Friday, March 29, 2013

FGCU, or the Power of Outliers

Any red-blooded American male knows that there's a formula for how a low-seeded team in the NCAA tournament pulls off an upset.  Takao maintains that all upstart double-digit seeds need a quick point guard, a lights-out shooter, and a bamma-ass power forward who tries hard.  It doesn't hurt when the favored team doesn't take the game seriously.  The lower seeded team hits 'em quick and builds a double digit halftime lead, usually because they hit a ridiculous percentage of their three-point shots, then holds on as the more talented favorite makes a run that may or may not make up for their early laziness.

If it doesn't, you have two days worth of heartwarming stories and images of the latest Cinderella, but two days is all they get.  Cinderellas can win one game in the NCAA tournament, but they rarely win two.  Most of these stories end with Cinderella absorbing a 20 point loss in their second game and everyone forgetting about them.

Needless to say, Florida Gulf Coast University is not following the formula.


This can't be stressed enough: Dunk City is far, far, FAR more athletic than the average low-seed Cinderella, and they didn't hit more than 8 three pointers in either of their first two tournament games.  Neither of their first two opponents took them lightly.  FGCU was basically tied with both Georgetown and San Diego State at the half.  The typical underdog script does NOT include blitzing the higher seed team after halftime with the type of free-wheeling athleticism seen above.

Because I'm a nerd, I think about the difference between the typical Cinderella and FGCU as a difference between types of outliers.  The script I described above is probably the easiest way for an outlier to manifest itself, and outliers are bound to manifest themselves.  Odds say that the worst Division III team in America could beat Louisville or Indiana if they played enough times, even if those odds are probably, literally, at least one million to one.  The odds of a real, live Division 1 program that won its conference getting hot enough from the 3pt line for a day and catching a good team napping are considerably higher.  However, no one pretends that the normal Cinderella is actually better than the teams they beat.  They get hot for one night, and they get a great win, but there's no doubt who would win if they played a best of seven series.  What makes FGCU an outlier among outliers is, I think they're actually better than the two teams they've beat.  Once FGCU stopped holding Georgetown on a pedestal, they blew them away.

Stats has come to dominate sports generally and the NCAA tournament selection process specifically, and we have really good stats that measure the quality of teams that can predict with startling accuracy how 99 percent of the games between two teams will go.  I'm starting to think that FGCU is the 1 percent of teams that advanced stats really can't capture.  This team isn't winning flukey, like most underdogs do.  Over the past week, they have been better than two really good teams, because they've been more athletic and more skilled.  Maybe the stats penalized them too much for playing in a crap conference.  Maybe FGCU just got a lot better over, say, the last month, and the advanced stats are not rewarding that improvement.

It's one of those craziest, most unpredictable things I've seen in sports, and anyone who tells you they know how this will end is lying.  All bets are off with this team.  If they can blowout Georgetown and San Diego State, they can absolutely beat Florida and the Michigan/Kansas winner.  Numbers tell you that they shouldn't have a chance against any of those three teams, but it's pretty clear that numbers fail to capture the magic of FGCU.


No comments:

Post a Comment