2012 Election | Barack Obama | Mitt Romney | Nate Silver

'Times' Chides Nate Silver, Who Still Predicts An Obama Win

The most recent numbers from New York Times statistician Nate Silver still show President Obama trouncing Mitt Romney next week.

According to Silver's probability calculations, Obama will claim 303.4 electoral votes, far over the 270 needed to win; Romney, according to this calculation, only gets 234.6 such votes.

If Obama wins, it will be the second presidential election that Silver has correctly called, and it could also mean that Joe Scarborough gives $1,000 to charity. That is, if Scarborough took Silver's bet, a bet that has already earned the ire of the Times' public editor. An excerpt from Margaret Sullivans' excoriating criticism:

Whatever the motivation behind it, the wager offer is a bad idea – giving ammunition to the critics who want to paint Mr. Silver as a partisan who is trying to sway the outcome. It’s also inappropriate for a Times journalist, which is how Mr. Silver is seen by the public even though he’s not a regular staff member.


ranted, Mr. Silver isn’t covering the presidential race as a political reporter would. But he is closely associated with The Times and its journalism – in fact, he’s probably (and please know that I use the p-word loosely) its most high-profile writer at this particular moment. When he came to work at The Times, Mr. Silver gained a lot more visibility and the credibility associated with a prominent institution. But he lost something, too: the right to act like a free agent with responsibilities to nobody’s standards but his own.

Feed This post's comment feed


  1. I was very concerned when Nate's 538 blog transitioned into a NY Times column. I didn't see the move as giving him an elevated status or visibility. He had already earned it in 2008. Wish he had kept the blog-he probably would have made more money while maintaining independence.

    Posted by: Sean Denoyer | Nov 2, 2012 6:49:22 PM

  2. It is a bit meaningless to call an election close. The popular votes in every election differ at most a few percentage points. Every time you can call it "close". But if you look at electoral college, it is definitely not close in this election.

    Posted by: simon | Nov 3, 2012 12:38:25 AM

  3. Nate-Silver has a recent article explaining:
    Nov. 1: The Simple Case for Saying Obama Is the Favorite
    The only reason those irrational evidence deniers can give is that all the polls are biased.

    Posted by: simon | Nov 3, 2012 6:49:14 AM

  4. "But if you look at electoral college, it is definitly not close in this election."

    That was one of the reasons the electoral college was born: it is a way to portray the president-elect as having carried more of the country than the popular vote would show.

    Posted by: Diogenes Arktos | Nov 3, 2012 1:41:20 PM

  5. Someone posting as BZ stated that "Were it not for gambling, there would BE NO probability and statistics. The whole field of probability and statistics was invented by professional gamblers," and this is simply not true.

    While a mathematician, Gerolamo Cardano, who liked to gamble, came up with some of the basic concepts, that was in the 1500s. Other contributors included Fermat, LaPlace, Huygens, and Pascal. The modern formulation based on measure theory was developed by Kolmogorov.

    It's uses go well beyond opinion polls and gambling.

    Posted by: Bill | Nov 3, 2012 5:11:42 PM

  6. One person commented that one of the original reasons for the electoral college was "to portray the president-elect as having carried more of the country than the popular vote would show." This is not the case. I'll refer interested readers to http://electoralcollegehistory.com/electoral/fecmemo.asp and also http://electoralcollegehistory.com/electoral/federalist68.asp (the latter URL is a link to a document written in 1788).

    Posted by: Bill | Nov 3, 2012 8:55:22 PM

  7. What Nate Silver is most likely doing is to use polling data from individual states to get the probability of Obama or Romney winning, the the randomness reflecting sampling error. He may also use linear regression to estimate the state-by-state vote given that how people would vote may be changing with time. Once you have those probabilities, you can run a monte-carlo simulation of the election and get a distribution of the electoral votes, or compute it analytically (but a monte-carlo simulation would likely be quicker to code if you have to write the software).

    Posted by: Bill | Nov 3, 2012 9:10:21 PM

  8. As a freelance journalist who loves following presidential elections every four years, I think Nate Silver's analysis is right on and a breathe of fresh air in the rather otherwise old world way of political punditry and covering elections.

    Wish Nate was around when I was a meeber of the Electoral College in the 1980's.

    Keep up the good work, Nate!

    Posted by: Scott Larsen | Nov 4, 2012 1:01:27 AM

  9. « 1 2

Post a comment


« «Watch Mitt Romney Get Angry While Explaining Mormonism: VIDEO« «