Mathematical programming as a tool for virtual soccer coaches: a case study of a fantasy sport game

Abstract

General awareness of the potential of mathematics to support sports decision making has been raised by the movie “Moneyball” through its portrayal of a real professional American baseball club that used mathematical tools to improve team performance. This article addresses the same challenge using as a test case a fantasy sport game organized by an Argentinian newspaper. Two mathematical programming models are presented that act as virtual coaches that choose a virtual team lineup for each round of the real Argentinian soccer league. The a priori design creates a competitive team for the game, while the a posteriori model determines what would have been its optimal lineup once the season is over and all the results are known. The a priori model was entered in the fantasy game, achieving results that positioned it among the highest scoring participants. Further development of such models would provide useful tools for supporting decision making by coaches or managers of real sports.