The Sweet Sixteen: Statistics to Ponder

If the tourney's seeding had borne out (in terms of expected wins), the median and mean seed of all the teams remaining in the Sweet Sixteen should have been 2.5 (all 1s, 2s, 3s, and 4s).  In actuality, the median  is 4.5 and the average is 5.

The average of the top half of remaining seedings is 2, which is only .5 less than should have been expected, while the average of the bottom half of remaining seedings is 8, which is 4.5 more than should have been expected (3.5).

Conclusion:  This year, the committee failed to properly distinguish second tier teams from third tier teams. I don't think the sample size is big enough to make the supporting data statistically significant, but this isn't a peer reviewed journal article, so it really doesn't matter.