In my new NCAA bracket over-rated index, Duke is clear champion
It pains me to say it, but...OVER-RATED
As you are racing to fill out your tournament brackets today (and remember, don't fall for the Quicken/Buffett ploy!), I have one piece of iron-clad, scientific advice to give you: Pick Duke to under-perform its tournament seeding. Duke is much more likely to bow out of the tournament early than other frequent tournament entrants. I've run the numbers
(I say this as a Duke fan, a Mike Krzyzewski admirer, and uncle to a current Duke student athlete. Sorry, Shannon).
For years, I've been fascinated by the over-seeding phenomenon. Is it real or not? The theory goes something like this: Duke (and other nationally-recognized teams) gets far too much attention, and its gets far too much benefit-of-the-doubt from the tournament selection committee. Well, that's the benign theory. The conspiracy theory is that it's in everyone's interest for Duke to advance, as Duke games produce high TV rankings, so Duke is always nudged towards a little easier road to the Final Four. I don't actually subscribe to that conspiracy -- just look at Duke's placement in the Region of Death this year.
But it seems incontrovertible that Duke, for whatever reason, takes an early exit from the tournament more than other teams.
Sign up for Bob Sullivan's free email newsletter.
To test this theory, I developed an overrated scale. It's pretty simple. A team that gets a No. 1 seed is expected to play at least five games in the tournament -- to make it to the Final Four. If that team lost in the regional final (the Elite Eight round), I give that team a -1. If it loses in the Sweet Sixteen (Regional Semifinal), I give it a -2, and so on. Conversely, I award points for teams that outperform seedings. So a No. 2 seed that makes the Final Four gets a +1. A 9th-seeded team that makes the second round gets a +1. (Because of the way seeds worked, I stopped at the Final Four. No extra points awarded for making the Final, or winning the tournament)
I looked at name-brand teams that made the tournament every year (well, almost every year), and counted up the points since 2000. Why that year? Hey, no one is paying me to do this. I think it's enough to make my point. Anyway, since 2000, Duke's scale was....
NEGATIVE 16!!!!
In the 14 years I considered, Duke under-performed its seed 9 times. The -16 was heavily influenced by a disastrous 2012, when Duke lost in the first round as a No. 2 seed. But their scoreboard is littered with plenty of -2's as well - in 2011, in 2008, in 2006, 2005, 2002 and 2000. That's a pretty clear pattern. Coach K has at times been a great tournament coach. But Duke has a serious tournament problem. (Duke, by the way, never outperformed its seed during those years.)
I couldn't find any other team with worse than a -8 during those 14 years. That would be Kansas, which does seem to suggest some kind of national program bias. Kansas also suffered a dreaded 2 vs. 15 loss, and a few -2's, but it has a smattering of +1's to distance itself from Duke.
Next on the list was Kentucky (obviously, the tournament is upset friendly), as a -6. After that, the committee did better than I thought. Syracuse was a -2. Connecticut was a -1 (aided by its 2011 National Championship run as a 3 seed, for an eagle...errrr a +2.
To balance out all these negative numbers, there have to be teams with positive numbers, right? Yes, but those upsets are spread out nicely among dozens of upstart college basketball programs (like Florida Gulf Coast). They don't fit into my review here because they don't have 12 or 13 tournament entries since 2000. You might think a lovable small school like Gonzaga would rate high in this index, but it did not (Gonzaga hasn't surprised people for years). The Zags have their own -3 disaster (losing as a 1-seed in the second round last year) and come in as a -2 overall since 2000.
On the other hand, Arizona enjoys a +2 overall. Worth considering as a you finish up your bracket tomorrow morning.
My data collection team (me) got tired at this point and I stopped rankings. If anyone wants to see my underlying data, I'll email you. I used the Washington Post's results tool to build my spreadsheet. If you want to throw a team at me that I should have rated, go ahead, and I'll update the story.
Good luck. BTW, I have Duke losing to Louisville in the Elite Eight, which would be the first time the team has overperformed since...1994! Logic be dammed.