Modeling March Madness — Part Three

first of all: this

even though we lost an elite 8 team with duke going down, i’m still so happy it happened.

enough gloating, though. here are the results from the round of 64:

round of 64 results.png

overall, our (read: Corey’s) model correctly predicted the winner of 24 of the 32 games, the same as nate silver’s model. our judgmental selections did one better (thanks harvard!), and would have done two better if i hadn’t been so stubborn in picking my alma mater (big letdown for #nebrasketball). not awful, considering the chaos. our bracket is currently in the 91st percentile of ESPN brackets, although losing an elite 8 team will definitely hurt us. thankfully, the rest of our sweet 16 picks are still alive…

also, here are some notes from Corey regarding the model’s performance:

Of the 8 games that the model missed, the average projected win probability for the team that lost was only 60%.  By projected win probability, the three biggest upsets were:

Dayton over OSU: 23% chance
Duke over Mercer: 25% chance
Stanford over New Mexico: 33% chance

All the other games were essentially coin flips (between 45% and 49% chance of upsets).

and without further ado, here are the selections for today’s games:


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s