College Football
There are some CFB projects I’ve worked on using data from collegefootballdata.com.
Disclaimer: Despite spending more years at Florida State (and being on campus for the 2013 championship season) while I worked on my PhD, my own college football loyalties lie with my undergraduate university, Texas A&M.
Simulating the College Football Playoff
The teams have been chosen; who is most likely to win the 2024 College Football Playoff? I simulate the entirety of the playoffs thousands of times using my college football team efficiency model. Who is most likely to win the national championship? What are the most likely matchups?
Predicting College Football Games
I use estimates of team efficiency on offense and defense and my game prediction model to simulate upcoming college football games. In addition to estimating the spread and probability of victory for the home team, I rate the expected quality/competitiveness of each game to find the most interesting matchups for the week.
Ranking College Football Teams
Who are the best college football teams right now? How have they changed during a season? I use my model of play by data to estimate opponent adjusted team efficiency after each week of the college football season. I convert these estimates into a team power score, which measures a team’s expected margin of victory against an average opponent. I monitor these power scores to track how teams are faring based on their on field performance.
Simulating College Football Seasons
Which teams are most likely to win their conference? Make the playoff? Win the national championship?
2022
Team Predictions // Game Predictions
Team Profile Examples
I simulate the remainder of the college football season 1000 times after every week of the regular season. These simulations come from an (adjusted) Elo model that I developed using historical game and play by data - the current iteration of the model uses historical game outcomes, team offensive and defensive efficiency metrics, and recruiting composites in order to estimate a team’s relative rating compared to other teams.