2025 March Madness recap - how did our Optimal Bracket do?
I hope everyone had a wonderful March Madness tournament!
While not the most exciting tournament - only 1 game-winning buzzer beater, few OT games, and no Cinderellas - the chalky all-#1 seed Final Four resulted in an amazingly competitive Finals weekend among inarguably the best four teams of the season. Houston’s unbelievable comeback win against Duke in the semi-finals will stand as one of the more incredible games in tournament history for years to come.
Optimal Bracket results
Unfortunately, the Men’s Optimal Bracket once again had a relatively down year, ending at the 81.4th percentile of all ESPN brackets, equivalent to 4th place in a typical 20-person bracket pool, likely putting you outside the prize money.
This was despite picking all Elite 8 and Final Four games correctly (not hard if you picked mostly favorites). It goes to show how much picking the last game correctly matters in tournaments with scoring that doubles per round. As a reminder, I had Houston winning it all and if they hadn’t turned over the ball four straight possessions in the Final, they likely would have won! I appreciated the plaudits on X from one of you after Houston’s improbable comeback against Duke in the semi-final game:
Generally, the process was sound: 8.8% of the crowd (per Yahoo) picked Houston to win while Nate Silver’s quantitative model suggested they had a 15.6% chance of winning, suggesting they were undervalued. The crowd had Florida winning 21.2% of the time, while Silver gave them only 13.8% chance of winning it all, suggesting they were overvalued.
That said, I’m excited about some changes I have in store for next year - during this year’s tournament I worked on a new model that simulates the tournament 1,000 times, which is technically “purer” than my current method. More to come on that next year.
In the meantime, enjoy the offseason, the 2025-2026 season, and I’ll be back in your inbox in March 2026!