Saturday, November 3, 2012

Two Worlds: Two Narratives

My biggest nightmare for Tuesday is that at the end of the night ... we don't know who won. I'm ready for this to be over--one way or another. However, whether Romney or Obama wins there are going to be a substantial percentage of people who think it couldn't have happened. Why? Because there are two very different worlds and whichever one you live in determines which candidate you think is winning.

Last time we looked at poll unskewing. Today it's election predictions.

On The Right
Here is Michael Barone, Fox political analyst and Sr. Political Analyst for The Examiner: "Romney Wins Handily". He calls Florida, Ohio, and Virginia for Romney ... among others.
Bottom line: Romney 315, Obama 223. That sounds high for Romney. But he could drop Pennsylvania and Wisconsin and still win the election. Fundamentals.
Barone is not the court-jester Dick Morris is: he's a reasonably respected voice in conservative politics.

Slightly more jester-iffic is the state map from Unskewed
Crimson Tide
In addition to FL, OH, and VA, he throws in Michigan and Iowa (and OR, NV, PA, and The Kitchen Sink's 1 EV for good measure).

On The Left
On the left, Princeton Election Consortium finds:
President Obama is peeling away. As you can see from the electoral vote (EV) estimator, he is the candidate with the momentum, not Romney. In terms of EV or the Meta-margin, he's made up just about half the ground he ceded to Romney after Debate #1. And the indicators are still headed straight up.
The Red Zone is the Likely Obama EV, Yellow is within, I think, 95%
 He gives Romney 6% chance to win.

FiveThirtyEight puts Obama at just over 80% chance to win. Here's his current state snapshot:
He estimates Obama at 305 EV
The other EV vote estimators vary between around 330 to 280 for Obama. Election, which is ideologically right wing--but uses the polls--puts Obama at a barely-winning 280 EV today.

What Does It Mean?
Votamatic takes a look at the prospect of poll bias:
But here’s a simple test. There have been hundreds of smaller organizations who have released fewer than a half-dozen polls each. Most have only released a single poll. We can’t reliably estimate the house effects for all of these firms individually. However, we can probably safely assume that in aggregate they aren’t all ideologically in sync – so that whatever biases they have will all cancel out when pooled together. We can then compare the overall error distribution of the smaller firms’ surveys to the error distributions of the larger firms’ surveys. 
If the smaller firms’ errors are distributed around zero, then the left-leaning firms are probably actually left-leaning, and the right-leaning firms are probably actually right-leaning, and this means that they’ll safely cancel each other out in my results, too. On the other hand, if the smaller firms’ error distribution matches either the left-leaning or the right-leaning firms’ error distribution, then it’s more likely the case that those firms aren’t significantly biased after all, and it’s the other side’s polls that are missing the mark.
Pretty!! Gray line is smaller firms ...
I think what this means is this: right now there is reason to think that there are methodologically left and right leaning polls and that that probably tracks somewhat to ideology. If that's true then ... averaging should work reasonably well.

On the other hand, it's quite possible that the foundational models that pollsters are using are, as the un-skewers suggest, quite wrong. There's no real way to know until election day.

What Do I Think?
If my biggest fear for Tuesday is that we don't know the answer by the end of the night, my second biggest is that the conspiratorial thinking that guys like Dick Morris and Unskewed polls have created on the right (assuming, of course, it isn't a Romney landslide--in which case, hey!) will delegitimize the election in many people's view. This will have impacts for years to come and within the losing party. My lean right now is towards the Obama-is-in-the-lead because I gravitate towards the assumptions that (a) pollsters are measuring something of value (even if, yes, with a margin of error) and (b) their ideological slants cancel each other out--in other words, I'm siding with the quants--but honestly? I could see it going either way (and, to be fair, even if I give 538 the nod a 20% chance of winning isn't nothing).

No comments:

Post a Comment