Construction and Application of an MDP-Based Dynamic Decision Framework for Professional Sports Team Management

Abstract

This study discusses the conflict between sports achievements and money of professional sports teams. Therefore, we construct a decision-making framework based on markov decision processes, taking WNBA as an example. The framework uses three-dimensional state space (competitiveness, financial status and market position) and reward function with multiple objectives. It combines three modules: player acquisition, fare optimization and injury management, TOPSIS is used to evaluate players, dynamic pricing algorithm and random forest is used to predict injuries. It also puts forward a three-stage strategic theory (reconstruction, champion containment and stable profitability). The results show that the framework increases the average team profit by 52.4%, the team value by 49.0%, the ticket income by 18%, the expected injuries by 15.6%, and the accuracy of evaluating players by 80%. These well balances sports performance and economic benefits, and provides decision support tools for professional sports teams.