Thursday, March 18, 2010

NCAA selection committee errors or bias in seedings: some data

From Russell Adams in the WSJ...

How much should you trust the people who build the bracket?

Every year, fans accuse the NCAA selection committee of overseeding (and underseeding) certain schools....

We recorded the scores of every tournament game played since 1985 to find out how close each game should be based on the seeds of the two teams playing. By establishing the average scoring differential for each seed pairing (for instance, 3-seeds beat 14-seeds by about 11.5 points), it's possible to compare every team's performance against a benchmark. Some teams consistently outperform this average-scoring margin, and others consistently underachieve...

Here are the teams in this year's tournament that have done the most and the least to justify their seeds since 1985. (They provide a full list of teams on-line.)

(Avg. Pts. Below Seed)
(Avg. Pts. Above Seed)
Wake Forest (5.68) West Virginia (4.08)
New Mexico St. (4.92) Texas A&M (3.57)
Clemson (4.65) Kansas (3.28)
BYU (4.47) Kentucky (3.25)
Tennessee (3.62) Louisville (2.79)

In my bracket, I'm hoping that they over-corrected with BYU this year. I have them going to the Final Four-- along with Baylor and Kansas who will beat Kentucky in the Finals.


Post a Comment

Subscribe to Post Comments [Atom]

<< Home