The election is close and so everyone is nervous. Well, everyone except Dick Morris who is convinced that Romney will win in a landslide as Oregon is now in play (I'm not kidding).
Obama is leading according to the polls. It is roughly tied in the national poll averages and Obama has clear leads in the polling averages in enough swing states to put him over the top for re-election.
So, if you're supporting the President, you have just two things to worry about:
1) Something changes
2) The polls are wrong
Are either or both of these things possible? Absolutely. Are they likely? No. But let's take a closer look at each possibility.
Something changes. What could change in the final days? Some argue that nothing is going to change. Sam Wang, for instance, points out that the cake is now fully baked. He provides statistical proof for this and I find the phrase oddly compelling so I figured I'd repeat it. But it is always possible something could change. For one, a jobs report is going to be released in a few hours. Economists expect that the unemployment rate will tick up to 7.9%. That's probably not enough to damage Obama significantly. But it is an awfully close race and what if it ticks up to 8.0%? Are the optics of that sufficiently bad to do damage? Doubtful, but possible. (UPDATE: The jobs report was better than expected with 171,000 jobs added but the unemployment rate did tick up to 7.9% as more workers re-entered the workforce). What about the so-called incumbency effect? Many (mostly Republicans) argue that most of the undecided voters will break for Mitt Rommey because if they were voting for the President, they'd have been with him already. Is this likely? No. Mark Blumenthal picks that argument apart pretty well.
What about some other wildcard? Hurricane Sandy seems like the last, biggest wildcard and that does not seem to have done any damage to President Obama. Arguably, it has made him look good. Mitt Romney seems to think Pennsylvania is in play and he's headed there for a campaign stop on Sunday. Don't get me started on that. Overall, I think the jobs report is Romney's best bet in this category.
The polls are wrong. The polling data is much more clear than it appears to the naked eye. Sam Wang, Simon Jackman, Nate Silver, Drew Linzer, and others have modeled the probability that Obama does, in fact, have a lead according to the polls and the general consensus is ... "YEAH." But the key phrase in all that is "according to the polls." If the polls are all systematically biased in some similar kind of way, that lead would be a mirage. How often does this happen? Not often. But it is important to remember that polling is at least partially an art. In this era of low response rates, it is not quite as easy as before to ensure you have an unbiased sample. If Obama's lead (and he DOES lead in the key state of Ohio in the polls) were bigger, this would be less of a concern. But it is a small lead.
Of course, the possibility of error goes in both directions. If you watched the video of Stan Greenberg I posted a few days ago, you know there are those who believe the pollsters are undersampling Obama voters. In addition, the Obama campaign claims their superior ground game will help them out-perform the polls. Of course, the Romney campaign claims the same things. The reality is that pollsters are smart people and have staked their reputation on what they're doing. The odds are they are right. Systemic polling failure is possible ... but unlikely.
In the end, there is enough uncertainty to worry about both of these things: Something could change. The polls could be wrong. How much should you worry? As of tonight, Nate Silver has Obama as an 81% bet to win.
That's a high point in the post-Denver (first debate) campaign for Obama. And as Nate points out, you'd rather be holding Obama's cards right now than Romney's.
Showing posts with label Drew Linzer. Show all posts
Showing posts with label Drew Linzer. Show all posts
Friday, November 2, 2012
Saturday, October 27, 2012
Where are the National Polls?
We all see the divergence between the national tracking polls, which generally show Romney with a small lead on average, and the state-level polls, which generally show Obama with small leads in most of the key swing states on average.
How could these two things be true at the same time? There's actually two things going on here:
First, it is entirely possible that the President is winning by narrow margins in swing states like Ohio and Colorado but is losing by more in "solid Romney" states than he is winning in "solid Obama" states. That would translate into a national popular vote deficit but, potentially, an Electoral College victory. Most of the sites that create election-prediction models like Sam Wang's Princeton Election Consortium and Drew Linzer's Votamatic use state-level polls either exclusively or more heavily than national polls because the state-level polls provide more information that is useful. They are generally right to do this in terms of predicting what matters - the Electoral College outcome. But those national tracking polls are still unpleasant to look at ... which brings me to my second point ...
Second, it is really striking to me how few "full" national polls there have been in this campaign, especially in the final weeks of the campaign. There are at least 8 different national "tracking" polls that are released just about every day. But these are not the same as a regular poll. I can't claim perfect expertise on the impact (or not) of this phenomenon but it does seem to me there is some important effect of looking at only tracking polls as opposed to regular "full" national polls. Tracking polls take a smaller sample every night for 3 or 4 or 5 or even 7 (in the case of Gallup) nights. They put these together to create a national polling number. Then, when a new night comes on board, the sample from the oldest night drops out. What you end up with is a "rolling average" of the most recent nights. One potentially problematic aspect of this is that, because each night is a distinct sample and because the pollster wants to keep the demographics in the full sample correct from night to night, you have to weight some very small samples night to night before adding them to the sample. This would seem to introduce some error as opposed to doing a regular 2 or 3-night poll and then weighting the whole sample from there. In effect, you'd need to do less weighting as you'd have stronger sub-samples to work with.
Now, I'm just kind of spit-balling on all that and I could be wrong. But I do think that regardless, it is notable that there have been so few full national polls. Indeed, there has not been one full national poll that I can find since the third debate. I'd be interested to see one of those from a good pollster who does live interviews with cell phones included.
UPDATE: In a post today, Nate Silver speaks to the question of the split between the national tracking polls and the state-level polling. The bottom line is that he says the state-level numbers are suggesting leads for Obama in swing states like Ohio that are statistically very meaningful. Mark Blumenthal makes some similar arguments in Pollster.com
How could these two things be true at the same time? There's actually two things going on here:
First, it is entirely possible that the President is winning by narrow margins in swing states like Ohio and Colorado but is losing by more in "solid Romney" states than he is winning in "solid Obama" states. That would translate into a national popular vote deficit but, potentially, an Electoral College victory. Most of the sites that create election-prediction models like Sam Wang's Princeton Election Consortium and Drew Linzer's Votamatic use state-level polls either exclusively or more heavily than national polls because the state-level polls provide more information that is useful. They are generally right to do this in terms of predicting what matters - the Electoral College outcome. But those national tracking polls are still unpleasant to look at ... which brings me to my second point ...
Second, it is really striking to me how few "full" national polls there have been in this campaign, especially in the final weeks of the campaign. There are at least 8 different national "tracking" polls that are released just about every day. But these are not the same as a regular poll. I can't claim perfect expertise on the impact (or not) of this phenomenon but it does seem to me there is some important effect of looking at only tracking polls as opposed to regular "full" national polls. Tracking polls take a smaller sample every night for 3 or 4 or 5 or even 7 (in the case of Gallup) nights. They put these together to create a national polling number. Then, when a new night comes on board, the sample from the oldest night drops out. What you end up with is a "rolling average" of the most recent nights. One potentially problematic aspect of this is that, because each night is a distinct sample and because the pollster wants to keep the demographics in the full sample correct from night to night, you have to weight some very small samples night to night before adding them to the sample. This would seem to introduce some error as opposed to doing a regular 2 or 3-night poll and then weighting the whole sample from there. In effect, you'd need to do less weighting as you'd have stronger sub-samples to work with.
Now, I'm just kind of spit-balling on all that and I could be wrong. But I do think that regardless, it is notable that there have been so few full national polls. Indeed, there has not been one full national poll that I can find since the third debate. I'd be interested to see one of those from a good pollster who does live interviews with cell phones included.
UPDATE: In a post today, Nate Silver speaks to the question of the split between the national tracking polls and the state-level polling. The bottom line is that he says the state-level numbers are suggesting leads for Obama in swing states like Ohio that are statistically very meaningful. Mark Blumenthal makes some similar arguments in Pollster.com
Labels:
Cell phones,
Drew Linzer,
Mark Blumenthal,
Nate Silver,
Sam Wang
Sunday, October 21, 2012
Nevada, Wisconsin, Ohio ... Oh My
There's a lot of good articles floating around in the last couple of days about how close the national polls are and how President Obama has an advantage in the state level polling in swing states. A particularly good piece was posted on Huffington Post by Simon Jackman of Stanford outlining a statistical model of both the present national vote/EV split and the history of this split (statistically-speaking).
It got me to thinking about how to express Obama's advantage in the simplest terms, so here it is: Wisconsin, Nevada, and Ohio.
Now let me flesh that out a bit. Take a look at the map below. It shows what I believe to be the state of the race in the most general terms.
Every red state and blue state on the map is highly likely to go to Romney (red) or Obama (blue). Put another way, I don't think there is a single red or blue state on that map that any reasonable person would dispute.
Obama has 237 electoral votes effectively "in the bank" and Romney has 191. There are 9 "battleground" states totalling 110 electoral votes. I've ranked these 9 battleground states below in order of their probability of going for Obama according to Nate Silver's blog:
WI - 79.0%
NV - 73.1%
OH - 70.3%
IA - 65.9%
NH - 62.6%
CO - 52.9%
VA - 46.9%
FL - 32.6%
NC - 15.0%
Notice the three at the top of the list; Wisconsin, Nevada, and Ohio. By the way, Drew Linzer also estimates that Obama would win these three states today predicting the President would get 51.7%, 52.0%, and 51.1% respectively in Nevada, Wisconsin, and Ohio.
Now you can see why I've been saying since ... forever ... that Ohio is the whole ballgame. Let's assume Obama wins the three of those battlegrounds he's most likely to win; Wisconsin, Nevada, and Ohio. Here's the new map:
Obama is now at 271, 1 more than what he needs to win. It isn't just that he needs 3 states out of 9. It is that he needs the 3 states that he's most likely to win. Unless Romney can pick off at least one of these states, he can't win.
It got me to thinking about how to express Obama's advantage in the simplest terms, so here it is: Wisconsin, Nevada, and Ohio.
Now let me flesh that out a bit. Take a look at the map below. It shows what I believe to be the state of the race in the most general terms.
Every red state and blue state on the map is highly likely to go to Romney (red) or Obama (blue). Put another way, I don't think there is a single red or blue state on that map that any reasonable person would dispute.
Obama has 237 electoral votes effectively "in the bank" and Romney has 191. There are 9 "battleground" states totalling 110 electoral votes. I've ranked these 9 battleground states below in order of their probability of going for Obama according to Nate Silver's blog:
WI - 79.0%
NV - 73.1%
OH - 70.3%
IA - 65.9%
NH - 62.6%
CO - 52.9%
VA - 46.9%
FL - 32.6%
NC - 15.0%
Notice the three at the top of the list; Wisconsin, Nevada, and Ohio. By the way, Drew Linzer also estimates that Obama would win these three states today predicting the President would get 51.7%, 52.0%, and 51.1% respectively in Nevada, Wisconsin, and Ohio.
Now you can see why I've been saying since ... forever ... that Ohio is the whole ballgame. Let's assume Obama wins the three of those battlegrounds he's most likely to win; Wisconsin, Nevada, and Ohio. Here's the new map:
Obama is now at 271, 1 more than what he needs to win. It isn't just that he needs 3 states out of 9. It is that he needs the 3 states that he's most likely to win. Unless Romney can pick off at least one of these states, he can't win.
Labels:
Drew Linzer,
Nate Silver,
Nevada,
Ohio,
Simon Jackman,
Wisconsin
Subscribe to:
Posts (Atom)


