Monday, April 5, 2010

Forget the NCAA Tournament's expansion to 96 teams -- what about the 2010 Tourney Rule Change?

The media is abuzz about the NCAA tournament's likely expansion to 96 teams next year. Equally intriguing (though not equally hyped) is the NCAA's near-simultaneous announcement that, in seeding teams for the tournament, it stopped weighting the last 12 games of the regular season when evaluating each team's record. Historically, teams that are winning at the season's end keep winning, justifying a higher seed. This year, the NCAA decided to ignore this fact in ranking the teams. So, while expansion has been in the spotlight, we ask: how has the rule change played out?

First a little background. This change in the committee's criteria came in the wake of a tournament with one of the fewest upsets ever.

The 2008-2009 Final Four featured two 1-seeds, a 2-seed, and a 3-seed. Besides Arizona (which, though a 12-seed had two future NBA prospects in the starting lineup), the lowest seed in the sweet sixteen was Purdue at 5.

One could argue then that, in seeding the teams, the NCAA did a near perfect job that year. A tournament with no upsets is a tournament seeded perfectly. Apparently, though, the NCAA thought it could do better by changing the rules. Or perhaps the NCAA changed the last-twelve-games criterion to create the illusion of “upsets” because there were too few of them last year to generate interest in the tournament, which translates directly into cash -- Nielsen ratings for the final game last year were down ~7% versus the prior year. People love an underdog... if it wins. And if you can't make a real underdog win, why not create the illusion of an underdog?

Though two years worth of data are hardly enough for a bulletproof statistical comparison, a cursory glance at the '09 and '10 tourneys does not seem to support the rule change, if fairness is the goal. (We'll wait and see if it supports the Nielsen ratings.) There have been many more "upsets" this year. But were they really upsets, had the teams been seeded by last year's criteria?

Consider Butler's status as a 5-seed. Butler has the nation’s longest winning streak (25 games); most of these wins came at the end of the season (they continued into the tournament). By the old rules of seeding, they should have received more weight. They were paired up with the "1-seed" Syracuse, which lost two games to close the season and suffered injuries at the end of its season.

There's also the case of Washington, which (like Butler, but not to the same extent) was on a tear winning games at the end of the season; they were a supposed 11-seed, but they pulled off two “upsets” to make the sweet sixteen.

Other stories? Thoughts? Please post a reply.