1. Scheduling the national soccer league.

Before 2019, in the Russian (Soccer) Premier League (the major domestic soccer club tournament) the tournament schedule had been created manually, according to the so-called Berger scheme. However, the Berger scheme does not allow taking into account various restrictions, including league and clubs’ preferences, climate, broadcaster’s preferences, international matchdays, etc. Jointly with my colleagues, Gleb Vasiliev and Arseniy Stolyarov, we have developed a computer algorithm based on the gradient descent method that allows finding the schedule that minimizes the penalty function. `The clever schedule’ that we first made in 2019, got a warm reception from the league and the clubs, and after a one-year contract for the 2019/20 season, it was prolonged for the next 4 years.

2. A recommendation system for soccer referees assignments.
In 2020, our Laboratory of Sports Studies signed an agreement of cooperation with the national Football Union, which gave a start to developing a computer system that helps the Referee Department to assign referees. The recommendation system is based on the Gale-Shapley algorithm which allows finding optimal (in some sense) matching of referees to games.

3. A model for predicting a transfer market value.
One of the leading football clubs asked us to develop a predictive model for future transfer market value of young players. The problem is that for younger players there are no expert estimations for the transfer market value. Therefore, a club has financial risks due to information asymmetry. We collected hundreds of objective performance variables in order to take into account all relevant available information. The project was successfully finished in 2022.

Economics, Game Theory, Computer Science, Sports Studies

Applied Projects