Resetting After the Round of 32
The opening weekend this year didn’t provide the same fireworks and Cinderella stories of years past. That doesn’t mean it has been any less enjoyable…okay it has been slightly less enjoyable. The Ducks are the only double digit seed left to root for this year. Duke dodged a 7’6″ Tacko Fall’s sized bullet. Millions of brackets would have been busted if UCF’s buzzer beater was a single inch lower on the rim as it circled round and round. But unfortunately we are stuck with higher seeds continuing to win across the board. Despite the lack of upsets, let’s check in how we performed so far.
Bracket Performance
You can read about our methodology outlined in Part 1 and our follow-up in Part 2.
The model avoided any significant bracket-busting upsets entering the Round of 32. Mississippi State and Marquette were the only real blemishes and even then they were pretty minor. We are set up pretty nicely for the rest of the tournament.
The model continued to like Oregon and somewhat surprisingly Murray State but went chalk outside of that.
The model missed on Murray State but they perfectly encapsulate one-man teams. They can certainly beat anyone, but are also prone to higher volatility. The model predicts nothing but chalk for the Sweet 16 and only 2 “upsets” the rest of the way. That fits with the narrative about how top-heavy teams this season.
The model figures the right side of the bracket figures to include the closest games but Michigan-Texas Tech may be the game of the weekend.
Monte Carlo Simulation
One of the trend we noticed with the Monte Carlo simulations is the bump it gives underdogs relative to the model above. Looking back at past years this made sense with more upsets. Obviously we have not seen much of that so it will be interesting to watch going forward.
Day 3 and 4 posted pretty solid results with the simulation correctly predicting 13 of the 16 matchups.
As expected, Gonzaga’s chances are creeping higher and higher. The Zags and Blue Devils are the clear favorites right now. We are looking forward to some heavyweight blows back and forth between the juggernauts, hopefully generating real drama and fireworks for us all to enjoy in the final four. Houston is an interesting team to watch though. The simulation has continually liked their chances IF they can get past Kentucky. For comparison, both Houston and Purdue reached the Elite 8 in 46% of simulations, however Houston wins the title 8.38% of the time vs a paltry 3.8% for Purdue. Meanwhile, Virginia looks poised to survive the South region but falls a bit flat the rest of the way.
Look for a detailed breakdown of the Sweet 16 matchups. Looks like we could see some real nail-biters.
