Wednesday, March 16, 2011

2nd Annual Fulminations of the Season

It’s mid-March again, time for the annual rant against the incompetence of the NCAA tournament selection committee. This year, I didn’t have time to work out “my” system, tracking every game by every team. And I couldn’t find a stat sheet showing each team’s record against top 50, top 100, and top 200 opposition. So here’s what I did: I tracked every team that made it to the NCAA tournament, got a vote in either the AP or coaches’ poll, or placed in the top 50 in RPI, the Sagarin ratings, or the Pomeroy ratings. I think those five systems are listed in increasing order of accuracy and provide a nice combination of the subjective (the polls) and the objective (the three computer-generated rankings). Moreover, the objective analyses include both disjunctive win/loss and more continuum-based analyses (a blowout is better than a squeaker). I took each school’s ranking in each area, multiplied it by 1, 2, 3, 4, and 5, respectively, and divided by 15. (For the two polls, anyone not getting any votes at all was given a 50.) The best a team could do, then, is 1; the worst somewhere in the mid-300s.

At the top, it all looks pretty good: the top three teams in order are Ohio State (1.20), Kansas (2.12), and Duke (2.87). After that it gets shakier. Teams that made the tournament and shouldn’t have: Georgia, UAB, Southern Cal, and Virginia Commonwealth: familiar faces, all. (At least three of the four were relegated to play-in games.) Those who should have got in and didn’t: Virginia Tech, New Mexico, St. Mary’s, Colorado. At least there wasn’t anything egregious this year, although Virginia Tech was better this year than eight teams with at-large bids, and a full dozen teams better than Virginia Commonwealth didn’t get in.

The biggest problems were the seedings within the field, however. Most abused was Utah State, who earned a 5 and got a 12. Really. They’d be a 4 according to their RPI, a 5 according to both polls and Pomeroy, a 6 according to Sagarin. Literally no one lists them below #21 in the country. So how’d they get dumped down to a 12 seed (numbers 45-48)? Allow me to quote myself from last year, when Utah State earned an 8 and got a 12 and UTEP earned a 9 and got a 12: “The crimes of these two teams? They’re from mid-major conferences, they aren’t Gonzaga or UNLV, and they don’t even have the decency to be from the Eastern or Central time zones. What do they expect?” And this year, those stupid Aggies are still located in Logan. They just won’t learn.

Next most under-valued: Belmont, whom Pomeroy actually lists as #18 in the country (a 5 seed), earned a 9 overall, and got a 13. They’re at least conscientious enough to have their campus in Nashville, but man, are they pushy. They seem to think that going 30-4, with all the losses on the road, three of them against tournament teams from the SEC, they should get a little respect. Sheesh.

Other teams seeded at least two rankings below what they deserve: Texas, Kentucky, UNLV, Gonzaga, Missouri, Richmond, Clemson. You know how I said “they aren’t Gonzaga or UNLV” last year? This year, those places get kicked around, too. Mizzou has at least had a strange season: they’ve played really well at home, horribly on the road. So at least I see the committee’s logic. (Besides, they’re Mizzou; don’t expect this loyal Jayhawk to weep much for them.) Clemson, however, earned a 9 (a 6 according to Pomeroy) but had to go to a play-in game. And besides, everyone knows the Big East is invincible, so those Big 12 and ACC teams are just delaying the inevitable. Don’t ask me about Kentucky; I got nothin’.

In the other direction, there’s Vanderbilt, Butler (hey, they were good last year), Michigan, UCLA, and Tennessee. Most over-rated: UCLA, who earned a 12 (a 13 from Pomeroy) and got a 7. No individual metric has them higher than a 9; none of the objective systems have them above an 11. Go figure. Plus, two teams from the SEC (not counting Georgia who shouldn’t be in the field at all and got a 10-seed) and one mediocrity from the Big 10 (deserved a play-in game, got an 8). No surprises there.

Of course, it’s not just the under-rated teams that suffer: in other words, not only should Utah State not have to play a 5-seed or Belmont a 4, but Kansas State and Wisconsin both deserve easier first-round games than the Aggies and the Bruins will provide. On the flip side, while Pitt is supposedly the overall 4th seed, no one familiar with the game thinks their bracket isn’t hands-down the easiest. The Panthers don’t deserve a top seed to begin with, yet they get a 2 who should be a 4, and a 3 who has two 18-point losses (one at home, one on a neutral site) in March. As the overall #4, they should have to face the overall #5 (or somebody who beats them) to make it to Houston. The best other team in their bracket is at #9 and fading fast.

Last year, Duke rode precisely this scenario, the overall #4 and by far the easiest bracket, to a national championship. It will be much harder for the Blue Devils this year, as their prospective 3rd and 4th round opponents are actually both under-rated. [EDIT: Apparently the NCAA has decided that the first game of the tournament for 60 of the 68 teams in the new format is now the 2nd round. It was the first round last year, but there are big-conference mediocrities in the play-in round this year, not just champions from less prestigious conferences, so of course we have to change the numbering system. Therefore, Duke's 4th and 5th round opponents are likely to be high-quality and under-rated teams. Sheesh.]

The real problem with the system is that teams like Utah State may well lose in the first round: they should be about evenly matched with Kansas State. And if KSU prevails, the talking heads will all babble about how USU wasn’t so good, after all. The point is, Utah State should have drawn Bucknell, not the Wildcats, who were rated #3 in the country at the beginning of the season. Similarly, significantly over-rated UCLA may win over (also over-rated, but less egregiously so) Michigan State; but they should have had to face Georgetown in the first round. (I'll also note that the Jayhawks might well face another second-round game against a team significantly better than their seed: hopefully, they won't play like they did last year.) Still, despite the structural advantage to being over-rated, I’ll predict a better tournament record for the teams I’ve identified as under-rated than those I’ve called over-rated, despite having to play higher-ranked opponents. Last year, they went 10-9 vs. 4-11, respectively.

1 comment:

nikolajm said...

ONly comment I'd make is about Butler. Didn't they take some unexpected victories late in the season? Might be the reason for including them where they did.

Otherwise: you know a hell of a lot more about this than I do, Rick. More power to you.