Through machine learning, the researchers Carlos Soares, Cláudio Sá, Ana Fernandes, Pedro Abreu, Fábio Pinto, Tiago Cunha and João Correia Pinto (CESE), together with Vítor Cerqueira (LIAAD) and José Carlos Coutinho, master student in Informatics and Computing Engineering at FEUP, are calculating the probability of certain events occurring, such as shots or goals, minutes before they happen.

The analysis of the matches is done through techniques of data mining that monitor the athletes’ trajectories and identify patterns that relate their behaviour with the events that occurred during the match and with the final score.

The identification of patterns on the athletes’ behaviour and how these relate to some moments of the game, such as shots, fouls or goals, can help coaches improve the performance of their players.

“The information acquired will be useful in identifying differences between the winning teams and the remaining ones”, explains Carlos Soares, also a professor at the Faculty of Engineering of the University of Porto (FEUP).

According to Cláudio Sá, also an INESC TEC researcher, the advantage of this tool relies on the fact that it can “analyse, in real-time, the position and the behaviour of several players at the same time, as well as identify patterns and measure to what extent they are reliable or not”, something that to the naked eye is not possible.

The data is gathered by the cameras installed on the football fields of the Euro, which started on the 16^th of July and will end on the 6^th of August and where the Portuguese team is participating. The data will be provided to the Royal Dutch Football Association.

The expertise acquired from this project may be applied to other competitions and to other sports.

In addition to INESC TEC and FEUP, the Leiden University, in the Netherlands (leader), also participates in the project.

The researchers mentioned in this news piece are associated with INESC TEC and UP-FEUP.