Friday, October 19, 2012

Robo-Polled by Rasmussen!

Rasmussen (the Only Unskewed Pollster(TM)) called The Omnivore's household last night with a few short questions. Was I following the election? Yeah? How closely? Was I likely to vote? Really likely? Things sometimes happen on election day--might I not vote if 'something came up'? And so on.

I was asked if I was a Republican or Democrat--not which party I identified with--but which I was registered with. That's usually the same for most people--but not always the same. I believe they sometimes switched Obama and Romney around on the "Press 1 for Obama and Press 2 for Romney" questions. Maybe to make sure I wasn't just pressing the 1-key over and over.

They didn't say if it was a national poll or a Florida poll but I guess we'll find out.

Some Thoughts
We are awash in data (numerous polls, betting markets, the RAND study, things like the Twindex and PoliticIT's Internet ratings) and so on. The press is taking the safe rout and saying it's very very close. Various Electoral Vote measurements put Obama at around 280-290--a win if he performs "on average" but he could lose the national vote.

A recent Gallup Poll shows Romney up +7 nationally--a huge lead. Their poll just post 2nd-Debate broke down demographically to Obama +4 to +6 in the East, Midwest, and West ... and +22 in the South: a massive lead but with a certain cap on EV (all those excess votes in the South don't give more EVs). If that's true then it means that while some demographics give Romney a massive lead the overall picture doesn't give him that much. That may be happening in other national polls: I don't know.

The Monkey Cage notes that the betting markets are sort of converging on the EV-prediction sites (like 538) and notes, as I did, that a lot of the bettors are reading 538. I think this is/has to be clearly true--however there are a few other things to note:

  1. Betting markets move much faster than polling cycles. We can see what people are thinking within a polling cycle or even within a few minutes (SELL PERRY!) by watching the markets.
  2. Bettors trying to make money will look for contrarian bets. If someone is betting heavily on Romney right now or heavily on Obama it's because they think they are under-rated. If the EV-sites are the "rating agency" for a candidate then the bettors are the stock traders and they will usually be acting on some knowledge that they think the EV-sites don't have (i.e. if I believe Unskewed polls I will be investing my life's savings in Romney right now. If I believe 538 and think that the market places Obama at about 65% chance to win ... it's not much of a bargain).
Finally, the Binders Of Women thing. Amazon's binder offering is now loaded with mock-reviews. I'll admit I found the site relatively funny (yes, it's juvenile but c'mon). I'm certain there are some funny sites for Obama out there--but I haven't seen them. I'm going to say that I don't think this is a distraction: it's not substantial in any way (everyone knows that Romney meant binders full of women's resumes--and he should be lauded for that if you believe it was true and that he was aiming for equality. You may ask if he did the same thing at Bain ... answer: apparently, no.)--but I don't think it's a distraction.

The Internet is interoperative. It's a conversation--and when information hits it, it responds like this. It's part of life now. The Romney campaign can't buy up every possible domain name to prevent a sneak attack--but maybe they're wishing they bought that one.

No comments:

Post a Comment