Last night's biggest winner wasn't Barack Obama. It wasn't marijuana or gay advocates, either. No, the biggest winner was the New York Times' resident statistician Nate Silver, who, after correctly predicting 49 out of 50 states in the 2008 election, correctly predicted all 50 states in 2012. As a measurement of Silver's popularity, his blog, FiveThirtyEight, grabbed 20% of the NYTimes' traffic on election night.
So is Silver, as supporter suggest, "a witch," "from the future"?
Um, no. One skeptic, Slate's Daniel Engber, explains that Silver's accurate predictions aren't unique to Silver. They derive from averaging state polls; and Silver did no better than other polls' averagers, like Real Clear Politics, at predicting the outcome. Plus Silver wasn't the only one to accurately predict the outcome - other statisticians managed it too.
Silver's supporters say his model is a victory of math over gut feelings, taking the human bias out the equation and allowing numbers to talk. But Silver wasn't accurate about all of the Senate races, undermining the idea of his model as magic.
Not really. All the predictions, polls , and pundits tossing their two cents worth make me weary. That said, I do not think Nate Silver is a witch form the future. Just your everyday run of the mill evil genius. The man is one smart cookie even with\out the stats.
@mimi57 Einstein wasn't bad looking when he was younger, but that's beside the point - some people are attracted to people because of their hearts and minds. I'll take a man that can engage in a deep, meaningful conversation with me any day over one who has nothing to offer other than sitting there looking pretty.
Somebody should make Dick Morris get dress up like when he was caught with a hooker in a garter belt and dog collar appear on Imus saying "I was wrong, hit me beat me, I was wrong." I really enjoyed Karl Rove's little hissy fit what a clown how much money did he piss down the drain to no avail
I thought Silver's principal contribution this year was his scoffing at the punditry and its ridiculous "Mittmentum" fable. Gov Romney never had momentum of any sort and those who called this a "razor thin" race (like Joe Scarborough) were, as Silver said last week, just entertaining the public, not making serious political judgements.
If you ever read this .. I want you to know it placed my bets on your work because I knerw it was good!! In terms of relying on information about the election I give credence to you and Chuck Todd, period!! I sent links to your work to my family and a few trusted colleagues. You are absolutely the best.
On this site I have told people I respect about your work also! Thanks for your work.
Of course, what else are we going to use to predict an outcome subject to chance events?
Nate Silver is doing things in a good defensible way. I'm sure he could publish his methods in a statistically-oriented journal if he wanted. But there's no such thing as "correct" in what he does. There are many different valid ways to do it and the only sensible metric is how often he predicts the correct outcome.