Randomness, uncertainty, and the optimal college football championship tournament size

Every year, there is a popular debate over how many teams should take part in the NCAA’s FBS-level college football championship tournament, and especially whether it be expanded from 4 to 8 or even 12. The inherent tradeoff that larger higher probability true best team included (“validity”), but lower will avoid being upset win tournament (“effectiveness”). Using simulation based on empirically-derived estimates of ability measure quality amount randomness each game, we show effect expanding could very small, an effectiveness decrease only 2-3% while increasing validity by 1-4%, 7-team provides slightly better tradeoffs. A 12-team would 5-6%.