We’ve been tracking a lot of numbers this March. Desperation Scores. Head Coach Scores. How conferences are performing against the kenpom spread, and how overs and unders are behaving.
A few updates, heading into the Elite Eight tonight and tomorrow:
First, we’ve got two more noteworthy head coaches to score. Mike Magpayo’s going to Fordham from UC Riverside. Kevin Willard might be going to Villanova? We ran the numbers on each.
In Magpayo’s five years at UC Riverside, he outperformed UCR’s previous historic average by 38%. That score—0.38—slots him behind Darian DeVries and Phil Martelli Jr. but ahead of Ryan Odom on Wednesday’s list. Strong.
In Willard’s eighteen years at Iona, Seton Hall, and Maryland (so far?), his weighted average is an 18% overperformance of his peers. That’s good, but it’s not as good as Odom or Fran McCaffery, let alone Will Wade.
Magpayo’s a great hire for Fordham, especially considering he was an assistant at Kyle Smith’s Columbia, meaning he knows what he’s getting into going from the Inland Empire to New York. Had Magpayo never coached on the East Coast before, we’d be more concerned.
Willard won’t necessarily be a bad hire for Villanova. A sample size of eighteen seasons is significant. Ben McCollum tops our list of head coaches on the move, but his sample is one year large. Villanova has different risk incentives than, say, Fordham. Fordham needs to swing for the fences. Villanova needs to get on base. Villanova could do better than Willard, but in the process of attempting that, Villanova could also do worse. I’m curious whom they called first, second, and third. I would hope they at least tried to land Odom and had a conversation about Wade.
Here’s that list, inclusive of Magpayo and Willard. Prior caveats about sample size and degree of difficulty still apply.
| Coach | Score |
| Ben McCollum | 0.67 |
| Ross Hodge | 0.62 |
| Will Wade | 0.49 |
| Bryan Hodgson | 0.45 |
| Phil Martelli Jr. | 0.40 |
| Darian DeVries | 0.40 |
| Mike Magpayo | 0.38 |
| Ryan Odom | 0.32 |
| Fran McCaffery | 0.26 |
| Kevin Willard | 0.18 |
| Niko Medved | 0.09 |
| Richard Pitino | 0.07 |
| Josh Pastner | -0.22 |
| Sean Miller | -0.72 |
Second, an update on conferences vs. kenpom.
Four of the five power conferences have, on average, outperformed their kenpom projections across NCAA Tournament and NIT games, at least entering tonight. Each of those four also has a winning record against the kenpom spread. This is small-sample stuff as always, but Pomeroy’s original prediction of SEC underperformance—a prediction we wholly bought ourselves—doesn’t seem to be playing out. The Big Ten’s been the class of the tournament by point differential (vs. the kenpom spread), but the SEC has the best record (vs. the kenpom spread), and the SEC’s point differential is in the middle of the power conferences, better than the Big 12’s and ACC’s. If there’s any story here, it’s that the most powerful power conferences might have improved over the course of the season relative to their mid-major and low-major peers. This provokes questions about depth and talent and training and continuity. There are a lot of possible explanations if the phenomenon is real. For right now, we’re mostly curious if it’s real. Were we capable of pausing time, we’d go run the numbers on the last ten years or so as well.
This list only includes conferences who’ve played at least seven NCAAT and NIT games. The SoCon and Big West will join that group on Tuesday. It’s an arbitrary cutoff, but these are the conferences people care about. Again: This does not account for which conferences played which conferences. The Mountain West did play three of its games against Big Ten competition.
| Conference | KP Spread W’s | KP Spread L’s | KP pt. diff. |
| Big Ten | 11.5 | 8.5 | 3.85 |
| Big East | 5 | 4 | 2.22 |
| WCC | 3.5 | 4.5 | 1.63 |
| SEC | 17.5 | 11.5 | 1.07 |
| Big 12 | 11.5 | 8.5 | 0.25 |
| AAC | 4 | 4 | 0.00 |
| ACC | 4.5 | 7.5 | -1.08 |
| MWC | 3.5 | 3.5 | -4.14 |
| A-10 | 5 | 6 | -5.64 |
Third, an update on totals.
The NIT numbers, to wet the beak:
- Overs (vs. kenpom): 19–9
- Average margin (vs. kenpom total): +3.7 points
- Median margin (vs. kenpom total): +7 points
Overtime is not included.
It’s worth noting that regulation overs went 0–4 against the kenpom total in the quarterfinals, missing by an average of twelve points. We point this out partly because we want to warn anyone thinking overs are cleaning up heading into the NIT Final Four. We point it out also to demonstrate just how much overs were cleaning up heading into the NIT quarterfinals. Is it a get–your–buckets tournament, as Greg Peterson offered on Free Hoops?
The NCAA Tournament:
- Overs (vs. kenpom): 25–30–5
- Average margin (vs. kenpom total): +0.7 points
- Median margin (vs. kenpom total): –0.5 points
Overtime, again, is not included.
Is the ball over-inflated? Maybe. If it is, other factors seem to be pushing totals back to roughly normal. Pomeroy’s pretty darn good at what he does.
**
