Poll of Polls are affected by their components - no?
The reason I asked the question in the first place was because of seeing commentators on FOX saying the polls themselves were being skewed by their (non?)models.
If I poll every single person I go to church with in this small East Texas town - Romney wins 95%+, but at family dinner with parents and sisters Romney - 14%, Obama - 86%
[quote]jnd wrote:
[quote]ZEB wrote:
[quote]jnd wrote:
[quote]treco wrote:
Think any of the polls are both scientifically derived (as much as a sample can truly represent 100MM voters) and non biased?
[/quote]
I wouldn’t trust any one poll. You need to look at the trend of all of the polls. Based on all of the polls, this election is over. Obama gets 4 more years.
jnd[/quote]
That’s absolutely false!
In 1980 Jimmy Carter lead Ronald Reagan almost right up to their first debate. When America saw the two next to each other they decided Reagan was a viable alternative. Ronald Reagan went on to carry 44 states and beat Carter by about 8 million votes. And just a few months prior to that he was down by 10 points! The voting public is very fickle.
Right now the very best polls show that Obama has about a 2-4 point advantage nationally. But what is more important Romney has made some great gains in various key states like Florida. Keep in mind that this is an electoral race not a popular vote race. Therefore. key states will determine the victor.
While I agree (and I’ve said from the start) that it will be difficult for Romney to defeat Obama there is no poll currently available that demonstrates that it is over. Not by a long shot![/quote]
There is nothing false about what I said. If you disagree with my view of the polls, that’s fine, but that does not mean what I said was false. Again- if you look at all of the polls as a whole, then this is over.
I’m willing to bet on it. Are you?
jnd
[/quote]
It’s my opinion that your opinion is false.
Are you really that offended?
If you actually follow it as you pretend to do then you know it can go either way. Right now you are convinced that Obama will win because he had a popular vote bounce after his convention. That is what the typical guy on the street thinks every time his (or the other) team pulls ahead. Congratulations on that spectacular insight.
But as Karl Rove and Joe Trippi and others who actually make a living following this stuff will tell you, the democrats usually get a 4.9 pt bounce after their convention. This held pretty true this time around. But as they will also tell you their bounce is less likely to stick than a modest republican bounce. Go figure…guys who follow it for a living know more than say YOU!
So, sure Obama could win. As I’ve been saying for months on end he is the favorite to win for many reasons. But claiming that he’s going to win because YOU spent exactly 1:00 glancing at the latest poll well…don’t get all upset if I’m not jumping up and down at your incredible prognostication ability.
So yeah based on your (lack of) analysis I think it is false.
[quote]ZEB wrote:
[quote]jnd wrote:
[quote]ZEB wrote:
[quote]jnd wrote:
[quote]treco wrote:
Think any of the polls are both scientifically derived (as much as a sample can truly represent 100MM voters) and non biased?
[/quote]
I wouldn’t trust any one poll. You need to look at the trend of all of the polls. Based on all of the polls, this election is over. Obama gets 4 more years.
jnd[/quote]
That’s absolutely false!
In 1980 Jimmy Carter lead Ronald Reagan almost right up to their first debate. When America saw the two next to each other they decided Reagan was a viable alternative. Ronald Reagan went on to carry 44 states and beat Carter by about 8 million votes. And just a few months prior to that he was down by 10 points! The voting public is very fickle.
Right now the very best polls show that Obama has about a 2-4 point advantage nationally. But what is more important Romney has made some great gains in various key states like Florida. Keep in mind that this is an electoral race not a popular vote race. Therefore. key states will determine the victor.
While I agree (and I’ve said from the start) that it will be difficult for Romney to defeat Obama there is no poll currently available that demonstrates that it is over. Not by a long shot![/quote]
There is nothing false about what I said. If you disagree with my view of the polls, that’s fine, but that does not mean what I said was false. Again- if you look at all of the polls as a whole, then this is over.
I’m willing to bet on it. Are you?
jnd
[/quote]
It’s my opinion that your opinion is false.
Are you really that offended?
If you actually follow it as you pretend to do then you know it can go either way. Right now you are convinced that Obama will win because he had a popular vote bounce after his convention. That is what the typical guy on the street thinks every time his (or the other) team pulls ahead. Congratulations on that spectacular insight.
But as Karl Rove and Joe Trippi and others who actually make a living following this stuff will tell you, the democrats usually get a 4.9 pt bounce after their convention. This held pretty true this time around. But as they will also tell you their bounce is less likely to stick than a modest republican bounce. Go figure…guys who follow it for a living know more than say YOU!
So, sure Obama could win. As I’ve been saying for months on end he is the favorite to win for many reasons. But claiming that he’s going to win because YOU spent exactly 1:00 glancing at the latest poll well…don’t get all upset if I’m not jumping up and down at your incredible prognostication ability.
So yeah based on your (lack of) analysis I think it is false.
[/quote]
So then you do not want to bet on this election???
If my analysis is wrong, then you should have no trouble betting… right?
jnd
[quote]ZEB wrote:
[quote]jnd wrote:
[quote]ZEB wrote:
[quote]jnd wrote:
[quote]treco wrote:
Think any of the polls are both scientifically derived (as much as a sample can truly represent 100MM voters) and non biased?
[/quote]
I wouldn’t trust any one poll. You need to look at the trend of all of the polls. Based on all of the polls, this election is over. Obama gets 4 more years.
jnd[/quote]
That’s absolutely false!
In 1980 Jimmy Carter lead Ronald Reagan almost right up to their first debate. When America saw the two next to each other they decided Reagan was a viable alternative. Ronald Reagan went on to carry 44 states and beat Carter by about 8 million votes. And just a few months prior to that he was down by 10 points! The voting public is very fickle.
Right now the very best polls show that Obama has about a 2-4 point advantage nationally. But what is more important Romney has made some great gains in various key states like Florida. Keep in mind that this is an electoral race not a popular vote race. Therefore. key states will determine the victor.
While I agree (and I’ve said from the start) that it will be difficult for Romney to defeat Obama there is no poll currently available that demonstrates that it is over. Not by a long shot![/quote]
There is nothing false about what I said. If you disagree with my view of the polls, that’s fine, but that does not mean what I said was false. Again- if you look at all of the polls as a whole, then this is over.
I’m willing to bet on it. Are you?
jnd
[/quote]
It’s my opinion that your opinion is false.
Are you really that offended?
If you actually follow it as you pretend to do then you know it can go either way. Right now you are convinced that Obama will win because he had a popular vote bounce after his convention. That is what the typical guy on the street thinks every time his (or the other) team pulls ahead. Congratulations on that spectacular insight.
But as Karl Rove and Joe Trippi and others who actually make a living following this stuff will tell you, the democrats usually get a 4.9 pt bounce after their convention. This held pretty true this time around. But as they will also tell you their bounce is less likely to stick than a modest republican bounce. Go figure…guys who follow it for a living know more than say YOU!
So, sure Obama could win. As I’ve been saying for months on end he is the favorite to win for many reasons. But claiming that he’s going to win because YOU spent exactly 1:00 glancing at the latest poll well…don’t get all upset if I’m not jumping up and down at your incredible prognostication ability.
So yeah based on your (lack of) analysis I think it is false.
[/quote]
I did not spend 1.00 minute looking at the latest poll and where did it say I was upset? I’m not even mildly bugged that you disagree. You are the one writing paragraphs about it, not me. I just looked at the trend from 138 polls and this was my conclusion. Sorry that you got your panties in a bunch.
jnd
[quote]countingbeans wrote:
[quote]jnd wrote:
Again- if you look at all of the polls as a whole, then this is over.
[/quote]
I am happy to see your input, but can you put a bit more effort into it than this?
Your statement by itself is just an appeal to authority at this point.
If you are right, can you explain why? And no “look at the polls as a whole” isn’t explaining it, seeing as when I read things like this http://www.nationalpolls.com/stories/2012/0911-examiner-romney.html your post seems… I don’t know, baseless.[/quote]
Beans-
That is a very fair request. If you head over to realpolitics you can get all of the data from this seasons polls. If you look at the majority of the polls (something like 138 in the last year, only a small number go towards Romney). The law of large numbers tells me that any one poll is useless (hope you read that ZEB), but the combined results from all of the polls is really telling. This perspective says that Romney carried less than 20% of the polls. All of them were close, but Obama leads in so many it is hard to ignore.
I might be wrong, but the combined effect of all of those polls going in the same direction gives me enough confidence that you don’t need a chi-square test to see the trend.
jnd
[quote]jnd wrote:
[quote]countingbeans wrote:
[quote]jnd wrote:
Again- if you look at all of the polls as a whole, then this is over.
[/quote]
I am happy to see your input, but can you put a bit more effort into it than this?
Your statement by itself is just an appeal to authority at this point.
If you are right, can you explain why? And no “look at the polls as a whole” isn’t explaining it, seeing as when I read things like this http://www.nationalpolls.com/stories/2012/0911-examiner-romney.html your post seems… I don’t know, baseless.[/quote]
Beans-
That is a very fair request. If you head over to realpolitics you can get all of the data from this seasons polls. If you look at the majority of the polls (something like 138 in the last year, only a small number go towards Romney). The law of large numbers tells me that any one poll is useless (hope you read that ZEB), but the combined results from all of the polls is really telling. This perspective says that Romney carried less than 20% of the polls. All of them were close, but Obama leads in so many it is hard to ignore.
I might be wrong, but the combined effect of all of those polls going in the same direction gives me enough confidence that you don’t need a chi-square test to see the trend.
jnd[/quote]
Obama’s biggest lead, according to here RealClearPolitics - Election 2012 - General Election: Romney vs. Obama since hitting 4.6 in April, was 4.6 on August 11th.
So they have tied been even since April.
Shit he got a 3.6 point lead after his convention, which is still bascially tied.
[quote]countingbeans wrote:
[quote]jnd wrote:
[quote]countingbeans wrote:
[quote]jnd wrote:
Again- if you look at all of the polls as a whole, then this is over.
[/quote]
I am happy to see your input, but can you put a bit more effort into it than this?
Your statement by itself is just an appeal to authority at this point.
If you are right, can you explain why? And no “look at the polls as a whole” isn’t explaining it, seeing as when I read things like this http://www.nationalpolls.com/stories/2012/0911-examiner-romney.html your post seems… I don’t know, baseless.[/quote]
Beans-
That is a very fair request. If you head over to realpolitics you can get all of the data from this seasons polls. If you look at the majority of the polls (something like 138 in the last year, only a small number go towards Romney). The law of large numbers tells me that any one poll is useless (hope you read that ZEB), but the combined results from all of the polls is really telling. This perspective says that Romney carried less than 20% of the polls. All of them were close, but Obama leads in so many it is hard to ignore.
I might be wrong, but the combined effect of all of those polls going in the same direction gives me enough confidence that you don’t need a chi-square test to see the trend.
jnd[/quote]
Obama’s biggest lead, according to here RealClearPolitics - Election 2012 - General Election: Romney vs. Obama since hitting 4.6 in April, was 4.6 on August 11th.
So they have tied been even since April.
Shit he got a 3.6 point lead after his convention, which is still bascially tied.
[/quote]
How is that basically tied? Over the number of polls the margin of error shrinks…It does not grow. With all of those polls (and the shrinking error term), that 3.6 lead is massive.
Am I missing something with your view of those polls?
jnd
[quote]jnd wrote:
[quote]countingbeans wrote:
[quote]jnd wrote:
[quote]countingbeans wrote:
[quote]jnd wrote:
Again- if you look at all of the polls as a whole, then this is over.
[/quote]
I am happy to see your input, but can you put a bit more effort into it than this?
Your statement by itself is just an appeal to authority at this point.
If you are right, can you explain why? And no “look at the polls as a whole” isn’t explaining it, seeing as when I read things like this http://www.nationalpolls.com/stories/2012/0911-examiner-romney.html your post seems… I don’t know, baseless.[/quote]
Beans-
That is a very fair request. If you head over to realpolitics you can get all of the data from this seasons polls. If you look at the majority of the polls (something like 138 in the last year, only a small number go towards Romney). The law of large numbers tells me that any one poll is useless (hope you read that ZEB), but the combined results from all of the polls is really telling. This perspective says that Romney carried less than 20% of the polls. All of them were close, but Obama leads in so many it is hard to ignore.
I might be wrong, but the combined effect of all of those polls going in the same direction gives me enough confidence that you don’t need a chi-square test to see the trend.
jnd[/quote]
Obama’s biggest lead, according to here RealClearPolitics - Election 2012 - General Election: Romney vs. Obama since hitting 4.6 in April, was 4.6 on August 11th.
So they have tied been even since April.
Shit he got a 3.6 point lead after his convention, which is still bascially tied.
[/quote]
How is that basically tied? Over the number of polls the margin of error shrinks…It does not grow. With all of those polls (and the shrinking error term), that 3.6 lead is massive.
Am I missing something with your view of those polls?
jnd[/quote]
Aside from the fact real life doesn’t always follow the polls, I don’t find a sudden spike after a convention after 6 months of basically neck and neck to be massive.
I mean, it has been awhile since my last stat class, but the margin of error is never zero, and a 3.6 point lead 60 days out isn’t some end game figure that couldn’t be well wrong in either direction.
I think you are confusing the 3.6 point lead with a trend that isn’t there. If it was to follow the trend that is there, it will shrink back down to even.
Ohio is the biggest thing for Romney to worry about right now. His own strategists told Politico last week that without Ohio, his road to the white house is extremely difficult. And it’s not looking great right now.
Honestly, while he may not have been my guy, Romney isn’t going to lose the election for Republicans. This election might be a tipping point. But, when I think about it now, we may have already passed that point. There might be far too many Americans whose lives are directly touched by Government ‘generosity.’ From cradle, to college, to death bed. Some of our libertarian friends here lament the social conservative platform.
Seeing it as having become a permanent impediment to victories that could open the nanny state to sizable reductions. Brick by brick, eventually torn down, and bulldozed flat. My opinion is that it’s that sentiment, an ideology attached to anyone with an (R), that allows Obama to win. Because it won’t be his record earning him a victory.
[quote]countingbeans wrote:
[quote]jnd wrote:
[quote]countingbeans wrote:
[quote]jnd wrote:
[quote]countingbeans wrote:
[quote]jnd wrote:
Again- if you look at all of the polls as a whole, then this is over.
[/quote]
I am happy to see your input, but can you put a bit more effort into it than this?
Your statement by itself is just an appeal to authority at this point.
If you are right, can you explain why? And no “look at the polls as a whole” isn’t explaining it, seeing as when I read things like this http://www.nationalpolls.com/stories/2012/0911-examiner-romney.html your post seems… I don’t know, baseless.[/quote]
Beans-
That is a very fair request. If you head over to realpolitics you can get all of the data from this seasons polls. If you look at the majority of the polls (something like 138 in the last year, only a small number go towards Romney). The law of large numbers tells me that any one poll is useless (hope you read that ZEB), but the combined results from all of the polls is really telling. This perspective says that Romney carried less than 20% of the polls. All of them were close, but Obama leads in so many it is hard to ignore.
I might be wrong, but the combined effect of all of those polls going in the same direction gives me enough confidence that you don’t need a chi-square test to see the trend.
jnd[/quote]
Obama’s biggest lead, according to here RealClearPolitics - Election 2012 - General Election: Romney vs. Obama since hitting 4.6 in April, was 4.6 on August 11th.
So they have tied been even since April.
Shit he got a 3.6 point lead after his convention, which is still bascially tied.
[/quote]
How is that basically tied? Over the number of polls the margin of error shrinks…It does not grow. With all of those polls (and the shrinking error term), that 3.6 lead is massive.
Am I missing something with your view of those polls?
jnd[/quote]
Aside from the fact real life doesn’t always follow the polls, I don’t find a sudden spike after a convention after 6 months of basically neck and neck to be massive.
I mean, it has been awhile since my last stat class, but the margin of error is never zero, and a 3.6 point lead 60 days out isn’t some end game figure that couldn’t be well wrong in either direction.
I think you are confusing the 3.6 point lead with a trend that isn’t there. If it was to follow the trend that is there, it will shrink back down to even.
[/quote]
Life does not follow one poll, but it does follow many polls (otherwise, no one would ever poll).
You are correct- your stats teacher did a nice job- the error never hits 0, but it does reduce in size with additional samplings (polls). Just look at the MoE for each poll, larger number of people sampled, smaller error. What do you think happens when you start adding each of the samples–error becomes even smaller- which is why polls typically get about 1000 people or so, MoE hits 3%.
That 3.6% lead is small, but the error is really small and that little romney line rarely gets above the obama line. When it does, it drops back down right quick.
jnd
[quote]jnd wrote:
Life does not follow one poll, but it does follow many polls (otherwise, no one would ever poll).
[/quote]
Yeah, and the trend here is Obama isn’t pulling away, nor is he staying particularly far ahead of the margin of error on any one poll.
My point being, this isn’t the run away, based on these polls or any, that you implied in your posts.
[quote]
That 3.6% lead is small, but the error is really small and that little romney line rarely gets above the obama line. When it does, it drops back down right quick.
jnd[/quote]
I see it as an incumbant falling off for the challenger right quick after each jump, not including smoking him all spring.
[quote]countingbeans wrote:
[quote]jnd wrote:
Life does not follow one poll, but it does follow many polls (otherwise, no one would ever poll).
[/quote]
Yeah, and the trend here is Obama isn’t pulling away, nor is he staying particularly far ahead of the margin of error on any one poll.
My point being, this isn’t the run away, based on these polls or any, that you implied in your posts.
[quote]
That 3.6% lead is small, but the error is really small and that little romney line rarely gets above the obama line. When it does, it drops back down right quick.
jnd[/quote]
I see it as an incumbant falling off for the challenger right quick after each jump, not including smoking him all spring.[/quote]
I never implied it was a runaway in terms of a huge number, but it is over for the reasons I already explained.
He does not have to smoke him all the way,just be above him in the vast majority of the polls and it is over.
I am sticking by my prediction all the way. Are you predicting a Romney victory? If so- why? What evidence do you have to suggest such a thing?
jnd
[quote]jnd wrote:
Are you predicting a Romney victory? [/quote]
I’m not predicting anything. I’m simply saying the phrasing you use implies it isn’t as close as it is.
Incumbants don’t lose often, but the irratic spikes in this race don’t seem to follow what happens, at least not with Bush (sr), Clinton, Reagan or Carter, when elections are “done” as you say, 60 days out.
[quote] If so- why? What evidence do you have to suggest such a thing?
jnd[/quote]
What evidence do I have that Romney might win?
Have you read a newspaper lately?
[quote]jnd wrote:
I never implied it was a runaway in terms of a huge number, but it is over for the reasons I already explained.
jnd[/quote]
We’re all adults here. This statement contradicts itself.
[quote]countingbeans wrote:
[quote]jnd wrote:
Are you predicting a Romney victory? [/quote]
I’m not predicting anything. I’m simply saying the phrasing you use implies it isn’t as close as it is.
Incumbants don’t lose often, but the irratic spikes in this race don’t seem to follow what happens, at least not with Bush (sr), Clinton, Reagan or Carter, when elections are “done” as you say, 60 days out.
[quote] If so- why? What evidence do you have to suggest such a thing?
jnd[/quote]
What evidence do I have that Romney might win?
Have you read a newspaper lately?
[/quote]
It will be close in terms of % of votes, but again- Obama is up consistently and based on this I believe he will win.
I’ve read lots of newspapers, but I prefer to look at the polls, since they are not influenced by my opinion about the state of our country.
You really didn’t answer my question—you just asked another question.
What evidence is there that Romney will win? I’ve given you my reasons for saying Obama, what is your evidence for Romney?
jnd
[quote]jnd wrote:
what is your evidence for Romney?
jnd
[/quote]
Why Romney has a chance:
“A man can fail many times, but he isn’t a failure until he begins to blame somebody else.”
-John Burroughs
The economy sucks, people are broke, out of work, and angry
I feel that Obama is walking a thin line with his #1, and promoting class warfare. All it will take is one miscalculation, just one. Now I doubt this will happen because he is very very good at it, has a whole party actually backing this shit, but see the last part of #2 coupled with he thinks he is as good as he is.
It doesn’t seem voter turnout will be as high as 2008.
The 2010 elections
Scott Brown still has a lead over Warren in MA after she spoke at her convention and he didn’t even go to his.
I live in MA, and am running into more and more conservatives than ever before. One of the best things Obama ever did for me was piss me off to the point where I actually informed myself. I’m not the only one.
The economy sucks
[quote]countingbeans wrote:
[quote]jnd wrote:
what is your evidence for Romney?
jnd
[/quote]
Why Romney has a chance:
“A man can fail many times, but he isn’t a failure until he begins to blame somebody else.”
-John Burroughs
The economy sucks, people are broke, out of work, and angry
I feel that Obama is walking a thin line with his #1, and promoting class warfare. All it will take is one miscalculation, just one. Now I doubt this will happen because he is very very good at it, has a whole party actually backing this shit, but see the last part of #2 coupled with he thinks he is as good as he is.
It doesn’t seem voter turnout will be as high as 2008.
The 2010 elections
Scott Brown still has a lead over Warren in MA after she spoke at her convention and he didn’t even go to his.
I live in MA, and am running into more and more conservatives than ever before. One of the best things Obama ever did for me was piss me off to the point where I actually informed myself. I’m not the only one.
The economy sucks
[/quote]
Some of these reasons are totally based on your opinion. 130+ polls are not my opinion, but the facts.
jnd
[quote]jnd wrote:
[quote]ZEB wrote:
[quote]jnd wrote:
[quote]ZEB wrote:
[quote]jnd wrote:
[quote]treco wrote:
Think any of the polls are both scientifically derived (as much as a sample can truly represent 100MM voters) and non biased?
[/quote]
I wouldn’t trust any one poll. You need to look at the trend of all of the polls. Based on all of the polls, this election is over. Obama gets 4 more years.
jnd[/quote]
That’s absolutely false!
In 1980 Jimmy Carter lead Ronald Reagan almost right up to their first debate. When America saw the two next to each other they decided Reagan was a viable alternative. Ronald Reagan went on to carry 44 states and beat Carter by about 8 million votes. And just a few months prior to that he was down by 10 points! The voting public is very fickle.
Right now the very best polls show that Obama has about a 2-4 point advantage nationally. But what is more important Romney has made some great gains in various key states like Florida. Keep in mind that this is an electoral race not a popular vote race. Therefore. key states will determine the victor.
While I agree (and I’ve said from the start) that it will be difficult for Romney to defeat Obama there is no poll currently available that demonstrates that it is over. Not by a long shot![/quote]
There is nothing false about what I said. If you disagree with my view of the polls, that’s fine, but that does not mean what I said was false. Again- if you look at all of the polls as a whole, then this is over.
I’m willing to bet on it. Are you?
jnd
[/quote]
It’s my opinion that your opinion is false.
Are you really that offended?
If you actually follow it as you pretend to do then you know it can go either way. Right now you are convinced that Obama will win because he had a popular vote bounce after his convention. That is what the typical guy on the street thinks every time his (or the other) team pulls ahead. Congratulations on that spectacular insight.
But as Karl Rove and Joe Trippi and others who actually make a living following this stuff will tell you, the democrats usually get a 4.9 pt bounce after their convention. This held pretty true this time around. But as they will also tell you their bounce is less likely to stick than a modest republican bounce. Go figure…guys who follow it for a living know more than say YOU!
So, sure Obama could win. As I’ve been saying for months on end he is the favorite to win for many reasons. But claiming that he’s going to win because YOU spent exactly 1:00 glancing at the latest poll well…don’t get all upset if I’m not jumping up and down at your incredible prognostication ability.
So yeah based on your (lack of) analysis I think it is false.
[/quote]
So then you do not want to bet on this election???
If my analysis is wrong, then you should have no trouble betting… right?
jnd
[/quote]
I just got through explaining to you that while I always thought Obama was the favorite this is still a close election and either candidate could win. Certainly not one that I would bet on either way.
You probably need to take your shallow shoot from the hip conclusion to a near by bar and see if you can find a fellow Bubba to bet with. This thread was designed to discuss the various subtleties of the race.
The next time there is a poll shift and you have no clue as to why it happened think back to this thread. Sure it’s fun to bet, but it’s also fun to dig deeper into the why’s and how’s in order to obtain a fundamental understanding of the race.
It all might just be beyond you though…
[quote]jnd wrote:
Some of these reasons are totally based on your opinion. 130+ polls are not my opinion, but the facts.
jnd
[/quote]
Serious with this post?
You asked me for my opinion first off. Secondly how on earth is a poll of other people’s opinion a “fact”?
It is a fact that a poll has presented other people’s opinions in the way it has. But just because you lump a small sample of opinions together doesn’t suddenly turn opinion into fact. It very well may be an accurate prediction of what will happen, but it in no way turns opinion into facts.