This was an independent study from my last semester at UVM, advised by Dr. Michael Tomas III and Dr. Barbara Arel. It's still a work in progress, and our goal is to submit to a journal later this year.
European basketball is different than the NBA in that teams play much different schedules, due to the nature of their scheduling format. Because of this, it’s difficult to analyze player statistics across leagues. By building an incomplete competition network from each player’s actual schedule, and using supervised learning to complete it by predicting player performance in hypothetical matchups, the competition network can be completed, allowing for player statistics to be standardized across all leagues in the future.