Basketball Prospectus' Kevin Pelton and Joe Sheehan debated via e-mail Sheehan's conclusion in a post on BBP Unfiltered that the selection committee hurt non-BCS teams by pairing so many of them against each other in the first round of the NCAA Tournament bracket.
Kevin: Joe, I had the same thought as you during the selection show--I even tweeted as much. But a reply on Twitter got me thinking whether those matchups really were inevitable. I went through and did the math and if you randomly matched teams up, there would be 4.5 of those games instead of the five there actually are.
Seeds Pro Act
1/16 0.0 0
2/15 0.0 0
3/14 1.0 1
4/13 0.0 0
5/12 2.0 2
6/11 0.5 0
7/10 0.5 1
8/9 0.5 1
Total 4.5 5
(Explanation: There are two No. 5 seeds from non-BCS conferences and all four No. 12s are from non-BCS conferences, so there was no alternative but two matchups of non-BCS teams among the 5/12 games. There was one No. 6 seed from a non-BCS conference and two No. 11 seeds from non-BCS conferences, so we'd expect 0.5 matchups of non-BCS teams if 6/11 matchups were done randomly. And so on.)
Joe: The thing is, it's *not* a random process. They're sitting down, doing a bracket and shoving these teams together, while also giving us Cal/Louisville and Clemson/Missouri. I stand by the idea that this is intentional, and has increased since the Wichita State/George Mason regional semi in '06. They simply don't want more than one, maybe two widely spaced Cinderellas getting to that second weekend.
Running at this from a different direction, follow the money. Conferences get shares based on how far their teams advance in the tournament. It's best for the dominant conferences if the smaller ones are capped in how many shares they can get.
I'd believe in this less if there was ever a year where we saw the other end of the distribution, where the mid-majors don't get matched up. I mean, they even wildly underseeded the A14 teams, and probably Butler, to make it happen.
It's bad for the tournament and for the game. Cornell, UTEP, St. Mary's, UNI...these teams can't get games against BCS schools in the regular season short of playing at their place, and sometimes not even then. The tournament should provide some of that opportunity, and it's aggressively not doing so, essentially protecting the scheduling practices of BCS schools by not potentially exposing them to quality lower-conference teams in the tournament.
It lessens the tournament.
Kevin: In a sense, though, the more respect the mid-majors get in the seeding process the more it creates these kinds of matchups because naturally there are no BCS teams among the lower seeds. This year, everyone 12 and lower was non-BCS; whether that was fair or not I'm not sure, but it means the two 5/12 games and the 3/14 with New Mexico had to be all-non-BCS matchups.
Richmond/St. Mary's and UNLV-Northern Iowa were 50-50 propositions. The selection committee missed out on another chance with Xavier and either San Diego State or Old Dominion in a 6/11 game, which was another 50-50 proposition. Two out of three 50-50 chances doesn't strike me as strong evidence of a conspiracy, even if the motive is surely there.
Joe: If your argument is that this is a problem created by the mid-majors getting too *much* respect, it fails. Any list of the most underseeded teams in this tournament includes Temple and Richmond at the top, then includes Northern Iowa, UTEP and Cornell. Had those teams--as well as Florida, Notre Dame and some ACC teams--been seeded properly, these matchups would have been that much less likely.
I think you've got the causal arrow wrong. Underseeding non-BCS schools is creating some of the problems that are being "solved" by a minitournament--BracketBusters II--inside the bracket.
I'm not saying building a bracket is easy. I'm saying that the way they're doing it diminishes the tournament in a way that is superstructurally favorable to the BCS schools, and that they do it every year, and if it's NOT intentional, it shouldn't happen every year as a matter of course. It does.
Kevin: I would say those are two separate but related issues. There's certainly a strong argument to be made that mid-majors were underseeded, but I think that's different than whether the committee is trying to pit non-BCS teams against each other. Underseeding the best of the non-BCS teams makes it more likely they play against BCS foes seeded 8-11, not less. I think the matchup argument was valid in 2008--when the only 7 from a BCS conference ended up playing the only 10 from a BCS conference, defying the odds--but I'm not sure I see it in the scheduling last year and the numbers don't show it this year.
Joe: I got a lot of feedback on my Unfiltered post from people applying math to the problem, noting that the number of matchups wasn't unusual compared to what you'd expect from a random process. I respect that viewpoint, data-driven and, quite frankly, more consistent with the Prospectus approach than my own opinions.
With that said, this isn't a random process, and there are people involved with some vested interests in the outcomes. I don't think it's an elaborate conspiracy; I also don't think it's completely innocent. I'll say it again: CBS never again wants to televise Wichita State and George Mason in prime time on the second Thursday, and structuring the tournament to minimize that chance pleases them. They want Cinderella, just not her whole sorority.
After the treatment of athlete-students, the imbalance between BCS leagues and the rest of Division I is the biggest challenge the NCAA faces. I'd like to see them not use their biggest stage to further disadvantage the schools outside the blessed 73.
Kevin Pelton is an author of Basketball Prospectus.
You can contact Kevin by clicking here or click here to see Kevin's other articles.
Joe Sheehan is an author of Basketball Prospectus.
You can contact Joe by clicking here or click here to see Joe's other articles.