On Sunday, the NCAA’s selection committee revealed the bracket for its 2017 Division-I Men’s Basketball Championship. While many praised the committee for getting the “right” 68 teams in the field, there was wide concern over seeding.
The most troubling seed appears to be Wichita State. When the committee revealed that the Shockers, 30-4 and the Missouri Valley Conference champions, were a 10-seed, it was a message that the committee does not care about your advanced metrics, nor will they entertain arguments that there are some teams in smaller conferences that are good enough to be deserving of quality seeds. It announced that the RPI is the be-all and end-all of tournament resumés.
I mean, that’s what you should have taken away, if you weren’t going to dig any deeper. Wichita State was 32nd in RPI, meaning that they really were only two spots under-seeded if it was based solely on RPI, which makes the 10-seed seem not as bad as if you were looking at KenPom, which has Wichita State listed as a top-10 team in the country. But then, if you let your eyes gaze just one spot lower, you’d see that a team from the same Missouri Valley Conference, Illinois State, which lost just two conference games, both to Wichita State (one in the regular season and one in the conference tournament), was ranked 33rd in the RPI. Alas, Illinois State got no bid.
So, if it’s not advanced metrics and it’s not the RPI, what is it?
Perhaps the answer is a little more impure than we’d all like to think this process to pick the participants of the greatest event in all of amateur athletics. We’d like to think that it’s “the eye test” or “gut feelings.” We’d like to think it’s comparing the best wins and the worst losses.
But the truth of the matter is that the selection committee is flawed in it’s composition.
The selection committee is a 10-member group made up of athletic directors from various institutions. This year, the committee body was comprised of the following nine members:
- Mitch Barnhart, University of Kentucky
- Janet Cone, University of North Carolina at Asheville
- Tom Holmoe, Brigham Young University
- Paul Krebs, University of New Mexico
- Bernard Muir, Stanford University
- Bruce Rasmussen, Creighton University
- Peter Roby, Northeastern University
- Jim Schaus, Ohio University
- Kevin White, Duke University
Holmoe (BYU), Krebs (New Mexico), Muir (Stanford), Roby (Northeatern), and Schaus (Ohio) are from schools that had no hope of making the tournament. Cone (UNC Asheville) didn’t have a team to put in either, but her situation is still different than those five.
Barnhart and White are exempt from what I am about to discuss. Kentucky was well-deserving of a 2-seed, and anything else, in either direction, would have been inappropriate. As for White, Duke tore through the ACC Championship and not only didn’t get a 1-seed out of it, but didn’t get the highest 2-seed and also saw its archrival, North Carolina, who Duke beat twice, get a 1-seed.
But the other two, their participation and influence could be called into question. Let’s start with Cone. Duke shouldn’t have been a 1-seed, no matter how well they played this week. And since that’s the case, North Carolina absolutely should not have been a 1-seed. The Tar Heels were one spot higher than Duke in the RPI, but for reasons explained further below, this committee didn’t adhere to the RPI.
By every metric other than RPI (except KenPom, but it is glaringly obvious that this committee has rejected the utility of KenPom’s metric), Duke was better or even with North Carolina. They both lost to two non-tournament teams. UNC won the regular season title. Duke won the conference tournament, though. And, again, Duke beat UNC twice, including last week.
So, one might be left to wonder whether Cone, an employee of the University of North Carolina system, might have given some assistance to the Chapel Hill campus on getting them to the 1-line. Again, it’s not that Duke should have been a 1-seed, but there were teams between Duke and North Carolina (we know that Duke wasn’t the 5th team on the board because they got paired with overall 1-seed Villanova). So, how did UNC end up ahead of Kentucky and Arizona, the 5th and 6th teams on the true seed list?
Then there is Rasmussen. Creighton had a nice season, and they were certainly deserving of a bid. There’s no dispute. But on January 18, the Blue Jays were 18-1 when Maurice Watson was injured. He wouldn’t return the rest of the year. From that point on? Creighton was 7-9. Every year we here about how the committee will knock teams down on seed lines for injuries, accounting for how they played without a certain player who won’t be available in the tournament. In fact, they did it to Oregon, who presumably would have been on the 2-line, at worst, if Chris Boucher hadn’t been lost for the season a couple days ago in the Pac-12 Tournament.
Creighton? They got a 6-seed. Sure doesn’t sound like the committee looked at them as a team without Maurice Watson, does it?
Then there is the coup de grâce. The 10th member of the committee? Michigan State Athletics Director Mark Hollis, who served as committee chairman.
This is where the RPI argument truly falls apart and one has to wonder about the committee’s impartiality. The Spartans are 50th in the RPI. Only three teams ahead of Michigan State in the RPI didn’t make the tournament — the aforementioned Illinois State, Monmouth, and UT Arlington.
So, of the 68 teams in the tournament, the Spartans were 47th based on that metric. With teams ranked 1-68 based on RPI, Michigan State should have been a 12-seed. Only three teams below Michigan State in the RPI earned at-large bids — Marquette, Providence, and Kansas State. So, in theory, the Spartans should be in Dayton on Tuesday or Wednesday.
But Michigan State suspiciously got a 9-seed and was listed 35th on the complete seed list. Did the committee use the blasphemous KenPom for Sparty? No. Because Michigan State is ranked lower than Indiana on KenPom, falling in at 43rd. That would have made them an 11-seed.
Wisconsin won six more games than Michigan State. They lost two games to non-tournament teams and played in the Big Ten final. Michigan State lost five games to non-tournament teams. Wisconsin was seeded one line higher.
Northwestern won five more games than Michigan State and lost only three games to non-tournament teams. The Wildcats were seeded one line higher.
So, how did Mark Hollis and his committee decide that Michigan State was worthy of a 9-seed?
Well, let’s look at what he said about Syracuse, who won one less game than Michigan State and played in an indisputably better conference, as a guide.
Committee chairman Mark Hollis sites SU's 2-11 road record and struggles in non-conference sked that wasn't that tough— Syracuse Basketball (@syrbasketball) March 12, 2017
Yeah, about that. . .
Michigan State was 2-8 on the road. The two? At Nebraska and in overtime at Minnesota. Syracuse’s two? Clemson and North Carolina State. Not much difference.
Well, what about the eight and the 11?
Michigan State’s road losses: Duke, Penn State, Ohio State, Indiana, Michigan, Purdue, Illinois, and Maryland.
Syracuse’s road losses: Wisconsin, Boston College, Virginia Tech, North Carolina, Notre Dame, Pittsburgh, Georgia Tech, and Louisville.
Syracuse lost to three non-tournament teams on the road. Michigan State lost to four non-tournament teams on the road.
Well, he did say non-conference too. Let’s take a look:
Michigan State beat Wichita State, who Hollis’s committee says is worse than Michigan State. They beat Florida Gulf Coast, their only other win over a tournament team. And they lost to Arizona, Kentucky, Baylor, Duke, and Northeastern.
Syracuse did not beat a tournament team. But they lost to three non-tournament teams: UConn, Georgetown, and St. John’s. They also lost to South Carolina and Wisconsin, two tournament qualifiers.
So, the question becomes, was Syracuse’s non-conference performance so much worse than Michigan State’s that they could have fallen from a 9-seed to the NIT?
No. Not at all. Especially when you consider that Syracuse beat Virginia, Duke, and Florida State, all in the top-17 of the committee’s true seed list. Michigan State? Their best win based on true seeds was against Minnesota, who was 18th. Their next best? Michigan, who was 27th.
Maybe my take on Cone and Rasmussen are wrong, maybe I’m reading too much into their connections and/or personal interests. But there is no question that Michigan State benefitted from Mark Hollis, their AD, chairing the selection committee.
So, what’s the solution?
Many believe there aren’t enough “basketball people” on the committee. Maybe that’s so. But really, there aren’t enough disinterested people on the committee. It should be the case, every year, that the 10 individuals that make up the selection committee are 100% disinterested and impartial. How? The NCAA should stop using Division-I athletic directors to select the field.
Perhaps there aren’t any good solutions as to who should replace the AD’s. There would be problems with using the national media, D-II and D-III AD’s, or any other group that could feasibly be assembled. But one thing is certain: this system leaves the NCAA’s premier event vulnerable to nepotism and speculation about voting and seeding for their own interests. I offer the 2017 bracket as Exhibit A.