NBA TICKET PACKAGE OPTIMIZATION, A CASE STUDY OF THE CLEVELAND CAVALIERS

Abstract

The purpose of this study is to create a genetic algorithm to further enhance the current half season ticket packaging process in the event industry through a case study with the Cleveland Cavaliers of the National Basketball Association. The focus of this study is on the integration of machine learning and heuristic methods to simulate the human decision making currently taking place across the industry. This study will cover the methodologies being proposed for the overall, integrated approach. The methods that we cover in this study surround the tiering of events using K-Medoids and the makeup of the genetic algorithm that was implemented to solve for optimal half season packages using the Cleveland Cavaliers home schedule. Then, using the interaction between these two methodologies, we analyze the results in collaboration with domain experts from the Cavaliers. This study will show how the usage of machine learning paired with a genetic algorithm can affectively simulate and improve upon the current process for determining half season ticket packages. Furthermore, future improvements, such as the addition of predictive analytics and fan behavior are explored to supplement this work and lead to future areas of research and development.