Thursday, March 19, 2009

More on Obama's Bracket

Nate Silver has some interesting observations on Barack Obama's NCAA tournament bracket and his propensity for picking upsets that seems to have some relation to the electoral map.

Looking at the hard copy of Obama's bracket lends some more credence to the theory. There are seven instances where he picked one team to win a game, then crossed it out and picked another team and in five of those cases the change could conceivably have an electoral calculus to it:

-In the final he initially picked Louisville and then changed to UNC. North Carolina was one of the trio of the closest states last year, Kentucky certainly wasn't.

-In the semifinal he initially picked Pittsburgh and changed to UNC. Same thing as above.

-In a second round match he picked Marquette then changed it to Missouri. Missouri, along with North Carolina and Indiana, was part of the triumvirate of the swingiest states last year.

-In another he picked Xavier then changed it to Florida State. While Ohio and Florida are both swing states, Florida was closer last year and the 'Noles have a considerably larger fan base within their state than the Musketeers have within theirs.

-Another changed second round pick, from Clemson to Oklahoma, would not seem to have any political implications with both states dark red.

-In a first round pick he had UCLA winning, then changed it to Virginia Commonwealth. Virginia is likely to be quite competitive at the Presidential level moving forward while California is solidly Democratic.

-He changed from Oklahoma State to Tennessee, another red to red shift.

Do I actually think Obama's bracket was motivated by politics? No. As a good basketball fan he knows that FSU made the ACC final last weekend while Xavier couldn't make the A-10 final. I picked that upset too. And this Tar Heel alum certainly agrees that we're going to be cutting down the nets on Monday night in Detroit. But the relationship is interesting.


Rasmus said...

I think Nates analysis was pretty weak. He failed to get a significant result at the 95% significance level, so what did he do? Just lower the confidence level and that's it.

In fact, Nate (and yeah, I know that you added some more) bases his analysis on three picks. Three.

VCU vs. UCLA- this game is a fairly popular upset.
Purdue vs. Washington- first this is a #5 vs. #4 game, and I don't think that this would qualify for a literal 'upset' either way.
Also, Washington isn't that good- in fact there is even a not-so-bad chance that they don't even make it to the second round.
LSU-Butler- this is a #8 vs. #9 game, so no real favorite there. In fact, I think that Obama will be correct here, and that Butler is favored to win this one.
Yahoo Tournament picks are 52-48% for Butler, and Butler has the better conference record (and I don't think that Horizon is inferior to the SEC).
Well, whatever, this is just a fun topic anyway.

drew said...

Aw, c'mon, Tom. The President just knows we 'Noles need something to collectively hang our hat on given our guys' awful performances on the football field in recent years. :)

Web Statistics