Predictive Football Analysis

In the fast-paced landscape of sports, technology, and algorithms are becoming pivotal forces, creating a new era of performance and allowing athletes to rise to higher levels. Offenses and defenses are constantly evolving and transforming, and with the help of technology, teams are becoming better than ever, and coaching is trickier than ever. I present two algorithms to approach coaching sports. In the current football arena, accurately predicting the next play has been a longstanding challenge. Therefore, I created an optimized Random Forest Model (RFM) to anticipate what play a team might run, enabling coaches to strategize and teams to create better defenses. I also developed a second network using Deep Double Q-Learning (DDQN) to simulate an offense that a coach would call for his players.