Nate Silver says 91% chance of Obama re-election

Pollster Nate Silver doubled down tonight by upgrading his statistical certainty of an Obama win. If you thought a 73% (or 86%) chance of victory was a bit high, get a load of this.

Before the Left canonizes St. Nate, it’s important to remember that an Obama win would not prove Silver right.  Most people agree that this race is close. Obama might win. Silver’s fallacy is his claim that an Obama loss is highly unlikely.

Silver will be no more right if Obama wins than if he loses, just lucky that he won a reckless gamble.

  • Dandee

    Guess they have an in on the tainted polling machines.

  • Hannah Abbott

    Nate Silver is societal waste. . . remember his name. . .mock him tomorrow and then shun him like the intellectual fraud that he is . . . .

  • brewers_rule

    More delusions from the proggies. Why aren’t we hearing how massive Obama crowds are like in 2008? All we hear is how many show up at Romney events and Obama’s trying to galvanize his base daily with trumped up “war on….” nonsense. This loss may not be Reagan-esque but it’s not going to be Bush/Gore and Obama is going DOWN tomorrow.

  • RAD1

    He’s smoking crack!

  • JC

    My data has Romney at exactly 100% to win, so if Romney wins will the NY times fire Silver and give me his job?

    • $21367552


  • JackiB

    The strategy of the left is to declare victory early and often to keep conservatives thinking it’s a lost cause and it’s not worth it to vote. DO NOT BELIEVE THE LIES! They did this in Wisconsin–all day they said Gov. Walker was losing the recall election, and he ended up winning big. VOTE LIKE YOUR LIFE DEPENDS ON IT!

  • Purple State

    The accuracy of Silver’s predictions can be measured actually — against his predictions for each state. He’s not pulling these numbers out of thin air. He’s tracking polls in each state and combining them with a bunch of other factors (some named, some not — his rationale is there for anyone to read) to go state-by-state and measure the Electoral College Votes. So as those states get called tomorrow night, we’ll know how right or wrong he was. That’s the fun of it.

    If a meteorologist said, “There’s a 91% chance of sun tomorrow” you’d wear short sleeves, but that doesn’t mean it couldn’t end up storming instead. He wouldn’t be a “societal waste” or “intellectual fraud” if there were thunder and lightning.

    It’s just a prediction.

    • Pete Futz

      Nate ain’t no meteorologist and I never believe those idiots anyway. I go outside to see what the weather is then get dressed.

    • yahneverknow

      The problem is… everything he is basing his possibly “perfect” math on is skewed. The pollsters are delusional to assume partisan voting. And they do not take into account any other factors that will sway voters. All they need to do is watch this Obama speech in Iowa that I am watching on TV while watching Romney speak in New Hampshire; it’s ludicrous how different the crowds are.

      As for Silver? He obviously has not figured out how to quantify people like me.
      Independent. Rational. Female.

      (Perhaps this is more of a personal issue for him.)

      • Purple State

        He has something in his formula apparently that accounts for possible skew, and he factors in a bunch of other numbers that affect voting. It’s not only polls, so he may know how to quantify based on gender, party identification, etc. We’ll know tomorrow!

        • yahneverknow

          “something in his formula”
          “bunch of other numbers”
          “may know how to quantify”

          Most of his ardent defenders point to how transparent his model is.


          My point is more that the polls themselves don’t hold up to any kind of poking around and that they and Silver don’t appear to be able to quantify the intangibles.

          Plus… if he had offered his wager with whatshisname (Scarborough?) at the odds he was predicting, I would give him more cred. He offered 1-1 money when it should have been more like 3-1 at that point. :)

          • Purple State

            His model isn’t transparent. He won’t release his code — it’s his livelihood, so why would he? But don’t take my vague understanding of it as a fault of his. I’m the last person who understands aggregations of statistics. He can quantify the intangibles — he’s just not revealing how he’s doing it because otherwise he’d be out of a job as the stats guy.

          • Smitty 

            Here is the thing. No matter what you say on Twitchy, or how accurate Mr Silver was in 2008, and 2010, and how right he has been with the Baseball models, nobody is going to care. Science is too big of a word for {most} of the people who read this site.

            They would rather take seriously the guy who runs “Unskewed Polls” who instead of using science based in reality, attacks Mr. Silver for being too “Effeminate” to crunch poll data.

            Good luck changing their minds though.

        • Kate

          Yes, but he has admitted that if his model is off by 0.8%, his prediction falls apart.

          Using your weather analogy: He may be predicting a 91% chance of sun, but I’m still packing an umbrella!

    • Kent Vig

      If you jam a bunch of bad polls into an equation you are going to get a bad result. In 08 he had access to Obamas internal polling. He doesn’t have that this time. If the turnout is the same as 08 then Obama wins but if its like what Ras and Gallup say then it will be a massive landslide for Romney.

    • tomtom1983

      One problem, its not the numbers that are to be criticize but the process that lead to those numbers. You can’t base actualities when you oversample your own party and then apply that sample on a State to Nationwide scale.

      It just isn’t accurate and very misleading. So the numbers given by Silver are not viable in any way.

      • Purple State

        It’s a prediction. We don’t know that it’s not accurate until the votes are counted. He’s done very well before predicting elections, poker games, baseball stats. We’ll see! Exciting stuff.

        • yahneverknow

          You actually won’t ever know how accurate his predictions are.

          They are probabilities. The only way to see if they are (mostly) correct is to have the same event happen over and over and over again, ad infinitum, or at least as close to infinite number of times as is possible, to even out the results curve.

          That can’t happen, so there is no way to see if he was accurate.

        • yahneverknow

          Note: Yes, he has this overall electorate map “predicting” things. But the way this guy works is all probabilities. Because of this and because of how the electoral college works, his math could be totally off, but he could still end up with the correct “prediction” for any particular state.

          It really depends on what part of his “predictions” you are talking about to determine accuracy.

          But, because everything is based on probabilities, Romney could win and Silver’s math could still be completely correct.

    • TomJB

      “He’s tracking polls in each state and combining them with a bunch of
      other factors (some named, some not — his rationale is there for anyone
      to read)”

      Never mind the fact that the “not named” “bunch of other factors” came from somewhere dark and moist, how can anyone read factors not named?

      • Purple State

        He’s not named the factors to outside observers. He knows all of his own factors. He won’t release the code he uses because that’s his livelihood.

        • TomJB

          No peer review necessary, just “trust him” I guess his popularity is to be expected amongst those who vote based on “hope”

        • TomJB

          No peer review necessary, just “trust him” I guess his popularity is to be expected amongst those who vote based on “hope”

  • yahneverknow

    I’ll take $10k for a $1k wager if he’s willing.

  • Mike Rogers

    I’m with @rsmccain. I want this fraud and his model DESTROYED by tomorrow’s vote.
    Then again, FRAUDian slips seem to be an occupational hazard of getting close to Obama!

  • mike_in_kosovo

    He must’ve believed that CNN poll.

  • Harlan Hikaru Fox

    Nate’s got the inside track on exactly how much fraud the Chicago Machine has spread across the country. That’s why he’s so confident.

  • tomtom1983

    Can someone please explain to these morons the concept of propaganda and that its unwise to believe your own deceit.

    Its hilarious but unwise. Lol what twits.

  • JD Olson

    My theory: Lefties want to be lied to. With their capacity for cognitive dissonance, they will never believe Obama lost.

  • tjp77

    Why even bother having elections anymore? Every year we can just ask Nate Silver who he likes and leave it at that.

    I mean, who needs democracy when we have statisticians?

    • Kate

      I love the Obamabots on HuffPo who site his states as finite proof that Obama will win.

      Even the best statisticians and pollsters know that you cannot account for human error.

  • Kevin Krom

    Hmmm… and my “model” just updated to show a 91% chance of Silver being unemployed and humiliated within 24 hours.

  • Bored Dalek

    People just don’t understand statistics. Nate isn’t saying that BO has a 91.4% chance of winning. He’s saying that there is a 91.4% certainty that a BO loss is due to Republican cheating, followed by a 91.4% chance of rioting – I mean – direct action.

  • Kate
    • Greg

      Glenn Beck does a good job of crying on queue also. Doesn’t mean I believe him.

  • Greg

    Well I guess we will find out in a matter of hours if Nate Silver is correct or not. Better head to the gun store GOPers cuz I think Obama is back for four more.