Elections, Damn Elections, and Statistics
Yesterday, November 7 President Barack Obama was re-elected for his second term. It was, as almost every opinion maker remarked, a very embattled campaign. If there is one thing I love reading about after election day it’s reading about the political strategies that led to the win. Slate‘s How Obama won Four More Years is a particularly interesting read, I would probably read Vanity Fair’s version when it comes out.
But one of the more interesting aspects for me about this election is how much the internet and the changing demographic changed the landscape. We saw this happen during Obama’s 2008 campaign, launching Social Media into the stratosphere, and we’re seeing it again. Obama announced his win over twitter and his tweet has now become the most retweeted during Election Day.
source
Nate Silver of FiveThirtyEight has certainly left his mark, to most people who knew of Silver this was hardly a surprise. Many pointed out that he’s accurately predicted elections before and it’s because he doesn’t include any political affiliations to the mix. It’s all data, baby.
Nate Silver started out crunching numbers for baseball and later shifted to using his methods to politics. It’s unconventional to say the least, and received a LOT of derision from pundits, especially Republican pundits. Silver shrugged off all the criticism because what he does have was data on his side.
It’s not perfect by any means but Silver has a way of, as one blogger acquaintance puts it siphoning out the subtleties in the polling sample and point out his data set into making things as pretty damn accurate as he can make it. Let me tell you, this is the first time I’ve had an appreciation of math. Silver won himself many fans and a cool $1,000.
(MSNBC‘s Joe Scarborough bet Silver $1,000 should he prove accurate. I guess Mr. Scarborough has now written him a check.)
I’m a bit of a late adopter myself so watching new models disrupt the traditional modes of thinking. in action, in such a broad level is really interesting.