Tuesday, March 17, 2009

You Never Took Statistics, Did You?

Every year around NCAA tournament time, some sportswriter seems to make an argument so bad that it really makes me wonder how he has a job. Last year it was Dennis Dodd saying that Tim Floyd was the best coach in the Midwest region. At the time, I noted how insane that view was and, sure enough, Floyd's USC team lost in the first round to Kansas St. (who then lost to my Badgers).

Now, apparently, it's Gregg Doyel's turn. His contention is that the NCAA should stop awarding games closer to home for the tournament's higher seeds. A perfectly legitimate view and one worth exploring.

What's problematic--and I'd say baffling--is the argument. Doyel points out that since 1999, teams playing de facto home games are 71-8, making them almost sure winners. Okay, I'm listening. But then he drops this beauty: "In the first two rounds, "home" teams are 60-5 (.909). That number is skewed by the number of high seeds who get those "home" games--but then again, it's all related."

Very good, Gregg, it is all related.

Indeed, the only relevant question--and one he never addresses--is whether "home" teams of comparable seedings have a significantly higher winning percentage than those who have to travel greater distances. With that information, one has the basis for an interesting argument. Without that information, all Doyel has shown is that a certain subset of higher seeded teams win more games, which is pretty much what we'd expect, isn't it? What we're left with, then, is a mildly interesting opinion supported by irrelevant evidence.

Must be nice to have a job with such a low bar of success.

1 comment:

Anonymous said...

My family every time say that I am wasting my time here at net, but I know I am getting experience everyday by reading such fastidious articles.

Here is my web site: http://balanceinmotion.com/