ESPN’s Peter Keating lists a few upset possibilities…
No. 13 Portland State (17.6 percent) vs. No. 4 Xavier (62.7 percent)
Portland State, whose adjusted statistics are only about as impressive as Iowa State’s or St. John’s, isn’t the ideal killer. But Xavier is a textbook example of a giant waiting to be slain by just about anyone: The Musketeers turn the ball over considerably more often than they generate turnovers (21.9 percent vs.18.9 percent of possessions). They are heavily reliant on making more free throws than opponents (9.1 FT margin per 100 possessions), which is always a concern in a one-and-done scenario. And they don’t score enough to blow past their flaws (adjusted offensive efficiency of 111.3). Considering everything working against Xavier, and the fact that Portland State at least has a puncher’s chance, according to our model, this is as good an upset pick as you’ll find in the first round.
No. 11 Temple (32.6 percent) vs. No. 6 Arizona State (27.9 percent)
The Sun Devils play sloooow (61.2 possessions per game, by far the fewest among giants), which means James Harden (20.8 ppg) is even more impressive than he looks. But ASU doesn’t grab offensive rebounds (30.6 percent of possessions, 250th in the country). And they give opponents too many open looks from 3-point land (27.4 percent of all field goals). Temple, on the other hand, keeps foes off the offensive glass, doesn’t turn the ball over and doesn’t rely on free throws to outscore opponents. Also, our model can’t distinguish between dominating scorers who can carry a team through the postseason and those who can be stopped as soon as they run into a good opponent, but senior guard Dionte Christmas certainly was the former in the Atlantic 10 tournament.
I don’t advocate picking either of these underdogs. Arizona State is playing good basketball and I think Xavier will have enough to get past the first two rounds. But it’s interesting to see that someone has developed a formula to predict these upsets. We’ll see how it fares.
