Is the NCAA Stacking the Deck Against the Mid-Majors?
by T.H. on Mar 14, 2010 7:48 PM EDT
in Basketball
Email
PrintJoe Sheehan is annoyed:
The committee did it again, matching up non-BCS schools aggressively and keeping them away from BCS schools. UNLV/Northern Iowa. Butler/UTEP. Temple/Cornell. Richmond/Saint Mary’s. The committee is taking one of the best things about the tournament–that the big guys ahve to play the little guys on a neutral floor–and destroying it, aggressively so. Defenders of the bracket and the committee will always point out that this isn’t intentional, but after it happening year-in, year-out, I simply don’t believe them. You can’t keep playing off the non-BCS schools one another every year and pretend it’s not a strategy. It very clearly is one, and it’s designed to prevent the possibility of the schools from smaller conferences showing that the main difference between them and the middle of the BCS leagues is home games. The commmittee and the NCAA should be embarrassed.
There are 15 non-BCS schools on seed lines 5-12 in this bracket. Eight of them are playing each other. Thanks, NCAA. Just what the fans want.
Damning evidence. But wait. 5-12 seeds? That's eight teams per region, thirty-two teams over all. Almost half of them are non-BCS, and half from the big six conferences. If, to make the math simple, there was 16 BCS and 16 others, and you paired them completely random, you'd end up with... eight non-BCS teams playing each other. And eight BCS teams playing each other. And the remaining eight games would pit one small team against one large one.
That's pretty close to what we got. Michigan State (5) plays New Mexico State (12). Tennessee (6) vs. San Diego State (11). Gonzaga (8) vs. Florida State (9). Brigham Young (7) vs. Florida (10). Xavier (6) vs. Minnesota (11). Texas A&M (5) vs. Utah State (12). Notre Dame (6) vs. Old Dominion (11). Seven games, balanced relatively fairly as to which team is the higher seed.
If you want to argue that the committee should only seed mid-majors against the middle ranks of the BCS conferences, that's one thing. But that seems just as hacky as doing it the other way around. There are a lot of problems with the seeds, but that there's a concerted effort to keep the little guy down isn't one of them.
2 comments
|
0 recs |
Do you like this story?
Comments
Some more math....
Since there is no replacement, the odds don’t work out quite so cleanly. I was curious as to exactly how it would turn out (and didn’t feel like doing the algebra by hand), so I wrote a script to simulate the random pairing of teams (you can do this by hand by randomly pulling pairs of cards from a deck and counting the red-red pairs you get each time through the deck). I ran each simulation 20000 times, which for the standard deviations we are dealing with (about 2.0 for each trial) means my values for the expected number of mid-majors playing each other is 95% likely to be within 0.03 of the true expected value.
If it were actually 16 mid-majors and 16 BCS schools, then we would expect slightly fewer than 8 of the mid-majors to be playing each other (it comes to 7.75). But it rounds to 8, and your argument works.
However, changing the field to 15 and 17 changes the math. With this field, we would actually only expect about 6.76 mid-majors to be playing each other, with a standard deviation of about 2.03. So it is not statistically significant to find 8 mid-majors playing each other, but it is above the expected value.
I went back through the last several seasons of the tournament and compiled this table comparing the actual number of mid-majors playing each other with the expected number if the 5-12 match-ups (and seeds) are randomly assigned.
Year/#Non-BCS/#BCS/#Non-BCS playing each other/Expected Value/StdDev
2010/15/17/8/6.76/2.03
2009/10/22/0/2.93/1.72
2008/13/19/8/5.02/1.95
2007/12/20/6/4.24/1.91
2006/14/18/4/5.86/2.00
2005/15/17/4/6.76/2.03
Looking at the data this way, there doesn’t seem to be much to suggest that the NCAA is deliberately avoiding letting mid-majors play against BCS schools. However, in looking through the past tournament brackets, typically the mid-majors in the 5-12 seed range tended to be concentrated in the 9-12 range, making it impossible for them to play each other in the first round. If I get a chance later, I’ll adjust the simulation to consider the makeup of the 5-8 seeds and 9-12 seeds and see if they are paired fairly.
But for now, it looks like math trumps conspiracy theories yet again.
by PrincetonBlackSquirrels on Mar 15, 2010 11:29 PM EDT reply actions
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
