Uncategorized

30 hours: Intersections in Real Time

I strongly urge people only to dismiss information if they have a good justification for believing that it’s wrong, or not relevant. Disliking or disagreeing with it isn’t enough. So I will from time-to-time listen to conservative talk radio.

You have to hand it to these guys, Limbaugh and Beck. They are smooth talkers. It’s like they are talking to you, not a radio microphone. It’s almost smoothing and it’s very disarming. It’s why they are as popular and successful as they are.

I credit Limbaugh today with being intelligent enough to know that polls don’t make sense to him, but he doesn’t understand why. Here’s why.

A national CNN poll out recently shows the race tied, with more Democrats in the survey than Republicans by about 11 points, and Romney winning Independents by like 22 points. That doesn’t make sense. If you’re winning with indys like that, you should be way ahead in the topline horse race number. He knows that something is wrong, but not what, other than the typical misguided complaints that D+11 is higher turnout than 2008.

There’s one problem and one answer. The problem is that D+11 in a poll does not project turnout. It’s simply a breakout of people’s opinions of themselves, no different than answering if you consider yourself a Christian/Jew/Muslim, or conservative/moderate/liberal. None of those things determine how you will vote or even what you think of the race 24 hours from now, which is why those things aren’t weighted (except by Rasmussen). So even with a D+11 poll, that doesn’t by itself mean that anyone is projecting 11% more Democrats to vote than Republicans.

The answer to all of this is simple math applied to the inconsistency. The reason that a 22 point lead with Independents normally translates into a big win and overall lead is that most Independents are real Independents. But not anymore. Since the rise of the Tea Party and the disappointing of the Bush 43 generation, a *lot* of Republicans have begun calling themselves Independents, even if they aren’t.

So these days it’s not that polls have too many Democrats in them. It’s that polls have too few self-identified Republicans, because so many Republicans are identifying as Independents. That simultaneously explains why Romney has such a strong and consistent lead with self-identified Independents, yet is either tied or losing, and why the relative difference between self-identified Democrats and self-identified Republicans seems to favor Democrats.

Which is basically what I said last night, only I find it worth repeating all of this because of Rush Limbaugh is talking about this and doesn’t get it, then it’s very likely that a lot of people don’t get it also. But I don’t want you to take his explanation as the answer. It’s not tied, and some people do have a good idea of what’s going to happen.

* * *

Polls have been moving all day and I’m trying to update this graph on an hourly basis. There have been seven polls in Ohio in the past two days and I’m beginning to think that my five poll average doesn’t have enough data to reflect everything that’s happening there. But some states, like North Carolina, aren’t getting polled often enough for a seven day average to be useful.

On top of that, movement in some of these polls is deceptive. Obama just dropped in New Hampshire from earlier this morning by about 0.2 points just because one new poll that had him up by 6 points fell off the five day average (it’s now the sixth most recent poll), replaced by one that had him leading by 4 points from a completely different pollster.

If there is any takeaway so far today, it’s that Florida is a statistical tie, and Ohio is moving from tossup to leans Obama. Virginia was doing the same thing as Ohio, but now it’s slipping back a bit. Yet even Real Clear Politics has given Virginia to Obama on their no-tossup maps, giving Obama the 303-235 electoral vote consensus I was projecting on Saturday, October 27th.

I feel really good about that number. The only state that I think could fuck that up in the next 30 hours (about how long until the first polls close) is Florida going from Romney to Obama, resulting in a 332-206 Obama win.

There will be little else to do or say now. Watch that spreadsheet until the late evening hours when there probably won’t be anymore polls for this election. Then we’ll who wins: math, or faith-based polling.

* * *

These are from 2010 but should be accurate for informational purposes. DO NOT RELY ON THESE FOR YOUR OWN VOTING!

Polls close in these battleground states (EST all):

7:00 pm: Florida, New Hampshire, Virginia

7:30 pm: North Carolina, Ohio

8:00 pm: Michigan, Pennsylvania (Michigan is for those with hopes and dreams)

9:00 pm: Colorado, Wisconsin (Wisconsin, same)

10:00 pm: Iowa, Nevada

* * *

You don’t have to follow me (you’ll benefit from it) on Twitter on election night, but I strongly urge you to follow my political list. Watch TV of course, but you’ll get far more valuable information and you’ll get it much faster doing this.

You don’t need to follow it, if you’re not a Twitter user. Just load the page.

* * *

Update – An anonymous source has leaked some Romney campaign internal polling to the media, which goes a long way towards explaining why Romney is spending time and money in Wisconsin and Pennsylvania:

Romney internalsI guess nobody told the Romney campaign that internals are almost always junk.

Update 2 – Nate Silver has Barack Obama up to 91% to win the electoral college based on today’s polling, which shouldn’t surprise you if you’ve been following me around today. And he still has polls to add yet.

Update 3 – And Nate’s model just tipped Florida from ~54.x% Romney win to 52.5% Obama win. I did NOT think that would happen. It has been light red for quite a long time, really since the Denver debate. This perfectly tracks with my own analysis today which saw Florida start at Romney +1.4 and drop into a tie *twice* just this evening. It really starts to make me wonder how the Romney campaign can actually be as close as they are and yet be incompetent enough to campaign in safe blue states like Pennsylvania, Wisconsin, and Minnesota, while Virginia and Florida slip away.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.