Romney, Obama, Rasmussen

[NotJesus]

'Turd Blossom' weighs in with Romney winning by three nationwide, 279 electoral, maybe more. He also mentions the poll that has Obama up by five in Ohio:



When one poll shows a partisan advantage it's not out of the ordinary, however when (virtually) all the polls from certain firms report that, either the electorate has in fact dramatically changed, or there's something wrong with their assumptions.

Can anybody show me a 1000 or more respondent poll from the past week or two with a Dem advantage of four or less that has Obama winning either the national vote or one of the swing states? All those ones with huge Dem partisan advantages are likely just indications they didn't do a very good job separating the 'likely voters' from the 'registered voters' as there's usually a significant number of Dems that don't turn out, thus of that +8 in Ohio half (or more) of it might just be the void between the 'registered voter' and the 'likely voter' with more meticulous methodology.

What are you on about? In the last two weeks there have been 24 polls in Ohio. 18 show an Obama lead, 5 show a tie and one (Rasmussen, 10/28) shows a Romney lead of 2 points. It's not Democrats who need to cling to one poll result.

 

[Kaosium]

Here's one of the principal interior decorators on the Titanic. It's particularly useful because you can see links on the right hand column.

Not only do the polls flicking by show that his article is based on a series of lies (because it shows Obama leading and going up in the states he dismisses when he says "only in NV does Obama still lead", but it links to his other silliness... articles like Why Romney Will Win, Pennsylvania is the New Ohio*, etc...

Dick Morris...

http://www.realclearpolitics.com/articles/2012/10/31/here_comes_the_landslide_115998.html

Heh, you know what that reminds me of? There's a rumor buzzing around that the Obama campaign might have a sex scandal in the offing. Poking around it appears some think it might be John Kerry having a 'love child.' I've not been paying close enough attention to understand just why Kerry would be considered all that close to the campaign anyway, albeit he did have a final night speech at the convention and some have speculated he's on his way to Foggy Bottom if Obama is re-elected.

Why your mention of Morris brought that to mind is that late in the '96 campaign, when it looked like Clinton was cruising to an easy victory, Dick Morris was caught with a prostitute in a hotel or somesuch. There were those that thought it might derail Clinton chances, however there were others who remembered the LBJ campaign with some campaign staff getting caught in the bathroom having sex and knew it wouldn't have much of an effect.

Oh, and Greta is now covering Reince and Paul in WI this week. They show the "crowd" in Green Bay. No front audience view, just the speaker with three rows of people behind him on the podium. I'm pretty sure that was the whole turnout. I also note that they're having trouble with the scenery. Are there no Latinos or Blacks in Green Bay? They didn't have the one dark face they always position in these shots. Maybe they can get Mia Love to pose in the background. Maybe a little green screen CGI?

Green Bay gets cold as hell and it snows a lot, some people have better sense than to stick around!

Oooh, Reince is still spinning a big win in WI, but the gusto is gone. Even Greta is asking "Why is WI so close?" A week ago (she's from WI) she was cheering with them about how they were going to take the state. At least she's avoiding last week's lie - they actually claimed that the Obama team went to WI to work for the recall AND FAILED. Everyone who remembers the news of the time remembers that WI progressives were complaining of the lack of support from the big dawg.

When the Beltline gets backed up from Gammon to John Nolan on the west side by charter buses, and the east side from John Nolan practically to the interstate exit (when it first started) I'd say they got plenty of support from somewhere to get those figures of (up to) 100k on the capitol lawn for weeks.

 

[Kaosium]

What are you on about? In the last two weeks there have been 24 polls in Ohio. 18 show an Obama lead, 5 show a tie and one (Rasmussen, 10/28) shows a Romney lead of 2 points. It's not Democrats who need to cling to one poll result.

Yeah, but most of those are small sample polls with dubious partisan ID data; it's ones like this that I was looking for, shows parity between Dem and GOP voters and is 1000 or more, with a result of +2 for Obama. However that's a strange split too, usually it's more along the lines of 35 Dem, 30 GOP, and 35 Independent. Showing only 11% of the electorate as independent makes one wonder about how they determined that, it musta been counting 'leaners' but they didn't delineate that anywhere from what I could see. Not that it invalidates anything, I'd just be curious at which side had more leaners.

Polls like this, on the other hand, are more indicative Romney will win, despite it appearing to suggest Obama has a four point lead.

 

[NotJesus]

So it's the, "The polls are skewed," argument.

 

[Foolmewunz]

Oh, and Greta is now covering Reince and Paul in WI this week. They show the "crowd" in Green Bay. No front audience view, just the speaker with three rows of people behind him on the podium. I'm pretty sure that was the whole turnout. I also note that they're having trouble with the scenery. Are there no Latinos or Blacks in Green Bay? They didn't have the one dark face they always position in these shots. Maybe they can get Mia Love to pose in the background. Maybe a little green screen CGI?

I don't know if anyone else caught this over-zealousness on my part, but I found the full video and it was a considerably larger crowd. I figure no sense in distorting things in the Politics sub-forum. Everyone's so even-handed and downright Solomon-like in these election year threads.


No view of the whole place but it was several hundred, I'm sure.

http://www.realclearpolitics.com/articles/2012/10/31/here_comes_the_landslide_115998.html

(Still had no luck at my "Find the Dark Complexion" game,... anyone have any better luck?)

Just as an aside and to keep my record in tact of finding something bad to say about any GOP gathering.... Except for true believers, I'm sure no one wants to go the full sixty minutes of a Republican rally, so just bear with me and check out the first couple of minutes. When the guy introducing the evening starts out, count the cheese heads. Hey, it's Green Bay. They go to funerals in cheese heads. But then click to anywhere near the end and count the cheese heads surrounding Ryan. Only about two, maybe three, visible. Do you think they were hot and sweaty? Or do you think someone said, "Psst, ***hole! You're going to be on national tv. Take that thing off, ya dumb schmuck!"

 

[Foolmewunz]

Heh, you know what that reminds me of? There's a rumor buzzing around that the Obama campaign might have a sex scandal in the offing. Poking around it appears some think it might be John Kerry having a 'love child.' I've not been paying close enough attention to understand just why Kerry would be considered all that close to the campaign anyway, albeit he did have a final night speech at the convention and some have speculated he's on his way to Foggy Bottom if Obama is re-elected.

Why your mention of Morris brought that to mind is that late in the '96 campaign, when it looked like Clinton was cruising to an easy victory, Dick Morris was caught with a prostitute in a hotel or somesuch. There were those that thought it might derail Clinton chances, however there were others who remembered the LBJ campaign with some campaign staff getting caught in the bathroom having sex and knew it wouldn't have much of an effect.

Maybe we could get Donald Trump on the Kerry case. I'm sure that will swing as many votes as his Big Reveal did last week. I heard a rumor that Teh Donald got washed out to sea during Frankenstorm Sandy. Unfortunately, it turned out to be a rumor based on someone finding a yellow mohair toilet-seat cover floating in the drink.

(See? Anecdotes and speculation are fun!)

Green Bay gets cold as hell and it snows a lot, some people have better sense than to stick around!

I was wrong about the crowd size. No need to make even light-hearted excuses. They turned out for Priebus and Ryan.

When the Beltline gets backed up from Gammon to John Nolan on the west side by charter buses, and the east side from John Nolan practically to the interstate exit (when it first started) I'd say they got plenty of support from somewhere to get those figures of (up to) 100k on the capitol lawn for weeks.

Wrong on several counts.

1. I was speaking of the Obama ground organization in WI, and so was Greta (and I believe it was Priebus she was talking to at the time, but it could've been any of their mouthpieces). They were trying to indicate that the GOGG (great Obama ground game) was a paper tiger and had been declawed in the recall vote. That's just not true. Obama did not unleash the mindless progressive zombies on Walker. He's saving him for a midnight snack for his minions in the next gubernatorial election.

2. You're talking about the public service unions heading to WI for the Grand Showdown. That was months and months prior to the recall.

 

[JoeTheJuggler]

Yup. But he does have some very good basis for using state polls -- not only are they "more" unbiased than the national polls, they also are within a margin of accuracy that is quite satisfactory:

RCP had a pretty good article on this topic (national polls vs. state polls)--even made the point that taking an average of the two isn't ideal either.

One point I'm curious about: national polls pretty much only ask about the two big candidates because they use the same question nationwide so they can't reflect who is on the ballot in only some states. Do state polls usually include all the candidates on the ballot in their question?

If so, I would be inclined to think the state polls are more accurate (if the question they're asking more closely reflects the question in front of voters on Election Day).

 

[Kaosium]

So it's the, "The polls are skewed," argument.

Lemme guess, you know some place where they sneer smugly at those philistines not as sophisticated as they are who wonder about the partisan ID of many of these polls? Tell you what, I think I know better than them, but if I told you why, you'd have to shoot me! That wouldn't be fun for me and you might get in trouble--we can't have that!

I disagree, polls at this juncture are only of value (to me) in predicting the outcome of the election. The ones that make unlikely assumptions should be scrutinized more carefully. If some firms are oversampling Dem voters for whatever reason it makes sense to be skeptical of their findings.

You're welcome to believe that the electorate of 2012 will be the same as the historic one of 2008 but with no influence whatsoever from the equally unprecedented 2010 election. I just think that unlikely, and there's data that indicates that as well.

 

[KoihimeNakamura]

Except we don't. Think you're arguing against a strawman.

 

[Kaosium]

Maybe we could get Donald Trump on the Kerry case. I'm sure that will swing as many votes as his Big Reveal did last week. I heard a rumor that Teh Donald got washed out to sea during Frankenstorm Sandy. Unfortunately, it turned out to be a rumor based on someone finding a yellow mohair toilet-seat cover floating in the drink.

(See? Anecdotes and speculation are fun!)

I guess it was Menendez instead, I don't see how that affects the presidential race and he's up 18 according to his last poll.

I was wrong about the crowd size. No need to make even light-hearted excuses. They turned out for Priebus and Ryan.

Did you read that 'Wisconsin Winter Wonderland' link? That one puts me in stitches! There is Truth^TM in that diary of the snow shoveler, especially about the snowplow!

Wrong on several counts.

1. I was speaking of the Obama ground organization in WI, and so was Greta (and I believe it was Priebus she was talking to at the time, but it could've been any of their mouthpieces). They were trying to indicate that the GOGG (great Obama ground game) was a paper tiger and had been declawed in the recall vote. That's just not true. Obama did not unleash the mindless progressive zombies on Walker. He's saving him for a midnight snack for his minions in the next gubernatorial election.

Somebody put together a million signatures in such a remarkably short period of time. I saw some of them, but in truth the ones I saw didn't look out of state like many of the revelers on the capitol lawn did. It did strike me as strange that regardless Walker got more votes the second time around, especially considering the press the WSJ gave him during that period.

2. You're talking about the public service unions heading to WI for the Grand Showdown. That was months and months prior to the recall.

Ah, OK. Part of the same election cycle as far as us poor bastards are concerned, it all mushes together. Imagine (more than!) two years of constant campaign commercials, telephone calls, politics dominating the newspaper much of each week, signs in yards---it's going to be a relief when it is all over. Lately I have Tommy and Tammy handbagging each other on my computer screen every site I go to now!

 

[Kaosium]

Except we don't. Think you're arguing against a strawman.

Is that the royal 'We?'

 

[KoihimeNakamura]

Yes. As Queen of Plushistan, I'm entitled

 

[Foolmewunz]

The recall vote:

This and Walker's election are what has the GOP hopes so high. Does anyone really belive that the northwestern-midwest has gone Tea Party and stayed Tea Party. Remember that MN and WI are a bunch of ornery cusses under the best of conditions. (Yeah, that winter probably has something to do with it.) They are known for putting rather, uh, ... unique... people on the ballot and even electing them. Sweet Joe McCarthy, Fighting Bob LaFolette, Jessie Ventura, Al Franken, Michelle Bachmann,... they do elect some one-off people.

The gains in 2009-10 by the GOP in WI were largely Tea Party disgruntled white folks. (They got a lot of them.) If the TP looked like any kind of a movement with staying power instead of a movement that's so lost its credibility that the GOP Big Dawg, after conning them into supporting him in the primaries, is now running away from them, then they might have a point. But I think that the election that got Walker in was on the coattails of the TP cranky old man revolt.

I do not, however, believe that was the reason that he won the recall vote by a larger margin than his election bid. I believe that WI is still a very progressive state at heart and that there are a lot of people who are not inordinately fond of recalls. I think he got votes that might've gone against him if there was a legitimate election on the books at that time, but were from citizens who said, "Hey, he was elected - give him a chance...". And I believe a bunch of those same staunch progressive-rooted individuals are going to look back on four years of Congressional dam-building and circling of partisan wagons and say the same thing about Obama.

 

[Kaosium]

Yes. As Queen of Plushistan, I'm entitled

What happens to the serfs in Plushistan?

Anyway there's some sort of disconnect going on this year between the polling, I think it's probably driven by how they're determining 'likely voters.' Some of them, like that PPP poll I posted, may just have accessed a voter list and figured since they voted last time, they were likely to vote again, there's no question regarding it in the poll. However that's not exactly how it works sometimes, I would think especially so with new voters in demographics that don't turn out as often.

I suspect that might be driving some of the disconnect between polling methodologies, which produces this conundrum where it appears there's two separate sets of assumptions, leading to those two different peaks. I think some of them are going off the last (presidential) election turnout model or list, and the others are using different methods and are more likely to be correct.

 

[Foolmewunz]

What happens to the serfs in Plushistan?

Anyway there's some sort of disconnect going on this year between the polling, I think it's probably driven by how they're determining 'likely voters.' Some of them, like that PPP poll I posted, may just have accessed a voter list and figured since they voted last time, they were likely to vote again, there's no question regarding it in the poll. However that's not exactly how it works sometimes, I would think especially so with new voters in demographics that don't turn out as often.

I suspect that might be driving some of the disconnect between polling methodologies, which produces this conundrum where it appears there's two separate sets of assumptions, leading to those two different peaks. I think some of them are going off the last (presidential) election turnout model or list, and the others are using different methods and are more likely to be correct.

Look, we all read our preferred websites and as Karl Rove proved again this morning, you can make statistics do anything you want, particularly when the conclusion of the article is "Hey, who knows?" But citing articles by guys who write books on "How the Democrats ********** over America" is probably going to be a stretch for most of us in the neutrality sweepstakes.

The more interesting (and cheering to me) but worrisome (to GOP supporters) factor is that in the past several general elections, local poll averaging has been fairly accurate in states where there were more than six DIFFERENT polls in the week leading up to the election. That's why Nate Silver is looking for more information out of places like Minnesota and Pennsylvania. They were taken off the attention map months ago, and numbers crunchers just don't have enough numbers to do any crunching to. Ohio, Colorado, Iowa, Virginia, Florida, and even NH to an extent, have a whole lot of polls. RCP can pick and chooses which polls to use in Ohio, there are so many.

 

[Kaosium]

The recall vote:

This and Walker's election are what has the GOP hopes so high. Does anyone really belive that the northwestern-midwest has gone Tea Party and stayed Tea Party. Remember that MN and WI are a bunch of ornery cusses under the best of conditions. (Yeah, that winter probably has something to do with it.) They are known for putting rather, uh, ... unique... people on the ballot and even electing them. Sweet Joe McCarthy, Fighting Bob LaFolette, Jessie Ventura, Al Franken, Michelle Bachmann,... they do elect some one-off people.

We also had Proxmire, Russ Feingold, (the only one to vote against the Patriot Act) as well as Gaylord Nelson, (Earth Day) not to mention that Patrick Lucey was John Anderson's running mate in '80, and Lee Dreyfus was the Republican who signed the nations first Gay Rights bill. There's a strong independent tradition, Minnesota gets the weirder ones in my view. I dunno what the 'Tea Party' means to you, but it probably doesn't mean the same 'round these parts.

The gains in 2009-10 by the GOP in WI were largely Tea Party disgruntled white folks. (They got a lot of them.) If the TP looked like any kind of a movement with staying power instead of a movement that's so lost its credibility that the GOP Big Dawg, after conning them into supporting him in the primaries, is now running away from them, then they might have a point. But I think that the election that got Walker in was on the coattails of the TP cranky old man revolt.

Actually I think this had more to do with the public employee unions pissing people off, notably the teachers. I hear the Mayor down in Chicago (I think Obama has heard of him--mebbe you too?) got his own taste of just how arrogant they can be, we've had a lot more of it.

I do not, however, believe that was the reason that he won the recall vote by a larger margin than his election bid. I believe that WI is still a very progressive state at heart and that there are a lot of people who are not inordinately fond of recalls. I think he got votes that might've gone against him if there was a legitimate election on the books at that time, but were from citizens who said, "Hey, he was elected - give him a chance...". And I believe a bunch of those same staunch progressive-rooted individuals are going to look back on four years of Congressional dam-building and circling of partisan wagons and say the same thing about Obama.

They might, but Wisconsin has a different sort of progressive tradition than Chicago.

 

[Kaosium]

Look, we all read our preferred websites and as Karl Rove proved again this morning, you can make statistics do anything you want, particularly when the conclusion of the article is "Hey, who knows?" But citing articles by guys who write books on "How the Democrats ********** over America" is probably going to be a stretch for most of us in the neutrality sweepstakes.

I was merely referring to his charts that illustrated what appeared to be going on to me, that you don't like him, or the books he wrote (that I was entirely unaware of--never heard of him until recently actually) doesn't mean on this point he is wrong. I read from all quarters, that's actually the best way to learn and to figure out what's going on. I think the worst thing about ignoring someone because you don't like whatever else they might think is you might just allow that person to be right where you are wrong! I hate that, thus I'm willing to read from everywhere!

At any rate if one knows how to evaluate information it doesn't matter where it comes from, or even if it's true, as if it's not, that tells you something too.

The more interesting (and cheering to me) but worrisome (to GOP supporters) factor is that in the past several general elections, local poll averaging has been fairly accurate in states where there were more than six DIFFERENT polls in the week leading up to the election. That's why Nate Silver is looking for more information out of places like Minnesota and Pennsylvania. They were taken off the attention map months ago, and numbers crunchers just don't have enough numbers to do any crunching to. Ohio, Colorado, Iowa, Virginia, Florida, and even NH to an extent, have a whole lot of polls. RCP can pick and chooses which polls to use in Ohio, there are so many.

Good point, I think everyone's looking for more info out of MN and PA, not to mention here. It does appear to me that there's two very different approaches being taken in polling this election, and Nate Silver prefers one set of assumptions, and another one seems to make more sense to me. What's interesting is Rasmussen, by Silver's own methods, would have been considered one of the most accurate pollsters, yet there has been a determined effort on Silver's site (and not necessarily just emanating from him) and...elsewhere...to 'discredit' Rasmussen. That might just be booster-ism, but along with the curious interest Obama's campaign manager and Attorney General displayed with Gallup--the other outfit to eschew certain assumptions--it might not be.

 

[remirol]

What's interesting is Rasmussen, by Silver's own methods, would have been considered one of the most accurate pollsters,

This assertion has been destroyed in numerous places, but because lying liars have to lie when they don't have facts, I'll just demolish it one more time. Plus it only takes like 5 minutes to do so. But just to remind you again...

2008 result: Obama +7.3
Rasmussen: Obama +6, a -1.3 point differential

The following polls were MORE accurate than Rasmussen in 2008:

RCP Average (+7.6 (+0.3))
FOX News (+7 (-0.3))
Ipsos/McClatchy (+7 (-0.3))
CNN/Opinion Research (+7 (-0.3))
IBD/Tipp (+8 (+0.7))
NBC News/Wall Street Journal (+8 (+0.7))

The following poll was EQUALLY as accurate as Rasmussen:

Pew Research (+6 (-1.3))

This puts Rasmussen in a tie for 7th most accurate. Not at all a compelling case for "accuracy", especially when you take into account their methodology flaws that persist and the horrendous 2010 performance of Rasmussen.
In addition, of the 16 "final" polls listed, Rasmussen was skewed further right than all but 2 -- both halves of the Battleground poll and Diageo. So it should be no surprise to anyone that Rasmussen effectively marks the "far right" extreme in any given set of polls.

Sourcey McSource Source

The methodology flaws listed above are specifically called out by Nate elsewhere, but the most obvious one is that Rasmussen doesn't call cellphones.

And, of course, there's Rasmussen's obvious outlier status in states like Ohio, Wisconsin, and Colorado -- note links to RCP poll compilations.

Seriously. The pretense that Rasmussen isn't tilted right is laughable at this point. Even Neally won't stick around to try to defend the assertion any more. There are other polls that are tilted just as badly left. It happens. That's why focusing on a single poll and insisting they're "right" is near-guaranteed to lead to disappointment.

 

[Frank Newgent]

When the Beltline gets backed up from Gammon to John Nolan on the west side by charter buses, and the east side from John Nolan practically to the interstate exit (when it first started) I'd say they got plenty of support from somewhere to get those figures of (up to) 100k on the capitol lawn for weeks.

The Beltline backed up from Gammon Road to John Nolen? That's a lot of support coming from Orchard Ridge and West Towne for sure.

Far as the Beltline being backed up from the interstate to John Nolen... how did Obama's motorcade get there? Or might that have had something to do with the backup?

 

[BenBurch]

So it's the, "The polls are skewed," argument.

Nate Silver says they pretty much have to be because State and National polls paint a different picture.