OHIO: MATH IS HARD

Posted on Comments

Nate Silver approaches a clue and almost gets it:

For Romney to Win, State Polls
Must Be Statistically Biased

Missed it by that much, as Maxwell Smart might say.

His headline summarizes exactly what Republican poll-mongers have been saying since September, as they see poll after poll with crazy oversamples of Democrats. As for example, the NBC/WSJ/Marist poll that has Obama winning Ohio by 6 points — SIX FREAKING POINTS! — about which Ed Morrissey says, “all you need to know is this: the D/R/I is 38/29/32.  In 2008, the exit polls showed a split of 39/31/30, and in 2010 36/37/28.”

Stop for a second and think about that: What this poll is telling us is that partisan ID has shifted 2 points toward Democrats since 2008, which was the best year for Democrats since LBJ won a landslide in ’64. Therefore, we must choose between two alternative explanations:

  1. Obama is headed toward a world-historic victory based upon the remarkable popularity of the Democratic Party; or
  2. The poll sample is fucked-up beyond all comprehension.

Take your pick, eh? When the sample is D+9 and yet Obama is winning by only 6 points, why should anyone doubt that the NBC/WSJ/Marist lead is entirely a function of the Democrat oversample? And if this is not purposeful bias (i.e., the pollsters actively “modeling” the sample), then we must ask ourselves, “Why would Democrats be so absurdly over-represented?” Could this have something to do with what Dan Collins reports from Wisconsin?

I’m subjected to endless phone calls from the RNC, AFP and other PACs, and the endless polls and pushes masquerading as polls. I may be unemployed, but I have work to do, so I pick up the phone, find out who’s calling, and hang up. I’m going to be poll observing, and my brother Tim and his wife have been knocking on doors and manning the phone banks when they’re not at work, helping to GOTV.
I imagine that there are a lot of people who’ve gone beyond the saturation point, like me, so I’m not putting a lot of stock in any polling coming out of Wisconsin, and I figure it’s got to be the same in OH and IA and VA. But Wisconsin was already suffering from politics fatigue before it once again became a swing state. I imagine, too, that there are just a lot of voters like me who have had it with the incessant calls, and who are no longer participating in polling or answering pitches from campaigns.

Think about it: Multiple entities on the Right are targeting GOP-leaning voters in the swing states with a non-stop flood of phone calls and it may be that, as a result, the pollsters simply can’t get through to self-identified Republicans, so there is an under-sample of Republicans which no pollster can verify or compensate for, without resorting to artificial assumptions about the actual composition of the electorate.

Anyway, I don’t have much interest in numbers-crunching, nor time to get down into the metrics, which is why I’m glad Ali Akbar did it today:

Ohio Math: Numbers Equal Trouble for Democrats

UPDATE: Michael Flynn at Breitbart.com Friday:

This morning, because the sun rose in the East, Nate Silver again increased his odds of Obama winning reelection. Silver now estimates Obama’s chances of reelection at a precise 83.7%.

Exactly. What is the basis for that very precise number?

30 comments

  1. I’ve been saying for a couple of months the odd Democratic slant to poll samples is because our people are refusing to participate.

    But Ohio is fairly clear. UNLESS it is bucking all the historical trends – being slightly more Republican than the electorate at large, Democratic need for early voting, crowd size (when Stevie Wonder only draws 200 to a free concert, they need to go back to handing out cigarettes and liquor) – it is breaking for Romney.

  2. Sometime Tuesday, Nate Silver will either be declared a genius or charletan. But if we are tracking metrics, just remember Nate got 2010 wrong. So basically Nate is redoing RSM’s roulette system. Which worked fine…until it didn’t.

  4. At this point it is obvious that Silver is doubling down calls out his methodology. Could he not be a more representative of a Obama sycophant? He might as well be saying, “Don’t call my bluff!”

  10. Omen: Alabama victory over LSU. Draw your own conclusions. And does Carville look like a shell-less albino snapping turtle?

  11. As long as it isn’t the one where I get my tires.

    “I calculate a 93.456783% probability you can drive around just fine with three flat tires.”

  15. Per Iowahawk: “Your daily reminder: all political polls are still garbage. #9percentresponserate”
    He has a point.

  16. He’ll be declared both. The reasons for the thumping Obama gets will be just as incomprehensible as the details of the attack on Benghazi. The MSM will appoint a special commission to investigate how Americans malfunctioned so badly and unexpectedly, and they’ll probably put Nate in charge of it.

  25. Ali gets off track by the assumption that the 47% of early voters who did not vote in the last party primaries are independents. This is not a valid assumption at all. By the time Ohio rolled around, Romney was the clear leader and the last opposition was Santorum. Obama had no opposition, so the other side’s last primary was for state and congressional seats only, and the statewide offices aren’t up.

    In fact, the odds are that most of those 47% are party members, but one would have to check back on previous primaries to check that. It’s the party members who’ve been pushed, cajoled, and ridden to death to vote early, pure independents not as much.

    So his analysis fails on that basis. There have been other measures which tend to confirm that the GOP is in relatively far better shape than in 2008, but without relying on such huge assumptions.

  30. You clowns are so funny. Nate Silver correctly predicted R gains in the 2010 midterms, and you bozos were all over his “lies” then, right?

Comments are closed.