I still think the polls are pretty inaccurate. I still find it difficult to believe that there were really that many undecided people out there and that this election was so volatile.
I should have mentioned that polling firms, should they wish to be hired for the next round of politics, need to demonsrate they called the results correctly. I would submit there will be considerable changes in the polls as november 6 rolls around. no one is going to hire a loser who can predict the outcome.
I am amused that conservatives who hated the polls earlier now think they are great.
Obama's going to lose. He was putting up similar numbers in 2008 against McCain.
The huge break toward Romney at the end is going to be endlessly studied. What I think happened is that the negative campaign run by the media and the Obama camp got too unmoored from reality.
Romney isn't crazy, isn't a cultist, isn't stupid, and isn't incompetent. When all the undecided voters got a look at Romney they saw that the caricature drawn by the Obama campaign wasn't true, and then stopped listening. People tune out lies. Now any negative ads against Romney will not work as well.
Also, class warfare doesn't work to win elections.
Finally, Obama hasn't answered the question of what he'd do in the next four years as President.
MadisonMan is right--the margin of error applies to both numbers (either could be off 2% up or down), therefore a seven point lead is three points outside the margin of error rather than five.
Or course that also means that it is possible the race is 55-43 Mitt as well 50-47. Just saying.
I note that Dark Eden and i am in agreement--if you want to see the way the election is going, look at how the polling firms start reporting their poll results. no political party is going to hire a firm who doesnt get it right.
"The polls are all fixed to favor Obama. Or don't we believe that anymore?"
-- There's only so much large D+ skews can hide; I think part of the reason we're seeing this switch is that the right called out places on D+14 type samples, which forced them to back off advocacy polling. That and the natural trend toward more realistic situations being polled as we get closer to the election so that the polling firms can keep their honor after selling out for months.
Margin of error is ±2, so that's 5 points beyond the margin of error.
I don't believe this is correct. The margin of error is on each candidate's numbers, not the spread. Thus, Romney 50 / Obama 47 is within the margin of error, and so the difference between them is only 3 points beyond the margin of error.
Roger J. said... I should have mentioned that polling firms, should they wish to be hired for the next round of politics, need to demonsrate they called the results correctly.
agree.
On the basic issue...three points
Reelections are an if-then-else loop.
1. does the President deserve reelection? 2. If yes, vote! 3. if no, go to 4 4. Does the Challenger scare me too much? 5. If no, vote! 6. If yes, go back to 1.
repeat till ballots close.
POTUS has no successful record and can't reveal his plan for a second term, so they have been counting on making Romney too scary.
The 1st debate ruined that plan. All OFA has left is a prayer for either a disaster to appear Presidential at, or to catch Romney sleeping with a dead boy.
second, undecided break for the challenger.
third, there is a big Bradley factor happening. I think states like PA and MA are in play for Romney and WI is a win for him. Right now too many union and white households refuse to give pollsters the truth about their Romney voting plans.
I am amused that conservatives who hated the polls earlier now think they are great.
I don't think conservatives hated the polls per se, but rather the fact that the polls, which were so heavily skewed towards Democrats, were being touted as "proof" that Obama was in the lead, when the reality was that Obama was not exactly holding a commanding lead.
Now that the same polls, which I believe are probably still stacked with higher Dem numbers, are showing Romney holding a lead and in some a quite substantial lead, the conservatives are correctly (I pray) interpreting the data as a Romney lead being much larger than even these biased and skewed polls are representing.
Actually, +/- 2 would mean that 95% of the time we could expect Romney's total to be between 54% and 50% while Obamas would be between 47% and 43%. So it could be an 11 point race or 3 point race or anywhere in between. Of course this is also dependent on what the implicit turnout model. If the sample is R+15, then this poll is really bad news for Romney because we aren't likely to see R+15 turnout. If the sample is D+8 (roughly 2008 turnout), then this is even better news than it appears on its face because we know that is not likely to occur. I'm having trouble finding what the underlying turnout assumption is for Gallup. But based on what we have seen from polling so far this year, it is more likely to be closer to the D+8 than R+ anything.
"Margin of error" is somewhat imprecise since the confidence level isn't explicitly defined. In a Gaussian distribution with a standard deviations (SD) of X and a mean(average) of M, M-2x through M+2X would cover 95%. In other words, if you were to gamble that the actual value is in that range, you should win 19 out of 20 times. 95% confidence is the usual standard, so it is likely that this range is the "margin of error," versus 90% or 99%.
So there is a 95% chance that Romney's lead is 50% to 54% and Obama is at 43% to 47%. Since these do not overlap, that indicates a significant difference.
Regarding the trend: The Gallop pole is beginning to look like Mann's Hockey stick!
I checked Intrade so garage doesn't have to. Obama was at about 79 on September 29th, and is around 65 today. There's still a chance to make money with a bet on Romney. I would, but I don't care to pay Intrade's monthly dues.
I don't believe the polls when they are good and I don't believe them when they are bad.
Pollsters are dealing with a 9% response rate. They can't really tell us anything.
I do believe that things are very, very different from 4 years ago. My indicator is even here in Cook County Illinois I have seen ZERO Obama signs and several Romney signs.
At this stage of the game, with 3 weeks left in the campaign, these poll number start to matter. Anything before that was bullshit and I've said that many times before. Unless something totally catastrophic happens, Romney is the next POTUS with taking at least 44 states electorally.
The telling story here is the margin of error. It used to be anywhere from +/-4 to 6 points. Now it's shored up to 2. That is a massive statistical swing.
John Lynch: "I am amused that conservatives who hated the polls earlier now think they are great."
John, your observation is neither an accurate reading of the earlier complaints by conservatives about earlier poll results nor current conservative opinion about recent polls.
What was clear (to any sentient being) earlier in the year is that the voter models being used by the MSM were even more democrat heavy than the actual dem dream turnout in 2008.
Given the reality of our economic situation, there was no way in hell that turnout model was going to be valid.
What made it all the more suspicious then was the clear alignment of the over-sampled dem polls in conjunction with the obama campaign strategy to paint the election as "over" from the summer on.
All along conservatives were saying that as election day neared those very same polling outfits would necessarily modify their turnout models to more accurately reflect voter sentiment in order to maintain their legitimacy.
The obama campaign gambit to end the game before it began failed.
The obama campaign gambit to paint Romney as so unacceptable that he should not even be listened to or considered has failed.
Those failures combined with obama's indisputable economic failures have created this "sort of preference cascade" for Romney at precisely the point where it occurred for Reagan in 1980.
I am amused that conservatives who hated the polls earlier now think they are great.
I don't think conservatives hated the polls per se, but rather the fact that the polls, which were so heavily skewed towards Democrats, were being touted as "proof" that Obama was in the lead, when the reality was that Obama was not exactly holding a commanding lead.
Now that the same polls, which I believe are probably still stacked with higher Dem numbers, are showing Romney holding a lead and in some a quite substantial lead, the conservatives are correctly (I pray) interpreting the data as a Romney lead being much larger than even these biased and skewed polls are representing.
The polls prior to 3 - 4 weeks before the election have been used by the left as a means to disinform and dishearten voters for Romney. They've lied about everything else and used erroneous polling data as means of electoral psychological warfare. It didn't work come the first debate because it showed the left that all of the negative ads that the Urkel campaign put together were a lie from the start. That the characterization that the Urkel campaign made of Romney/Ryan was a house of cards that was never true and Urkel looked like a dear in headlights when presented with those facts. Hence, his speech the next day after the first debate that this wasn't the Romney he knew and had to have his campaign scramble to label Romney a liar on his positions again, which have failed. That's all this was. A $400 million dollar campaign to smear a guy with nothing but lies that evaporated on the first debate. What a waste, but not if you are POTUS with no record to speak of and can't use it as a means to get yourself re-elected.
It would be interesting to compare the October polling trendlines for the last few elections. I bet they would all show a Republican surge as they have to start resembling reality.
"therefore a seven point lead is three points outside the margin of error rather than five"
Except... The two numbers are essentially the same number. They're not independent.
If Obama goes +2, Romney goes -2 (or really close, given "undecided" and third parties, though no third party candidate this time, sadly, has ANY traction whatever).
I'm not sure what that means about how the margin of error is calculated or expressed, but I don't think any of us have it right.
"Pollsters are dealing with a 9% response rate. They can't really tell us anything."
This.
I mean, what is the probability that the 1 in 11 people who respond, represent a random sample?
"Fuck off!" is almost a political philosophy in the US. What are the odds that there is zero correlation between it, and other political leanings? Pretty slim...
Just keep believing the lies you've told yourself and have been told over and over and over and over again. Repeat them as much as you want, but it will never change the reality on the ground. Maybe in the alternate reality you live in this may seem like a bad dream, but hey, we never gave a shit what you thought anyway.
I am amused that conservatives who hated the polls earlier now think they are great.
This is a crude representation of the prior discussion. The question has always been how the polls could be so far at variance w/ the economic fundamentals.
The current opinion numbers are simply starting to line up w/ what the predictions based on economic stats have been saying all along. Nothing hypocritical about Repubs taking comfort from that.
I'm more impressed that Romney just went ahead in the RCP electoral map: http://www.realclearpolitics.com/epolls/2012/president/2012_elections_electoral_college_map.html
The polls are all fixed to favor Obama. Or don't we believe that anymore?
I'm still not a big believer in polling. When polling was tight, to me it looked like the Dem-loving media were highlighting polls that showed their guy Barry in the lead. So I guess I don't feel the polling companies are necessarily fixed, but the media is too eager to run with any poll showing their guy ahead, even if it's heavily weighted towards +D participation.
I don't know enough about polling to opine on whether they should abstain from posting pollings that are overly imbalanced. I guess that's their call, as long as they publicly post all weightings.
The polls are all fixed to favor Obama. Or don't we believe that anymore?
You're making the same mistake a lot of liberal pundits have. If the polls were skewed toward Obama, there's every reason to believe they still are and that makes this number all the most devastating.
Mary: "Pollsters are dealing with a 9% response rate. They can't really tell us anything."
The 9% response rate is another source of continued anxiety by most of us watching the proceedings along with the technology changes (cell phones vs land lines) which could/might/may add to polling inaccuracy.
However, to say that the polls can't tell us anything is wildly overstated.
The polls, when applied consistently using a "reasonable" model can tell us an awful lot about trends, campaign inflection points (and thus campaign strategies/impacts), etc.
Of course, that's why the obama's Justice Dept going after the Gallup organization in an election year, followed by the otherwise inexplicable decision by Gallup to alter their methodology with only a month to go in the race tells us alot.
It was a last ditch effort by the obama-ites to return to their earlier strategy of making obama's reelection seem inevitable and create a last minute cascade toward Teh One.
It would be interesting to compare the October polling trendlines for the last few elections. I bet they would all show a Republican surge as they have to start resembling reality.
The fact tha Romney is Rhino-est of Rhinos is what will be the reason for his landslide victory.
Obama can try to paint him as an arch-conservative when most true conservatives don't think he is one at all but are willing to swallow him to win the election.
So many people who could not swallow a Palin or a Gringrich or even a Bush will be able to either voter for him or stay home.
The Trill is Gone for the Jug Eared Jesus.
It's time to move on to the second husband. You know the boring one. Who has a job.
"The polls, when applied consistently using a "reasonable" model can tell us an awful lot about trends, campaign inflection points (and thus campaign strategies/impacts), etc."
That's a good point.
When I said that they can't tell us much, I was thinking about the polls that purport to tell us absolute results.
It's possible, under the current scenario, that Obama is up 60-40, or Romney is up 60-40, for all we know from polls.
But if the polls show movement, consistently, in a particular direction, that does tell us something.
Me neither. I understand why the campaigns do it, but they have the sophistication and the motivation to use it. As a general news item, it's more like a freak show.
Baron: "Obama can try to paint him as an arch-conservative when most true conservatives don't think he is one at all but are willing to swallow him to win the election."
This comment touches on another key aspect of this election: the Chicago Gang that couldn't shoot straight (unlike the actual gang-bangers setting murder records in Rahm's/obama's Chicago as we speak) STILL can't settle on a single message of "Romney is an arch-conservative!!!" vs "Romney is actually a moderate who keeps flip flopping!!1!11 eleventy!!!!"
What thats really done is help "undecideds" say to themselves, "hey, this guy really isn't all that conservative, so I, a wonderfully important and self-reverential middle of the roader can vote for him with a clear conscience".
Romney should send gift baskets to the obama's chicago HQ's when this is all over.....except the baskets would probably be stolen long before they ever arrived.
"It's the waning of the natural advantage of the incumbent."
The incumbent has a huge natural advantage, in times are really good.
If things are middling, the incumbent has a natural advantage, since there's no reason to make a big change.
If times are bad, the incumbent has a disadvantage, along with the incumbent's whole party -- but the challenger still has to overcome the fact that a President will look like one and a challenger will look like a challenger, usually. That's not worth all that much, unless the challenger is really, really bad (like John Kerry, who nevertheless came close, even in relatively good economic times but with unpopular wars in progress).
Interesting -- it doesn't look like a Republican surge so much as a challenger surge.
Yes, exactly right. Thanks for the clarification.
It is very striking, though, how predictable a shift toward the challenger is in October, and if the there is no incumbent, then the shift goes against the incumbent party.
Tyrone: "It is very striking, though, how predictable a shift toward the challenger is in October, and if the there is no incumbent, then the shift goes against the incumbent party."
I did calling for the, okay, I think they are the Latino branch of Americans for Prosperity, anyhow! So we called with a one-question push poll. "Now that the national debt has reached $16 Trillion dollars, do you think the administration's policies have helped or hurt the economy."
I'd guess I got 50% hang-ups. What was funny was the people who would talk to me but refuse to answer the question because they didn't know how the data collected would be used. Almost paranoia.
The people who answer any poll are self selecting.
I guess the media will soon be telling us that Romney won because he ran a brilliant campaign when in fact Romney won because the voters have had four years to figure out that Obama is an incompetent dope.
Does a 14% negative swing in Obama's numbers in two weeks make you feel good about this?
There has been no Romney movement. The polls are just becoming less skewed.
The 14% is not referring to polls, but to Intrade, which you love to cite. It's my guess Intrade is a lagging indicator anyway, since bets placed early in the cycle are still affecting the numbers today.
I find it very interesting, though, that you have come around to the idea that the polls are skewed.
garage: "There has been no Romney movement. The polls are just becoming less skewed."
When I first read this I had to go back and find the quote. This sounds like a right wing point. Is garage admitting that the polls have been skewed up until now?
You wingnuts won't be strutting once the "Binders" meme kicks in.
That will bury Romney...in binders, piles and piles of binders, which somehow Romney puts women in, and which have filters that siphon off 28% of their pay. The evil bastard.
I don't think I've ever cited Intrade. But, carry on.
Yes, the specific comment I had in mind when I wrote that turns out to have been written by Paulio . My apologies. I guess I have a tendency to attribute all comments of this caliber to you.
The only poll that matters is the one on November 6th. At that point, the American people can choose between four more years of the same with Obama or taking a different course with Romney.
Are you better off now than you were four years ago?
VOTE! (And send Obama into a richly deserved early retirement.)
You wingnuts won't be strutting once the "Binders" meme kicks in.
That will bury Romney...in binders, piles and piles of binders, which somehow Romney puts women in, and which have filters that siphon off 28% of their pay. The evil bastard.
-------------------
But it's HILARIOUS. Binders b-b-b-binders! With women in them!
Somehow, I think the misguided Obama campaign has blundered again with the snark on "binders full of women".
As in Lefty men snarking to women ..."Can you believe, as a Person with Lady Parts, that stupid fool Romney said he went into office with binders full of women he was looking to hire? Binders!!"
Human being with Lady Parts thinks:
"Awwww, got it..he tripped over his tongue..But what I like is he was saying he personally went out looking to hire lots of women to jobs. Their resumes filling his binders. I like that! I could use a better job, and my live at home college grad daughter - any job at all!! Go Romney, my Lady Parts are telling me, finally!"
garage mahal said... The 14% is not referring to polls, but to Intrade, which you love to cite
I don't think I've ever cited Intrade. But, carry on.
Yeah. You did. About a month ago you were crowing over the intrade odds.
You know, I thought so too, but I've spent the last hour looking for evidence (Google search garage intrade site:althouse.blogspot.com) and did not find any. garage is right.
garage, your credibility on the use of facts on here is shot. Not because we often disagree with you, because you're hardly alone in that regard. I haven't paid attention to any "evidence" you've cited or used since at least 2008.
Its a reputation that's perhaps not always fair but its well established. You have no reason to be upset. I think you should turn your attention to the Waukesha County vote counting fraud which you are sure to be on top of in a few weeks. Plenty of trunks to fill with votes.
There has been some movement but much of the change is in the unskewing. Right now it looks like Romney is up 7.x and that if the election were held today -- assuming a conservative 2/1 move to Romney by undecideds -- the vote would be at 54.75% R to 45.25% O. I'd call that blowout territory. I don't see how Obama can possibly come back at this point. But, still no cockiness here. It ain't over until its over and Obama plays dirty.
http://fivethirtyeight.blogs.nytimes.com/2012/10/18/gallup-vs-the-world/ How good is Gallup, when it disagrees with other polls. Nice level of detail from Nate Silver.
According to the maps at MikeR's link New Mexico is 95% likely to go to Obama.
There is ONE Obama sign I pass on my 20 min commute to and from town. There aren't any Romney signs to speak of either. There are hardly any signs at all. Most of the signs there are, not all by any means, but *most* signs are for the Republican candidate for Congress. There are a few for Heather Wilson for Senate. There are fewer for her opponent but the two of them seem to take up nearly all of the purchased television time.
I suppose there is some way that the people who do this stuff figure it out, but people have to go to the polls.
In my little Congressional district I don't know if anyone not voting for Janice is going to bother.
"Freder Frederson said... The polls are all fixed to favor Obama. Or don't we believe that anymore?"
They are shifting to "likely voters" which allows them to pretend they never were trying to push Obama ahead. If Romney blew up their polls, they would have top admit stuff, like the exit polls in 2004. Poor Kerry believed them but he might have been the only one.
I think this is a crucial fact. It's not that he has no plan. It's just that the land slide would be bigger if he revealed what it is. We got a hint in that aside to Medvedev that he thought was off mike. That one incident may cost him the election more than any other single thing.
Forgive me if I am posting things that others have posted. I haven't read all 127 comments.
On October 27th, 2000, GW Bush had a 13 point lead in Gallup and we all know who won the popular vote in 2000...
Gallup is sometimes way out of line with actual sentiments among voters, especially when they contradict or are in conflict with the majority of the polls.
Nate Silver has an excellent write up on "Gallup vs. the World" http://fivethirtyeight.blogs.nytimes.com/2012/10/18/gallup-vs-the-world/
But not having visited Queen Ann's blog in so long, I'd forgotten how much I dislike it and how much it's just basically part of the right wing hate machine.
Willy - I pay attention to the RCP Electoral College poll which shows Romney ahead 206-201 right now. That is the single most important poll in the business.
I am a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for me to earn fees by linking to Amazon.com and affiliated sites.
Encourage Althouse by making a donation:
Make a 1-time donation or set up a monthly donation of any amount you choose:
127 comments:
...!...
The polls are all fixed to favor Obama. Or don't we believe that anymore?
Just so. All the signs have pointed to this inevitable shift, and I have been confident this would happen.
I wonder how garage must feel about his Intrade odds now?
Freder Frederson said...
The polls are all fixed to favor Obama. Or don't we believe that anymore?
Excellent point. Yes, we still do believe that, and this means that Romney's margin of victory will be far larger than 7%.
Great news and as it should be. I hope/think Romney will bury Obama in the Foreign Policy debate; Obama doesn't have a leg to stand on.
I find the interesting number is that of those not affiliated with any party Romney leads by 9 points.
I still think the polls are pretty inaccurate. I still find it difficult to believe that there were really that many undecided people out there and that this election was so volatile.
#ObamaOneTermPresident
#ForOnceObamaGetstheGradeHeEarned
#YouCanFoolSomeofthePeopleAlloftheTime
#GarageMahal:ObamaTrappedRomney!
#AffirmativeActionHireFail
#NoMoreFore!Years
No, we still believe it Freder. Which means it's going to be a wipe-out.
Seven Machos, for many months, has been uber confident in predicting a sizable Obama loss. It looks like he was spot on.
The presidential flatus is escaping its empty suit.
It does appear that polling data is showing a shift toward Romney. Nontheless, the only poll that counts is on November 6.
No, we still believe it Freder. Which means it's going to be a wipe-out.
Well this close to the election the pollsters have to start telling the truth, or getting closer to it anyway.
Well within the margin of fraud. Team O could probably reverse a 60%–40% loss.
so that's 5 points beyond the margin of error.
Not quite the right statistical interpretation.
If the sampling were done repeatedly, Romney's numbers would range from 50-54%, and Obama's from 43-47%.
(That's how I understand it, at least)
I should have mentioned that polling firms, should they wish to be hired for the next round of politics, need to demonsrate they called the results correctly. I would submit there will be considerable changes in the polls as november 6 rolls around. no one is going to hire a loser who can predict the outcome.
"I hope/think Romney will bury Obama in the Foreign Policy debate; Obama doesn't have a leg to stand on."
True enough, but he might have a Star Wars villain to hide behind, like in debate #2.
Landslide
Obama has jumped the binder!
Well, this is getting interesting, isn't it?
I am amused that conservatives who hated the polls earlier now think they are great.
Obama's going to lose. He was putting up similar numbers in 2008 against McCain.
The huge break toward Romney at the end is going to be endlessly studied. What I think happened is that the negative campaign run by the media and the Obama camp got too unmoored from reality.
Romney isn't crazy, isn't a cultist, isn't stupid, and isn't incompetent. When all the undecided voters got a look at Romney they saw that the caricature drawn by the Obama campaign wasn't true, and then stopped listening. People tune out lies. Now any negative ads against Romney will not work as well.
Also, class warfare doesn't work to win elections.
Finally, Obama hasn't answered the question of what he'd do in the next four years as President.
MadisonMan is right--the margin of error applies to both numbers (either could be off 2% up or down), therefore a seven point lead is three points outside the margin of error rather than five.
Or course that also means that it is possible the race is 55-43 Mitt as well 50-47. Just saying.
I note that Dark Eden and i am in agreement--if you want to see the way the election is going, look at how the polling firms start reporting their poll results. no political party is going to hire a firm who doesnt get it right.
Awesome! Now the remaining undecideds will vote for Romney so they can hang with the cool kids.
"The polls are all fixed to favor Obama. Or don't we believe that anymore?"
-- There's only so much large D+ skews can hide; I think part of the reason we're seeing this switch is that the right called out places on D+14 type samples, which forced them to back off advocacy polling. That and the natural trend toward more realistic situations being polled as we get closer to the election so that the polling firms can keep their honor after selling out for months.
Margin of error is ±2, so that's 5 points beyond the margin of error.
I don't believe this is correct. The margin of error is on each candidate's numbers, not the spread. Thus, Romney 50 / Obama 47 is within the margin of error, and so the difference between them is only 3 points beyond the margin of error.
It's about time.
Roger J. said...
I should have mentioned that polling firms, should they wish to be hired for the next round of politics, need to demonsrate they called the results correctly.
agree.
On the basic issue...three points
Reelections are an if-then-else loop.
1. does the President deserve reelection?
2. If yes, vote!
3. if no, go to 4
4. Does the Challenger scare me too much?
5. If no, vote!
6. If yes, go back to 1.
repeat till ballots close.
POTUS has no successful record and can't reveal his plan for a second term, so they have been counting on making Romney too scary.
The 1st debate ruined that plan. All OFA has left is a prayer for either a disaster to appear Presidential at, or to catch Romney sleeping with a dead boy.
second, undecided break for the challenger.
third, there is a big Bradley factor happening. I think states like PA and MA are in play for Romney and WI is a win for him. Right now too many union and white households refuse to give pollsters the truth about their Romney voting plans.
I am amused that conservatives who hated the polls earlier now think they are great.
I don't think conservatives hated the polls per se, but rather the fact that the polls, which were so heavily skewed towards Democrats, were being touted as "proof" that Obama was in the lead, when the reality was that Obama was not exactly holding a commanding lead.
Now that the same polls, which I believe are probably still stacked with higher Dem numbers, are showing Romney holding a lead and in some a quite substantial lead, the conservatives are correctly (I pray) interpreting the data as a Romney lead being much larger than even these biased and skewed polls are representing.
Actually, +/- 2 would mean that 95% of the time we could expect Romney's total to be between 54% and 50% while Obamas would be between 47% and 43%.
So it could be an 11 point race or 3 point race or anywhere in between.
Of course this is also dependent on what the implicit turnout model. If the sample is R+15, then this poll is really bad news for Romney because we aren't likely to see R+15 turnout. If the sample is D+8 (roughly 2008 turnout), then this is even better news than it appears on its face because we know that is not likely to occur.
I'm having trouble finding what the underlying turnout assumption is for Gallup. But based on what we have seen from polling so far this year, it is more likely to be closer to the D+8 than R+ anything.
This would reflect one day post debate?
"Margin of error" is somewhat imprecise since the confidence level isn't explicitly defined. In a Gaussian distribution with a standard deviations (SD) of X and a mean(average) of M, M-2x through M+2X would cover 95%. In other words, if you were to gamble that the actual value is in that range, you should win 19 out of 20 times. 95% confidence is the usual standard, so it is likely that this range is the "margin of error," versus 90% or 99%.
So there is a 95% chance that Romney's lead is 50% to 54% and Obama is at 43% to 47%. Since these do not overlap, that indicates a significant difference.
Regarding the trend:
The Gallop pole is beginning to look like Mann's Hockey stick!
The polls will get real a week from Monday. Until then, take them for what they are, entertainment.
It is interesting to see Obama plateau at 45% and Romney steadily rising.
I checked Intrade so garage doesn't have to. Obama was at about 79 on September 29th, and is around 65 today. There's still a chance to make money with a bet on Romney. I would, but I don't care to pay Intrade's monthly dues.
I don't believe the polls when they are good and I don't believe them when they are bad.
Pollsters are dealing with a 9% response rate. They can't really tell us anything.
I do believe that things are very, very different from 4 years ago. My indicator is even here in Cook County Illinois I have seen ZERO Obama signs and several Romney signs.
At this stage of the game, with 3 weeks left in the campaign, these poll number start to matter. Anything before that was bullshit and I've said that many times before. Unless something totally catastrophic happens, Romney is the next POTUS with taking at least 44 states electorally.
The telling story here is the margin of error. It used to be anywhere from +/-4 to 6 points. Now it's shored up to 2. That is a massive statistical swing.
According to Ras, without "leaners", it's R 49 - O 46.
Lookin' rough out there.
And that's with that great "win" for Barry Tuesday night.
I checked Intrade so garage doesn't have to. Obama was at about 79 on September 29th, and is around 65 today
So that's making you feel good about this?
John Lynch: "I am amused that conservatives who hated the polls earlier now think they are great."
John, your observation is neither an accurate reading of the earlier complaints by conservatives about earlier poll results nor current conservative opinion about recent polls.
What was clear (to any sentient being) earlier in the year is that the voter models being used by the MSM were even more democrat heavy than the actual dem dream turnout in 2008.
Given the reality of our economic situation, there was no way in hell that turnout model was going to be valid.
What made it all the more suspicious then was the clear alignment of the over-sampled dem polls in conjunction with the obama campaign strategy to paint the election as "over" from the summer on.
All along conservatives were saying that as election day neared those very same polling outfits would necessarily modify their turnout models to more accurately reflect voter sentiment in order to maintain their legitimacy.
The obama campaign gambit to end the game before it began failed.
The obama campaign gambit to paint Romney as so unacceptable that he should not even be listened to or considered has failed.
Those failures combined with obama's indisputable economic failures have created this "sort of preference cascade" for Romney at precisely the point where it occurred for Reagan in 1980.
Dust Bunny Queen said...
I am amused that conservatives who hated the polls earlier now think they are great.
I don't think conservatives hated the polls per se, but rather the fact that the polls, which were so heavily skewed towards Democrats, were being touted as "proof" that Obama was in the lead, when the reality was that Obama was not exactly holding a commanding lead.
Now that the same polls, which I believe are probably still stacked with higher Dem numbers, are showing Romney holding a lead and in some a quite substantial lead, the conservatives are correctly (I pray) interpreting the data as a Romney lead being much larger than even these biased and skewed polls are representing.
The polls prior to 3 - 4 weeks before the election have been used by the left as a means to disinform and dishearten voters for Romney. They've lied about everything else and used erroneous polling data as means of electoral psychological warfare. It didn't work come the first debate because it showed the left that all of the negative ads that the Urkel campaign put together were a lie from the start. That the characterization that the Urkel campaign made of Romney/Ryan was a house of cards that was never true and Urkel looked like a dear in headlights when presented with those facts. Hence, his speech the next day after the first debate that this wasn't the Romney he knew and had to have his campaign scramble to label Romney a liar on his positions again, which have failed. That's all this was. A $400 million dollar campaign to smear a guy with nothing but lies that evaporated on the first debate. What a waste, but not if you are POTUS with no record to speak of and can't use it as a means to get yourself re-elected.
It would be interesting to compare the October polling trendlines for the last few elections. I bet they would all show a Republican surge as they have to start resembling reality.
"therefore a seven point lead is three points outside the margin of error rather than five"
Except... The two numbers are essentially the same number. They're not independent.
If Obama goes +2, Romney goes -2 (or really close, given "undecided" and third parties, though no third party candidate this time, sadly, has ANY traction whatever).
I'm not sure what that means about how the margin of error is calculated or expressed, but I don't think any of us have it right.
"Pollsters are dealing with a 9% response rate. They can't really tell us anything."
This.
I mean, what is the probability that the 1 in 11 people who respond, represent a random sample?
"Fuck off!" is almost a political philosophy in the US. What are the odds that there is zero correlation between it, and other political leanings? Pretty slim...
garage mahal said...
So that's making you feel good about this?
Just keep believing the lies you've told yourself and have been told over and over and over and over again. Repeat them as much as you want, but it will never change the reality on the ground. Maybe in the alternate reality you live in this may seem like a bad dream, but hey, we never gave a shit what you thought anyway.
garage mahal said...
So that's making you feel good about this?
Let me play turnabout. Does a 14% negative swing in Obama's numbers in two weeks make you feel good about this?
think states like PA and MA
big whoopsee. I meant Michigan.
I am amused that conservatives who hated the polls earlier now think they are great.
This is a crude representation of the prior discussion. The question has always been how the polls could be so far at variance w/ the economic fundamentals.
The current opinion numbers are simply starting to line up w/ what the predictions based on economic stats have been saying all along. Nothing hypocritical about Repubs taking comfort from that.
I'm more impressed that Romney just went ahead in the RCP electoral map: http://www.realclearpolitics.com/epolls/2012/president/2012_elections_electoral_college_map.html
But I will not get cocky, Pops.
The polls are all fixed to favor Obama. Or don't we believe that anymore?
I'm still not a big believer in polling. When polling was tight, to me it looked like the Dem-loving media were highlighting polls that showed their guy Barry in the lead. So I guess I don't feel the polling companies are necessarily fixed, but the media is too eager to run with any poll showing their guy ahead, even if it's heavily weighted towards +D participation.
I don't know enough about polling to opine on whether they should abstain from posting pollings that are overly imbalanced. I guess that's their call, as long as they publicly post all weightings.
Outliers outliars everywhere!
The polls are all fixed to favor Obama. Or don't we believe that anymore?
You're making the same mistake a lot of liberal pundits have. If the polls were skewed toward Obama, there's every reason to believe they still are and that makes this number all the most devastating.
Mary: "Pollsters are dealing with a 9% response rate. They can't really tell us anything."
The 9% response rate is another source of continued anxiety by most of us watching the proceedings along with the technology changes (cell phones vs land lines) which could/might/may add to polling inaccuracy.
However, to say that the polls can't tell us anything is wildly overstated.
The polls, when applied consistently using a "reasonable" model can tell us an awful lot about trends, campaign inflection points (and thus campaign strategies/impacts), etc.
Of course, that's why the obama's Justice Dept going after the Gallup organization in an election year, followed by the otherwise inexplicable decision by Gallup to alter their methodology with only a month to go in the race tells us alot.
It was a last ditch effort by the obama-ites to return to their earlier strategy of making obama's reelection seem inevitable and create a last minute cascade toward Teh One.
Didn't work.
Didn't come close to working.
Why?
Obama has a record.
Everything is just noise.
Okay, this is brilliant funny:
bad lip syncing the first presidential debate
Come on, Althouse, get this video on Instapundit!
"This is a crude representation of the prior discussion."
And your point?
Oh dear, I missed that obvious typo:
The Gallop poLL is beginning to look like Mann's Hockey stick!
Real Clear Politics has Romney up in the electoral count. I believe that comes from moving NC to "leans R."
At the beginning of October, the President was up 88. Now, he's down. Stunning.
garage mahal said...
So that's making you feel good about this?
Tyrone said...
Let me play turnabout. Does a 14% negative swing in Obama's numbers in two weeks make you feel good about this?
If Garage were a baseball team, he'd only be designated hitters. He doesn't like to play defense.
Dark Eden said...
It would be interesting to compare the October polling trendlines for the last few elections. I bet they would all show a Republican surge as they have to start resembling reality.
Your wish is my command
And you are, of course, correct.
Everything ELSE is just noise.
Sorry.
Tim in Vermont: "And your point?"
That Ohio maple syrup is every bit as good as Vermont maple syrup?
Yes, I said it.....that just happened....
Mike got there first. Originally.
The fact tha Romney is Rhino-est of Rhinos is what will be the reason for his landslide victory.
Obama can try to paint him as an arch-conservative when most true conservatives don't think he is one at all but are willing to swallow him to win the election.
So many people who could not swallow a Palin or a Gringrich or even a Bush will be able to either voter for him or stay home.
The Trill is Gone for the Jug Eared Jesus.
It's time to move on to the second husband. You know the boring one. Who has a job.
"Your wish is my command "
Interesting -- it doesn't look like a Republican surge so much as a challenger surge.
It's the waning of the natural advantage of the incumbent.
"The polls, when applied consistently using a "reasonable" model can tell us an awful lot about trends, campaign inflection points (and thus campaign strategies/impacts), etc."
That's a good point.
When I said that they can't tell us much, I was thinking about the polls that purport to tell us absolute results.
It's possible, under the current scenario, that Obama is up 60-40, or Romney is up 60-40, for all we know from polls.
But if the polls show movement, consistently, in a particular direction, that does tell us something.
"Yes, I said it.....that just happened."
Hah!
"I'm still not a big believer in polling."
Me neither. I understand why the campaigns do it, but they have the sophistication and the motivation to use it. As a general news item, it's more like a freak show.
Salry data from U.S. Senator Obama's office. Of the ten highest paid employees, seven were men.
http://www.legistorm.com/member/76/Sen_Barack_Obama/59.html
The next debate will be a blow out.
Barry will go full retard and there will be no way to come back from it.
Does a 14% negative swing in Obama's numbers in two weeks make you feel good about this?
There has been no Romney movement. The polls are just becoming less skewed.
Baron: "Obama can try to paint him as an arch-conservative when most true conservatives don't think he is one at all but are willing to swallow him to win the election."
This comment touches on another key aspect of this election: the Chicago Gang that couldn't shoot straight (unlike the actual gang-bangers setting murder records in Rahm's/obama's Chicago as we speak) STILL can't settle on a single message of "Romney is an arch-conservative!!!" vs "Romney is actually a moderate who keeps flip flopping!!1!11 eleventy!!!!"
What thats really done is help "undecideds" say to themselves, "hey, this guy really isn't all that conservative, so I, a wonderfully important and self-reverential middle of the roader can vote for him with a clear conscience".
Romney should send gift baskets to the obama's chicago HQ's when this is all over.....except the baskets would probably be stolen long before they ever arrived.
Better yet, Chik-fil-a gift baskets.
Tyrone, thank you for posting that! Very interesting!
Less skewed?
Oh you mean they stopped using loaded dice.
Nice of them.
"It's the waning of the natural advantage of the incumbent."
The incumbent has a huge natural advantage, in times are really good.
If things are middling, the incumbent has a natural advantage, since there's no reason to make a big change.
If times are bad, the incumbent has a disadvantage, along with the incumbent's whole party -- but the challenger still has to overcome the fact that a President will look like one and a challenger will look like a challenger, usually. That's not worth all that much, unless the challenger is really, really bad (like John Kerry, who nevertheless came close, even in relatively good economic times but with unpopular wars in progress).
garage: "There has been no Romney movement. The polls are just becoming less skewed."
LOL
Hey, the train is not moving, it's that darn platform that is.....
clint said...
Interesting -- it doesn't look like a Republican surge so much as a challenger surge.
Yes, exactly right. Thanks for the clarification.
It is very striking, though, how predictable a shift toward the challenger is in October, and if the there is no incumbent, then the shift goes against the incumbent party.
There has been no Romney movement. The polls are just becoming less skewed.
Obama is not less popular. His appeal is becoming more selective.
Tyrone: "It is very striking, though, how predictable a shift toward the challenger is in October, and if the there is no incumbent, then the shift goes against the incumbent party."
This.
I did calling for the, okay, I think they are the Latino branch of Americans for Prosperity, anyhow! So we called with a one-question push poll. "Now that the national debt has reached $16 Trillion dollars, do you think the administration's policies have helped or hurt the economy."
I'd guess I got 50% hang-ups. What was funny was the people who would talk to me but refuse to answer the question because they didn't know how the data collected would be used. Almost paranoia.
The people who answer any poll are self selecting.
I guess the media will soon be telling us that Romney won because he ran a brilliant campaign when in fact Romney won because the voters have had four years to figure out that Obama is an incompetent dope.
" "There has been no Romney movement."
There are no tanks in Baghdad.
"The polls are just becoming less skewed."
This is a fascinating statement. Has garage always admitted that the polls are skewed?
garage mahal said...
Does a 14% negative swing in Obama's numbers in two weeks make you feel good about this?
There has been no Romney movement. The polls are just becoming less skewed.
The 14% is not referring to polls, but to Intrade, which you love to cite. It's my guess Intrade is a lagging indicator anyway, since bets placed early in the cycle are still affecting the numbers today.
I find it very interesting, though, that you have come around to the idea that the polls are skewed.
garage: "There has been no Romney movement. The polls are just becoming less skewed."
When I first read this I had to go back and find the quote. This sounds like a right wing point. Is garage admitting that the polls have been skewed up until now?
If Garage were a baseball team, he'd only be designated hitters. He doesn't like to play defense.
If garage were a basketball team, he'd be the Washington Generals.
There has been no Romney movement. The polls are just becoming less skewed.
And that's better for you ... how?
The 14% is not referring to polls, but to Intrade, which you love to cite
I don't think I've ever cited Intrade. But, carry on.
If garage were a basketball team, he'd be the Washington Generals.
If he were a college football team, he'd be Eastern Michigan.
We're all skewballs now.
Americans are racists.
If he were military, he'd be the 13th Light Brigade.
The Clinton clan revenge for Hillary's dive under the Benghazi Bus?
You wingnuts won't be strutting once the "Binders" meme kicks in.
That will bury Romney...in binders, piles and piles of binders, which somehow Romney puts women in, and which have filters that siphon off 28% of their pay. The evil bastard.
garage mahal said...
I don't think I've ever cited Intrade. But, carry on.
Yes, the specific comment I had in mind when I wrote that turns out to have been written by Paulio . My apologies. I guess I have a tendency to attribute all comments of this caliber to you.
Judging by the popularity of 50 Shades of Grey, thoughts of Romney and Ryan putting them in binders might get more women to vote for them.
I guess I have a tendency to attribute all comments of this caliber to you.
gee, thanks.
I do like GM's avatar. That's high quality.
I imagine he has it on the side of a conversion van. (jk)
I guess I have a tendency to attribute all comments of this caliber to you.
Paulio's comments would need more lols and wheeeees!
garage mahal said...
The 14% is not referring to polls, but to Intrade, which you love to cite
I don't think I've ever cited Intrade. But, carry on.
Yeah. You did. About a month ago you were crowing over the intrade odds.
The only poll that matters is the one on November 6th. At that point, the American people can choose between four more years of the same with Obama or taking a different course with Romney.
Are you better off now than you were four years ago?
VOTE! (And send Obama into a richly deserved early retirement.)
Rusty said...
garage mahal said...
The 14% is not referring to polls, but to Intrade, which you love to cite
I don't think I've ever cited Intrade. But, carry on.
Yeah. You did. About a month ago you were crowing over the intrade odds.
Garage caught lying again with his short memory? Shocking.
You wingnuts won't be strutting once the "Binders" meme kicks in.
That will bury Romney...in binders, piles and piles of binders, which somehow Romney puts women in, and which have filters that siphon off 28% of their pay. The evil bastard.
-------------------
But it's HILARIOUS. Binders b-b-b-binders! With women in them!
AHAHAHAHAHAHAHAHA... oh god...*whimpers*
Newspaper that employed Algore years ago (and is still a liberal paper) is endorsing Romney.
Garage caught lying again with his short memory? Shocking.?
Prove it, pigfucker.
pig fucker
Good one.
Time to double-down on Big Bird...and Binders.
The pollsters do all kinds of market research for companies, so their credibility is on the line.
They have to start addressing reality or they get no business.
Obama Campaign is going after Gallup. And has been.
Somehow, I think the misguided Obama campaign has blundered again with the snark on "binders full of women".
As in Lefty men snarking to women ..."Can you believe, as a Person with Lady Parts, that stupid fool Romney said he went into office with binders full of women he was looking to hire? Binders!!"
Human being with Lady Parts thinks:
"Awwww, got it..he tripped over his tongue..But what I like is he was saying he personally went out looking to hire lots of women to jobs. Their resumes filling his binders. I like that! I could use a better job, and my live at home college grad daughter - any job at all!! Go Romney, my Lady Parts are telling me, finally!"
Bag0hYou wingnuts won't be strutting once the "Binders" meme kicks in.
...and Big Bird. And 'bortion. And Benghazi. and Bain.
The Killer B's.
Rusty said...
garage mahal said...
The 14% is not referring to polls, but to Intrade, which you love to cite
I don't think I've ever cited Intrade. But, carry on.
Yeah. You did. About a month ago you were crowing over the intrade odds.
You know, I thought so too, but I've spent the last hour looking for evidence (Google search garage intrade site:althouse.blogspot.com) and did not find any. garage is right.
Yowsah, Yowsah, Yowsah!
Don't get cocky kid.
"garage is right."
Didn't you mean, "garage is right, pigfucker"?
garage, your credibility on the use of facts on here is shot. Not because we often disagree with you, because you're hardly alone in that regard. I haven't paid attention to any "evidence" you've cited or used since at least 2008.
Its a reputation that's perhaps not always fair but its well established. You have no reason to be upset. I think you should turn your attention to the Waukesha County vote counting fraud which you are sure to be on top of in a few weeks. Plenty of trunks to fill with votes.
garage may not have cited Intrade, but he did crow about Obama's alleged win probability.
There has been some movement but much of the change is in the unskewing. Right now it looks like Romney is up 7.x and that if the election were held today -- assuming a conservative 2/1 move to Romney by undecideds -- the vote would be at 54.75% R to 45.25% O. I'd call that blowout territory. I don't see how Obama can possibly come back at this point. But, still no cockiness here. It ain't over until its over and Obama plays dirty.
Didn't you mean, "garage is right, pigfucker"?
Pigfucker is kind of endearing.
For all us skewballs out there
That WaPo/abc with a +9 D has Romney up 4.
Whoa.
http://fivethirtyeight.blogs.nytimes.com/2012/10/18/gallup-vs-the-world/
How good is Gallup, when it disagrees with other polls. Nice level of detail from Nate Silver.
It's 1800 in MN and neither the WaPo, nor the NYT has managed to find this story! Maybe tomorrow.
According to the maps at MikeR's link New Mexico is 95% likely to go to Obama.
There is ONE Obama sign I pass on my 20 min commute to and from town. There aren't any Romney signs to speak of either. There are hardly any signs at all. Most of the signs there are, not all by any means, but *most* signs are for the Republican candidate for Congress. There are a few for Heather Wilson for Senate. There are fewer for her opponent but the two of them seem to take up nearly all of the purchased television time.
I suppose there is some way that the people who do this stuff figure it out, but people have to go to the polls.
In my little Congressional district I don't know if anyone not voting for Janice is going to bother.
"Freder Frederson said...
The polls are all fixed to favor Obama. Or don't we believe that anymore?"
They are shifting to "likely voters" which allows them to pretend they never were trying to push Obama ahead. If Romney blew up their polls, they would have top admit stuff, like the exit polls in 2004. Poor Kerry believed them but he might have been the only one.
"can't reveal his plan for a second term"
I think this is a crucial fact. It's not that he has no plan. It's just that the land slide would be bigger if he revealed what it is. We got a hint in that aside to Medvedev that he thought was off mike. That one incident may cost him the election more than any other single thing.
New tag:
Obama is like the Yankees
For comparisons, You gotta love nbc.com's current headline: Obama Holds Leads in Iowa and Wisconsin
Forgive me if I am posting things that others have posted. I haven't read all 127 comments.
On October 27th, 2000, GW Bush had a 13 point lead in Gallup and we all know who won the popular vote in 2000...
Gallup is sometimes way out of line with actual sentiments among voters, especially when they contradict or are in conflict with the majority of the polls.
Nate Silver has an excellent write up on "Gallup vs. the World" http://fivethirtyeight.blogs.nytimes.com/2012/10/18/gallup-vs-the-world/
But not having visited Queen Ann's blog in so long, I'd forgotten how much I dislike it and how much it's just basically part of the right wing hate machine.
Willy - I pay attention to the RCP Electoral College poll which shows Romney ahead 206-201 right now. That is the single most important poll in the business.
Post a Comment