FROM 2012 INTO 2013 POLITICS THREAD

Re: THE 2012 POLITICS THREAD

It's not so much that they have to be flipped. It's that the assumptions that pollsters have been making have to turn out to be wrong in Romney's favor about as much as possible.

 

Re: THE 2012 POLITICS THREAD

You are correct.

And, yet, it will not matter one bit.

 

Re: THE 2012 POLITICS THREAD

And I predict this: If Obama wins, Silver will still get excoriated because he "influenced the voters."

 

Re: THE 2012 POLITICS THREAD

OK, retired Admirals and Generals. The guy you support wants less government. So pony up your pensions.

 

Re: THE 2012 POLITICS THREAD

Less government except for defense.

 

Re: THE 2012 POLITICS THREAD

Yep, science and learnin' is bad. Plus, he might be a sissy. And he prefers baseball to football.

 

Re: THE 2012 POLITICS THREAD

I think the Washington Generals are now 0 - 41 years or something like that.

Supposedly, they do play the Trotters competitively, but since some of the routines involve the Globies scoring baskets, that's how they win all the time.

I went to a "game" a couple of years ago, and the Trotters only won by like, 4 points or something like that. One or two things go wrong, and the arena might have rioted.

 

Re: THE 2012 POLITICS THREAD

The right-wingers will certainly say that.

 

Re: THE 2012 POLITICS THREAD

Good read here on Politico.

Unsurprisingly, it rips Politico to shreds

http://www.thebaffler.com/past/come_on_feel_the_buzz

In other news: I went to early vote today and the line was circled around the building. Twice. I'll vote at my regular polling place in the morning.

 

Re: THE 2012 POLITICS THREAD

I'm inclined to agree with Silver when he says that Romney has a chance, but given the relatively few paths to a Romney victory (citing that NY Times infographic previously posted) it hinges on the state polls all exhibiting a systemic bias, and further that that bias is in favor of Obama (when it could just as easily favor Romney). Hence the low (~15%) probability of a Romney victory. If Romney pulls it off, though, the pollsters and modelers are all going to have to completely reevaluate their approaches to data gathering and processing. I just don't see that happening.

 

Re: THE 2012 POLITICS THREAD

The issue is that the pollsters have to reevaluate every election.

A poll can give you a remarkably confident report of what percentage of people would respond the same way as their sample did in the same circumstances. Give me 1000 random phone numbers in Ohio, and let me call and ask those who answer who they plan to vote for, and I can be incredibly confident that the answers I get will look pretty similar to if I called every phone number in Ohio.

But, unique to election polling, nobody's interested in response. We're interested in predicting behavior. There's a difference between answering a poll question and physically going out to a polling station and casting a ballot.

So election pollsters have to devise all sorts of statistical tricks and screening questions to try to guess how the reported numbers will show up in voting. They have to change these every year based on changing moods and demographics.

They have a pretty good educated guess about how to do the conversion, but it's still just a guess and there is a certain amount of error. Historically, in national elections, it can be as much as 3 percentage points either way. So if it just happens to be about 3 percentage points in Romney's direction, he can still win.

 

Re: THE 2012 POLITICS THREAD

Silver is far from the only polling aggregator/modeler out there. Some have much higher percentages on the likelihood of Obama winning.
The thing is, in a close election, the polls can be "right" and still be wrong. If a poll calls for a one or two point Obama victory and Romney wins instead or vice versa, that's math, not error. But if they're ALL off by 2 points or more, that IS error.
As of today (I think Rasmussen does one tomorrow, too and the ABC/Post tracker has yet to come out), all the polls are either Obama plus one or more, tied, or Romney plus one (Rasmussen and Gallup). Frankly, I just don't see how the aggregators come out with such a certainty of Obama winning off those numbers, I mean, I do see how they do it. I just don't believe it. I guess politically speaking I'm just a Derek Jeter IS a Gold Glove shortstop guy.