputed scenes in the recorded video footage more quickly. The sometimes inefficient search for the decisive frames and the resulting slow interruptions to the game should be significantly reduced, so the promise - from an average of 70 to 25 seconds.

The technology is now being used for the first time at a European Football Championship. The system has not changed fundamentally since its World Cup premiere; the ball developed by Adidas/Kinexon and the AI-supported camera system for player tracking remain the same. The sensor ball definitely made its grand entrance in the World Cup match between Portugal and Uruguay, when it proved that it wasn't Cristiano Ronaldo who scored the opening goal, but his teammate Bruno Fernandes. While the video images were too blurred, the sensor data revealed it: Ronaldo could not have touched the ball with his head at all because the built-in sensor had not registered any shock. After Fernandes released the ball, the recorded measurement curve quickly flattened out and did not stop until the ball hit the ground.

c't compact

• The Video Assistant Referee tracks the player and ball and sounds the alarm if offside is suspected.
• Within seconds, a virtual simulation of the match scene is ready with guidelines to help the referee make a decision.
• AI helps to assign players' body extremities in confusing game situations.

Data analysis and processing are much more complex when it comes to offside detection. Using machine-enhanced video images, the human assistants in the referee catacombs should be able to assess as quickly as possible whether an illegal offside position actually exists. So that the audience in the stadium can also understand the decision, the position and camera data calculated by the system is then visualized in 3D so that the scene can be shown and explained directly on the stadium monitors from different angles. Because the technology can only identify potentially suspicious player constellations, but cannot interpret them according to the rules, FIFA refers to this as semi-automated offside technology (SAOT).

This cannot be achieved with yesterday's hardware, which is why venues and even the ball had to be properly upgraded: The World Cup ball "Al Rihla" and its European Championship successor "Soccer Love" conceal several sensors. The pitch is surrounded by 12 to 24 antennas that receive their signals. And at lofty heights under the stadium roof, a ring of 8 to 32 video cameras monitors the action. They collect precise position data for real-time player tracking, which a few clever algorithms evaluate and ultimately convert into virtual scenes.

This turns soccer arenas into gigantic 3D scanners – and gigantic production sites for match and player data, which soccer associations can now monetize themselves: with the media, clubs, player consultants, for scouting and training management as well as lucrative secondary exploitation in interactive games.