Monday, November 5, 2012

The Omnivore's Electoral Map

Since everyone else is doing it ... here is the Omnivore's prediction:
A Middle Of The Road Prediction
Why did I go this way?

It's a tossup according to the polls. Although Debbie "Downer" Wasserman-Schultz claims Dem victory in Florida, I am being more conservative here and assuming that Obama doesn't win it in "one-fell swoop." Romney has spent a lot of time here. It's crucial, and the polls show a Romney tilt. Obama has a good ground game--but Romney does this time around too and he has focused here. It's also worth noting that in Florida, again, we are already in the courts around early voting.

Virginia and Ohio
Virginia is a big deal: as Nate Silver quipped: if Obama wins OH and VA it's 'OHVA' (get it? GET IT?). VA is in the toss-up zone with Obama showing a persistent lead (if, mostly, a narrow one) in all recent polls. So I am giving it to him. Combined with Ohio, this is enough to give Obama the win before we pass into Central Time.

I clocked Ohio in for Obama: the Romney narrative about expanding the map (that is: he's won most of the swing states so now he's going for wins in hard-to-get states like PA and MN) doesn't ring any truer to me than Debbie's claim the Dems have won Florida: there, simply put, is no reason to believe anything that comes from a campaign--no campaign is going to straight-talk you (even the late, lamented straight-talk-express, alas) and the "behavior" of running ads in other states is completely neutral (Romney is hugely funded and a hail-marry attempt to win a blue state is indistinguishable from "expanding the map" from where I'm seated).

Iowa and Wisconsin
I dithered on Iowa. Wisconsin shows a much heavier favorite for Obama than Iowa--in the end, I flipped a coin and gave both to Obama. If I had to kick a state out to Romney, it would be Iowa.

Colorado and New Hampshire
Colorado I gave to Romney. Things are tight there and while the polling shows them less tight in NH, I figure that Romney is going to win some-but-not-all of the toss-ups he's close in. So Colorado goes Red. NH goes Blue.

Nevada and Oregon
The night ends with a whimper, not a bang as these states, in my guess, go blue. Why? The polling suggests Team Blue has a not insignificant lead here. While I had to pick, it doesn't matter: after EST the election, in this model, is over.

What Do I Think?
I'm on the edge of my seat along with everyone else wondering if it'll be complete poll failure or if the polls will turn out to be right. My picks suggest they'll be pretty much on target so, well, you know what I think--but there's a chance it could be POLLS-LIE. That's what the Red-Team is saying. So we'll see.

But here's a couple things to note:

  1. The National-Polls Normalization Scenario Happened ... kinda: the conservative theory was that the state polls would come in line with the national polls. That has, kinda, happened. But not the way that it was supposed to. Today the Pew poll shows Obama +3. Every recent poll shows either a national tie or Obama up +1 to +3. 
  2. suggests that pollsters might be "herding." This is where they "cheat" and make sure their polls are in line with everyone else's. The math suggests: maybe--not by much, but certainly possible. It's worth looking at and thinking about. No one wants to look stupid on election day. Bonus question: if they're herding now ... when did they start?
One more thing: I have been really, really troubled with a lot of the Conservative predictions this cycle. Yes, they may turn out to be right--and if they do, fine: they will have made a case for (a) the failure of polls and (b) the pervasive slant of the media narrative. That's actually a good thing: if it happens their way eyes will be opened. But not, they are contributing to the war on math and science and that's not good. I'll write more about this later but I wanted to point to some analysis I actually liked: The Daily Caller: For the most part, these are measured, intelligent predictions. For the most part they don't throw out polling (although one or two pundits predict a Romney Landslide) but rather show a close race with, mostly, a Romney victory. That's fine--it certainly could happen that way. There's one though--that I'm going to quote here (from Gregg RE):
Nate Silver doesn’t realize that the models and simulations he tirelessly iterates to the immense satisfaction of his superiors actually have a tragic, practical purpose. On Nov. 6, the electoral map will precisely match Silver’s state-by-state predictions, revealing to everyone in the New York Times’ secret command school that his gift for destroying the Republican threat is as real as it is powerful.
“I will program your battles now, not the computer,” General Petraeus will tell the former baseball statistician, as CNN holograms systematically call
Pennsylvania, Virginia, Ohio and Wisconsin for Obama, by massive margins. “From now on the enemy is more clever than you. From now on the enemy is stronger than you. From now on you are always about to lose… but you will win. You will learn to defeat the enemy.”
For all the world like Ender Wiggin, Silver will immediately leave his jeesh and enter into a deep depression. Mankind, lacking a new enemy half as fearsome as Mitt Romney and the Republicans, will begin waging brutal, neverending war against itself. Millions will die, and Silver will become hunted for his talent.
Years later, Silver will come across an egg laid by Ann Romney on Newt Gingrich’s moon base. Through telepathy, Silver will learn that Romney and the Republicans had come in peace. Taking the egg in hand, with a grin so wide it will threaten to split his face into two bloody halves, Silver will promise the queen that he will help start the Republican Party again on a new, more hopeful planet elsewhere in the Solar System.
When I say we need more of this in political reporting I mean we really need more of this in political reporting. 

No comments:

Post a Comment