Sports data analysis is the subject of an academic seminar
Team optimization model presented by UFRJ professor
10/4/2025

Data science entered the field on Wednesday (9) at IMPA Tech. Pedro Henrique González, professor at UFRJ and PhD in computer science from UFF, presented the seminar “Optimal team selection: an integrated approach to AI and Integer Programming”.
González presented an interdisciplinary optimization model based on sports data. The tool seeks to build an ideal soccer team using concepts from mathematics and computer science.
“It all started with the following question: Given a set of metrics, can we put together an ‘optimal’ team? In the model, each player is seen as a time series where the values are their respective scores over the rounds,” explained the professor.
The aim is to maximize the metrics using artificial intelligence tools and an integer optimization model. However, the first difficulty faced was access to data. “Sports data is very scarce. And most of what we have is closed and very expensive. A year’s worth of data from the Brazilian soccer league costs 20 thousand dollars. Imagine if we needed historical data?”
It was from there that Cartola FC, a fictional game in which people assemble their teams with real-life soccer players, came into the research. “Instead of working with real data, we decided to work with fantasy games. Cartola FC was the platform from which we started,” said González.
Student Daniel Couto thought the presentation was innovative. “I really liked the topic. Statistics and data science are areas that interest me. It was nice to see the practical applications of data and the difficulties faced in the process, such as communication between teams. As data science is new in the sports field, the professionals who work with the data are unable to communicate in the best way to the technical team that makes the final decisions,” he said.

González pointed out that some soccer clubs already have data analysts and teams responsible for processing this information, seeking to optimize the teams’ actual results. However, there is still room for growth in this area of sport.
The optimization model presented to the students was developed by Pedro Henrique González alongside other researchers at ACE Laboratory, a laboratory co-founded by the professor that works with data science applied to sport.
“ACE is a joint laboratory between the Systems and Computer Engineering Program at COPPE/UFRJ and the School of Informatics and Computing at CEFET/RJ. We work with a hybrid team of undergraduate, master’s and doctoral students from both institutions, and we carry out research on a wide range of topics in the field of sport.”
The collaboration between the institutions is bearing fruit and it is hoped that the laboratory will soon open a selection process for students from other institutions interested in the research area.
The presentation at IMPA Tech is part of the cycle of seminars organized by Professor Uéverton Souza. Every week, researchers and professors from IMPA or other institutions are invited to work on interdisciplinarity. “This exchange between different institutions is very important for strengthening research in the state of Rio de Janeiro. We’re close, but we often don’t communicate,” said González.
The activity also marked the ACE Laboratory researcher’s first visit to Port Maravelley, and he praised IMPA’s bachelor’s degree facilities. “The structure is very nice, quite different from what we’re used to seeing at other institutions.”
Read more: Financial day: meetings discuss young people’s relationship with money
