Every Tuesday in the second half of the college football season, the College Football Playoff Selection Committee releases its ranking of the 25 best teams in the country. The top four teams at the end of the season will go on to a four-team playoff to determine the national champion.

CFPMachine uses historical data and machine learning techniques to predict what the committee will (or would) do each week of the season (but not what they should do!). Not only does CFPMachine consider on-the-field success, it also considers off-the-field information (program prestige, average kickoff time, conference strength, etc.). See the About page for more details.