Some polls now have Romney ahead.

[Todd33]

He's not down that much.

If Romney is up 3 in popular vote then Ohio won't matter.

ABC just posted their poll

Obama was up 3 on 10/10 to 10/13 but now Romney is up 1.

You seem to have a serious disconnect on how this works. You should not hold onto any single poll, let alone a national one.

You need to do some homework on pol aggregators and statistics.

 

[buckshot24]

You seem to have a serious disconnect on how this works. You should not hold onto any single poll, let alone a national one.

You need to do some homework on pol aggregators and statistics.

I know exactly how this works. Where am I suggesting a single poll is right?

Here's a hint for you. Just because I want Romney to win isn't evidence that every post I make is attempting to prove that he will.

 

[buckshot24]

Yes, most of whom live in states where the outcome is already settled. Only the swing states matter.

Romney appears to have gotten a 2 point bounce from the first debate. Most of these national polls have him a head because of LV screens. And I think a lot of it is also right-wingers in red states coming around to voting for the stooge, which is meaningless.

Are you now saying the pollsters don't know how to run their polls but you do?

Would you agree that if Romney wins the popular vote by 3 points that he will most likely win at least 270 EC votes? Ohio won't matter if this is the case.

Also Gallup and Rasmussen have Romney up 4 in the swing states that doesn't include a single "red state". You're wrong Charles.

 

[Charles Kozierok]

Are you now saying the pollsters don't know how to run their polls but you do?

No, I'm saying that most people don't understand how to interpret the pollsters' polls.

Would you agree that if Romney wins the popular vote by 3 points that he will most likely win at least 270 EC votes?

Likely, yes. But not necessarily. I trust the state polls more, and I don't think Romney is up by 3 points in the PV.

Also Gallup and Rasmussen have Romney up 4 in the swing states that doesn't include a single "red state".

Rasmussen always has a heavy Republican lean. I haven't seen Gallup claim that Romney is up 4 in the swing states.

Right now Gallup's national poll has Romney up 51-46, but Obama's approval rating at 51-45. Do you think that makes any sense? I do not.

 

[Farang]

http://nymag.com/daily/intel/2012/10/romney-says-hes-winning-its-a-bluff.html

I know Chait's a liberal, and as an Obama supporter I want to careful not to descend into the delusion some conservatives did in late September (unskewed polls anyone?), but this is basically my take on the race at this point. Though we're both biased I think the facts are there..

Romney needs to do better than tied in VA, CO, and needs to overcome deficits in OH, IA, and WI. Not ALL of these things, of course, but unless any of them are happening, he isn't winning.

 

[jpeyton]

Are you now saying the pollsters don't know how to run their polls but you do?

Would you agree that if Romney wins the popular vote by 3 points that he will most likely win at least 270 EC votes? Ohio won't matter if this is the case.

Also Gallup and Rasmussen have Romney up 4 in the swing states that doesn't include a single "red state". You're wrong Charles.

Tisk tisk, continuing to cherry pick polls.

RCP's averages have Romney up in two swing states: Florida and Virginia.

They have Obama up in every other swing state. Obama's early voter turnout is running ahead of his 2008 numbers: http://www.politicususa.com/bad-news-romney-ohio-early-voting-turnout-obama-gop.html

 

[Doppel]

http://nymag.com/daily/intel/2012/10/romney-says-hes-winning-its-a-bluff.html

I know Chait's a liberal, and as an Obama supporter I want to careful not to descend into the delusion some conservatives did in late September (unskewed polls anyone?), but this is basically my take on the race at this point. Though we're both biased I think the facts are there..

Romney needs to do better than tied in VA, CO, and needs to overcome deficits in OH, IA, and WI. Not ALL of these things, of course, but unless any of them are happening, he isn't winning.

Fake it till they make it. EC is still substantially behind on Romney and as time gets closer it just becomes more of a lock, as many are voting every single day.

 

[buckshot24]

No, I'm saying that most people don't understand how to interpret the pollsters' polls.

How should we interpret these likely voter polls then?

Likely, yes. But not necessarily. I trust the state polls more, and I don't think Romney is up by 3 points in the PV.

OK, fair enough.

Rasmussen always has a heavy Republican lean. I haven't seen Gallup claim that Romney is up 4 in the swing states.

http://www.usatoday.com/story/news/...tates-poll-women-voters-romney-obama/1634791/

2008 Rasmussen wasn't heavy Republican.

Right now Gallup's national poll has Romney up 51-46, but Obama's approval rating at 51-45. Do you think that makes any sense? I do not.

Sure it does. They ask seven questions about their likelihood of voting. According to their models more people are going to show up (right or wrong) and support Romney. The job approval rating doesn't take this into account. Registered voters are much closer in Gallup's poll.

 

[Charles Kozierok]

2008 Rasmussen wasn't heavy Republican.

In 2010 he was. And he appears to be now as well.

Sure it does. They ask seven questions about their likelihood of voting. According to their models more people are going to show up (right or wrong) and support Romney. The job approval rating doesn't take this into account. Registered voters are much closer in Gallup's poll.

That's a 12-point discrepancy. I'd say more of a problem with Gallup's LV model than with Obama's polling.

 

[buckshot24]

Tisk tisk, continuing to cherry pick polls.

RCP's averages have Romney up in two swing states: Florida and Virginia.

They have Obama up in every other swing state. Obama's early voter turnout is running ahead of his 2008 numbers: http://www.politicususa.com/bad-news-romney-ohio-early-voting-turnout-obama-gop.html

Nice spin article. Only taking what the Obama camp is saying while not contacting Romney's people at all. Hackery.

We were talking about the national polls and he makes some stuff about Red state voters pulling for Romney. The national polls that have Romney up big are Gallup and Rasmussen for the most part. Me posting the swing state data refutes his contention that all the national bump is coming from "red states".

 

[buckshot24]

In 2010 he was. And he appears to be now as well.

Why don't you compare apples with apples? 2010 was an off year election.

That's a 12-point discrepancy. I'd say more of a problem with Gallup's LV model than with Obama's polling.

So now you're saying you know more about polling than pollsters. Maybe you should give them a call and offer to help.

 

[buckshot24]

Fake it till they make it. EC is still substantially behind on Romney and as time gets closer it just becomes more of a lock, as many are voting every single day.

I'm sorry but this is just silly. Those who are voting early would vote the same way on election day. I could have voted January 20 2009 for instance.

 

[Farang]

I'm sorry but this is just silly. Those who are voting early would vote the same way on election day. I could have voted January 20 2009 for instance.

That's one way to discount it. However Obama wins among non-voters, so there is substantially more for a ground game to draw from on the Democratic side. A lot of the early voters could be people that traditionally you couldn't get out of bed on election day.

 

[Charles Kozierok]

So now you're saying you know more about polling than pollsters.

No, but I know more about it than you.

Gallup's LV model has been controversial this year and the subject of much debate and discussion (none of which, I'm betting, you are aware of). For one thing, it takes into account rather subjective measures of "enthusiasm". The Gallup poll has been an outlier for much of October.

You might find this worth perusal.

 

[buckshot24]

Gallup's LV model has been controversial this year and the subject of much debate and discussion (none of which, I'm betting, you are aware of). For one thing, it takes into account rather subjective measures of "enthusiasm". The Gallup poll has been an outlier for much of October.

Your little barbs aside, being controversial isn't being wrong. You're still saying you know more about polling than Gallup.

You might find this worth perusal.

They were waiting for the other shoe to drop based upon the second debate. Nothing changed. Romney's vote is more motivated to vote which is what I have been saying.

 

[Charles Kozierok]

Your little barbs aside, being controversial isn't being wrong.

No, it doesn't. But what I'm saying -- as Nate Silver is saying -- is that Gallup's LV model is an outlier. And while outliers sometimes are correct when everyone else is wrong, that's not usually how it works out.

Romney's vote is more motivated to vote which is what I have been saying.

If true, that should be showing up in the state polls. It's really not. We'll see in two weeks, but if the state polls still show Obama ahead and the Gallup/Rasmussen nationals still show Romney ahead, I don't think Romney will have a happy Election Night.

 

[buckshot24]

No, it doesn't. But what I'm saying -- as Nate Silver is saying -- is that Gallup's LV model is an outlier. And while outliers sometimes are correct when everyone else is wrong, that's not usually how it works out.

A model can't be an outlier. Data points can.

This is just a quibble and it doesn't really have anything to do with the race.For something to be labeled an outlier it takes more than it just being different. You have to justify when you dismiss data points that seem out of whack.

A true outlier isn't merely something that disagrees with the rest of the data.

All of this being said, there IS agreement with Gallup.

Gallup has the race 51-46 Romney
Rasmussen has the race 50-46 Romney
Monmouth/SurveyUSA/Braun 48-45 Romney
Politico/GWU/Battleground 49-47 Romney

There is a lot of agreement that Romney has a 2-5 point lead.

There are polls that say Obama is up however but that doesn't mean there is no agreement with Gallup.

 

[Charles Kozierok]

A true outlier isn't merely something that disagrees with the rest of the data.

That is the definition of an outlier. Look it up. They can result in different ways, but it just means something outside the normal distribution.

All of this being said, there IS agreement with Gallup.

And a couple of days ago, Gallup was at +7 for Romney when Rasmussen was at +1 and IBD was at -6 and another poll was at -2 or something.

The numbers will move around. And Rasmussen will always read high for Romney because of his Republican bias. The other polls show a much tighter race at the national level.

And, once again -- I see no evidence that Romney is closing the gap at the state level. He peaked a week ago there. If you want to convince yourself that Romney's ahead, be my guest.

 

[buckshot24]

That is the definition of an outlier. Look it up. They can result in different ways, but it just means something outside the normal distribution.

I don't need to look it up. In analyzing data in science you can't simply say "its different" so it must be an outlier and we can ignore it.

And a couple of days ago, Gallup was at +7 for Romney when Rasmussen was at +1 and IBD was at -6 and another poll was at -2 or something.

That doesn't change the fact that you are calling a poll an outlier when there is some agreement with the other polls.

 

[Charles Kozierok]

I don't need to look it up. In analyzing data in science you can't simply say "its different" so it must be an outlier and we can ignore it.

I didn't say we can ignore it. I said: "[W]hile outliers sometimes are correct when everyone else is wrong, that's not usually how it works out".

Which is what quite a few statisticians analyzing the polls are saying.

If the other polls stabilize at Romney being in the +5 or +6 range over the next few days, then I'll be happy to admit I was wrong and reassess. But if that happens without significant movement in the state-level polls, it will just leave the situation muddled.

 

[buckshot24]

I didn't say we can ignore it. I said: "[W]hile outliers sometimes are correct when everyone else is wrong, that's not usually how it works out".

Make up your mind! If it is an outlier then we can ignore it. But you keep calling it an outlier. I did say it was a quibble....

Which is what quite a few statisticians analyzing the polls are saying.

The case for Gallup being an outlier has diminished greatly over the last few days. You can't really call it that anymore. There is agreement.

If the other polls stabilize at Romney being in the +5 or +6 range over the next few days, then I'll be happy to admit I was wrong and reassess. But if that happens without significant movement in the state-level polls, it will just leave the situation muddled.

I think you're misunderstanding me here. Gallup and Rasmussen and all the rest that have him up 2+ points may be dead wrong. Whatever I say here won't change the vote count come November 6th.

 

[buckshot24]

Intrade is moving towards Romney.

54.7 - 45.5

 

[TraumaRN]

Intrade is moving towards Romney.

54.7 - 45.5

And that holds very little bearing on the outcome of this race...state polls need to move for Romney.

 

[jpeyton]

And that holds very little bearing on the outcome of this race...state polls need to move for Romney.

It has even less bearing on the outcome considering that it still predicts an Obama win, just like it did yesterday (and every other day this year).

 

[buckshot24]

And that holds very little bearing on the outcome of this race...state polls need to move for Romney.

I agree that intrade means nothing. It was thrown in my face a few times as if it did.