Modeling March Madness — Part Four (By Corey Berg)

First thing’s first: What a set of games yesterday!  The model got 5 out of the 8 and our judgmental selections got 6 out of the 8.

The model only gave Dayton a 21% chance of beating Syracuse and UConn a 25% chance of beating Villanova.

Other notes:

• Our judgmental selection of Oregon nearly came through, as Oregon managed to lose after holding a 12-point halftime lead over Wisconsin
• The model did give Harvard much more of a chance than many “experts” did and Harvard nearly pulled it off.  Great effort by the Crimson!

We’re in full upset mode today, as the model suggests both Stephen F. Austin and Mercer have great chances to continue their runs!

kyle’s notes: our bracket took a big hit yesterday, as we lost another elite 8 team (Oregon)…and early on in sunday’s games, we’re in danger of losing another with Kansas struggling late in the second half (although it’s worth noting that the model has this Kansas/Stanford game being pretty close to a toss-up). our only consolation is that hopefully everyone else’s bracket is blowing up too…