Why the polls cannot be trusted - Operation Chaos

Some of you might have participated in my previous discussion on why the polls are flawed, including a look at the Bradley Effect and oversampling of Democrats. Well, this one is related to those, since it deals with why Democrats are being too heavily weighted in these polls. As many know, part of this is because of the number of new registered Democratic voters. What the prognosticators fail to realize, though, is that much of this is because of REPUBLICANS voting in the Democratic primaries as part of Rush Limbaugh's Operation Chaos. If this effort had such a big impact on the primaries (really, it never should have been close after Iowa), then of course it can have an effect on the poll numbers by giving a false impression of Democratic support.

I post this with the hope that Democrats might better understand the outcome on election night. Currently, Obama faces an uphill battle for the white house. Obama has no real chance winning any of the so-called swing states. Florida, Ohio and North Carolina should be solid for McCain. McCain will win all these states, but he also has to pick up a state that the media is saying is leaning to Obama, and New Mexico or Virginia should easily be won. That makes McCain the next president of the United States, meaning NOBODY is going to be spreading MY wealth around!

well good for you...

what is it with all these "Who do you think will win" thread, can't you just wait for the results?

Of course, things could change between now and the election, but it looks like the momentum is swinging in McCain's favor. At least now they're openly calling Obama a socialist. It's only a matter of time before more Americans realize that this is what Obama's all about.

well good for you...

what is it with all these "Who do you think will win" thread, can't you just wait for the results?

markop2003

As I said, I post these in an effort to reduce the shock on election night.

well good for you...

what is it with all these "Who do you think will win" thread, can't you just wait for the results?

markop2003

I think he wants to save the world.

[QUOTE="markop2003"]

well good for you...

what is it with all these "Who do you think will win" thread, can't you just wait for the results?

dave123321

I think he wants to save the world.

not just this thread but there's been a bunch over the months leading up too the election

Personally i think the process should take no longer than a month, they have a half hour speach at the begining and a half hour speach at the end before voting commences in the mean time you can read their polocies and that would be it, none of this touring the country and such

You just don't give up, do you?GabuEx

You just don't give up, do you?GabuEx

What does this contribute to the conversation? Of course I don't give up. I see so many people around here so sure of an Obama victory, because all they see are the biased polls on CNN. Well, I wish to tell them what's behind these polls, and why this race is closer than most people think.

[QUOTE="GabuEx"]You just don't give up, do you?Funky_Llama

Waiting to pounce? What are you planning to do, post it just to annoy me after the election if Obama wins? Real mature.

If you think I would come here after a McCain victory and rub it people's faces, then you must not know me well enough. We must move on after the elections and do what's best for America, not sit around acting like children.

You just don't give up, do you?GabuEx

You won't ever give McCain a chance will you?

[QUOTE="GabuEx"]You just don't give up, do you?mysterylobster

What does this contribute to the conversation? Of course I don't give up. I see so many people around here so sure of an Obama victory, because all they see are the biased polls on CNN. Well, I wish to tell them what's behind these polls, and why this race is closer than most people think.

Every single time I respond to one of your claims that the polls can't be trusted, you just find something new to bank on. I could go to all the trouble to show why Operation Chaos was not the huge force that you're asserting it is, but there'd be little point in doing so.

You won't ever give McCain a chance will you?

Nills

Historical precedent and analysis of the race thus far says both that polls can be trusted and that it is very likely at this point that Obama will be elected president. That has nothing to do with the fact that I personally support Obama.

You just don't give up, do you?GabuEx

Democrats are spewing crap on here too constantly. On the first page we have two threads: "Palin institute for Abstainance".. and "Do you think Cindy McCain looks evil?".

[QUOTE="GabuEx"]You just don't give up, do you?mysterylobster

What does this contribute to the conversation? Of course I don't give up. I see so many people around here so sure of an Obama victory, because all they see are the biased polls on CNN. Well, I wish to tell them what's behind these polls, and why this race is closer than most people think.

what does it mater?

Just wait to the election and make your vote, surveeys ain't gonna do anything so why pay attention to them?

what does it mater?

Just wait to the election and make your vote, surveeys ain't gonna do anything so why pay attention to them?

markop2003

Yep.

[QUOTE="mysterylobster"]

[QUOTE="GabuEx"]You just don't give up, do you?GabuEx

What does this contribute to the conversation? Of course I don't give up. I see so many people around here so sure of an Obama victory, because all they see are the biased polls on CNN. Well, I wish to tell them what's behind these polls, and why this race is closer than most people think.

Every single time I respond to one of your claims that the polls can't be trusted, you just find something new to bank on. I could go to all the trouble to show why Operation Chaos was not the huge force that you're asserting it is, but there'd be little point in doing so.

You won't ever give McCain a chance will you?

Nills

Historical precedent and analysis of the race thus far says both that polls can be trusted and that it is very likely at this point that Obama will be elected president. That has nothing to do with the fact that I personally support Obama.

A 5 point lead in the polls is not signifant at all. It's a close race. Not to mention, Republicans generally have a better voter turn out. There are still a lot of factors.. and the race is still close.

What would you rather have a socialist or a neocon fascist? I'd prefer neither and really can hardly fathom why so many americans would ever support either of the two.jer_1

they're politicians, when yhou vote you're simpley voting for what flavor of terrible you'ld prefer

Some of you might have participated in my previous discussion on why the polls are flawed, including a look at the Bradley Effect and oversampling of Democrats. Well, this one is related to those, since it deals with why Democrats are being too heavily weighted in these polls. As many know, part of this is because of the number of new registered Democratic voters. What the prognosticators fail to realize, though, is that much of this is because of REPUBLICANS voting in the Democratic primaries as part of Rush Limbaugh's Operation Chaos. If this effort had such a big impact on the primaries (really, it never should have been close after Iowa), then of course it can have an effect on the poll numbers by giving a false impression of Democratic support.

I post this with the hope that Democrats might better understand the outcome on election night. Currently, Obama faces an uphill battle for the white house. Obama has no real chance winning any of the so-called swing states. Florida, Ohio and North Carolina should be solid for McCain. McCain will win all these states, but he also has to pick up a state that the media is saying is leaning to McCain, and New Mexico or Virginia should easily be won. That makes McCain the next president of the United States, meaning NOBODY is going to be spreading MY wealth around!

mysterylobster

Lol you're going to be proven so wrong in a few days..

A 5 point lead in the polls is not signifant at all. It's a close race. Not to mention, Republicans generally have a better voter turn out. There are still a lot of factors.. and the race is still close.

The_Mac_Daddy

Depends on when you have the five-point lead. If the five-point lead were in summer, it'd be totally insignificant. When you have a five-point lead with only two and a half weeks before election day, however, all of a sudden it becomes much more significant on account of the fact that voters' minds tend become progressively harder and harder to change as the election year drags on and people's heels get dug into the ground.

Historically speaking, no presidential candidate has ever been down five points in the polls halfway through October and then went on to win the election. The only notable exception to that is Harry Truman in 1948, and polling has become much more rigorous and scientific since then.

Historically speaking, no presidential candidate has ever been down five points in the polls halfway through October and then went on to win the election.

GabuEx

I don't get it. I mention past voting patterns, and everyone, yourself included, says you can't compare this election to those that have come before, yet here you are doing just that.

Yes, there are some major differences in this election, but very few favor the Democratic candidate.

Of course, things could change between now and the election, but it looks like the momentum is swinging in McCain's favor. At least now they're openly calling Obama a socialist. It's only a matter of time before more Americans realize that this is what Obama's all about.

[QUOTE="markop2003"]

well good for you...

what is it with all these "Who do you think will win" thread, can't you just wait for the results?

mysterylobster

As I said, I post these in an effort to reduce the shock on election night.

lol if Obama's socialist, what about the UK, France, Germany, Switzerland, Canada, New Zealand, Austrailia? are they socialist as well?

I don't get it. I mention past voting patterns, and everyone, yourself included, says you can't compare this election to those that have come before, yet here you are doing just that.

Yes, there are some major differences in this election, but very few favor the Democratic candidate.

mysterylobster

You mentioned turnout in the last election. Turnout is something that can and does vary in every single election, so pointing to the last election and expecting this one to be the exact same makes no sense. Presidential candidates not making up five-point deficits in the last two weeks of the campaign, however, is not something that has changed over time.

Historical precedent can be observed only where it makes sense to do so given the evidence.

[QUOTE="mysterylobster"]

I don't get it. I mention past voting patterns, and everyone, yourself included, says you can't compare this election to those that have come before, yet here you are doing just that.

Yes, there are some major differences in this election, but very few favor the Democratic candidate.

GabuEx

You mentioned turnout in the last election. Turnout is something that can and does vary in every single election, so pointing to the last election and expecting this one to be the exact same makes no sense. Presidential candidates not making up five-point deficits in the last two weeks of the campaign, however, is not something that has changed over time.

Historical precedent can be observed only where it makes sense to do so given the evidence.

I don't know about five-point deficits, but Reagan definitely picked up at least 5 points on Carter, since he won by three points, and was trailing 10 days before the election (by as much as 8 points, according to one Gallup poll). So, historically speaking, it can be done.

I don't know about five-point deficits, but Reagan definitely picked up at least 5 points on Carter, since he won by three points, and was trailing 10 days before the election (by as much as 8 points, according to one Gallup poll). So, historically speaking, it can be done.

mysterylobster

The Gallup poll was one single poll; if you look at the aggregate polling of the 1980 election, you'll see that Reagan never trailed once in October.

[QUOTE="mysterylobster"]

I don't know about five-point deficits, but Reagan definitely picked up at least 5 points on Carter, since he won by three points, and was trailing 10 days before the election (by as much as 8 points, according to one Gallup poll). So, historically speaking, it can be done.

GabuEx

The Gallup poll was one single poll; if you look at the aggregate polling of the 1980 election, you'll see that Reagan never trailed once in October.

Another aggregate poll (which seems to be more comprehensive and better sourced) says Carter led by as much as three points in the final two weeks. Since Reagan went on to win by three points, that means he had a late 6 point turnaround.

Another aggregate poll (which seems to be more comprehensive and better sourced) says Carter led by as much as three points in the final two weeks. Since Reagan went on to win by three points, that means he had a late 6 point turnaround.

mysterylobster

I'm not sure if you know what "aggregate" means - it means you average everything out. Polls are subject to outliers - polls that unavoidably failed to gather a fully representative sample - so looking at one single poll is not really a very scientifically rigorous endeavor. If you take a look at all the polls in the study linked to by the article you provided, you'll find that the "Carter up 3" results were singular polls, not averages. If you average out the results for each week (thus effectively lessening the effect of outliers), you'll find that not once did Carter lead on balance.

By the way, that article you linked to is trying to dispel the notion that McCain might re-enact the "Reagan in 1980" effect, so I'm not sure why you're attempting to use it as evidence in favor of that very notion.

Well. Now I plan to start Operation Disturb the Order. To counter act the effect of Operation chaos.

And also Cause the Davidley effect to counter act the Bradley effect.

If you take anything Rush says seriously you must be on more pills than he is.

Heavy LOLz.

So you're hoping for the Bradley Effect?

You're hoping people are racist? Because that's surely what it sounds like...

So you're hoping for the Bradley Effect?

You're hoping people are racist? Because that's surely what it sounds like...

TenP

Operation Chaos is something different - it refers to a proposal by Rush Limbaugh whereby Republicans were supposed to vote for Clinton in the Democratic primary, thus dragging it on longer and causing the Democrats' chances to be lessened due to a bloody and expensive primary.

Given that the vast numbers of Republicans who voted in the Democratic primary did so for Obama (Hillary's strongest supporters were Democrats), I would say that the assertion that it had any noticeable effect is tenuous at best.

So you're hoping for the Bradley Effect?

You're hoping people are racist? Because that's surely what it sounds like...

TenP

So you're saying Obama will only lose because of racism? That's absurd.

Operation Chaos is something different - it refers to a proposal by Rush Limbaugh whereby Republicans were supposed to vote for Clinton in the Democratic primary, thus dragging it on longer and causing the Democrats' chances to be lessened due to a bloody and expensive primary.

Given that the vast numbers of Republicans who voted in the Democratic primary did so for Obama (Hillary's strongest supporters were Democrats), I would say that the assertion that it had any noticeable effect is tenuous at best.

GabuEx

Ahh... so he's hoping that the second biggest piss stain on the republican party (first is Ann Coulter) has enough influence on his sheep to make a difference.

[QUOTE="mysterylobster"]

Another aggregate poll (which seems to be more comprehensive and better sourced) says Carter led by as much as three points in the final two weeks. Since Reagan went on to win by three points, that means he had a late 6 point turnaround.

GabuEx

I'm not sure if you know what "aggregate" means - it means you average everything out. If you take a look at all the polls in the study linked to by the article you provided, you'll find that the "Carter up 3" results were singular polls, not averages. If you average out the results for each week (thus effectively lessening the effect of outliers), you'll find that not once did Carter lead on balance.

By the way, that article you linked to is trying tothe notion that McCain might re-enact the "Reagan in 1980" effect, so I'm not sure why you're attempting to use it as evidence in favor of that very notion.

"Aggregate" doesn't mean you average everything out. It means you take something as a whole. You can average out the results of an aggregate poll, but the averaging isn't what makes it aggregate.

Based on this aggregation of the polls, you can see that there were polls that showed Carter led by as much as 3 points.

So you're hoping for the Bradley Effect?

You're hoping people are racist? Because that's surely what it sounds like...

TenP

No, I'm hoping that they won't put political correctness over what's right for this country. A vote against Obama isn't racist.

it looks like the momentum is swinging in McCain's favor.

mysterylobster

when did the momentum start swingin to mccian? lately he's been getting mocked w/ all the joe the plumber stuff and people have been noticing all of his hate rallies...

but yah the polls cant be trusted...

"Aggregate" doesn't mean you average everything out. It means you take something as a whole. And based on this aggregation of the polls, you can see that there were polls that showed Carter led by as much as 3 points.

mysterylobster

ag·gre·gate
-adjective

1. formed by the conjunction or collection of particulars into a whole mass or sum; total; combined: the aggregate amount of indebtedness.

When you aggregate something, you take each individual item and put it all together for a summary look at the topic at hand. When the topic involves numbers, that means you average out the numbers and come up with one final aggregate number that takes all the data into account.

Aggregating polls means that you look at all the polls simultaneously; it does not mean that you cherry-pick one single poll that is a favorable result for your position while effectively ignoring all of the polls that say that it was a simple outlier. The fact of the matter, whether or not you wish to argue over the definition of a word, is that the average "poll of polls" for every week showed a lead for Reagan. Thus, it cannot be said that Reagan overcame a five-point deficit in the second half of October, because he never had such a deficit in the first place.

The final conclusion? 1980 is not an exception to the rule I laid out.

I hope for the sake of this country that you're right.bungie93

As do I. You would have thought people would've wised up by now. *shudders*

but yah the polls cant be trusted...

Ultima_5

People say this all the time, but polls predict the vast number of winning candidates every single election year. The idea that polls can't be trusted, quite simply, has no basis in fact.

[QUOTE="Ultima_5"]

but yah the polls cant be trusted...

GabuEx

People say this all the time, but polls predict the vast number of winning candidates every single election year. The idea that polls can't be trusted, quite simply, has no basis in fact.

the problem is that people can lie. i cant tell you how many telephone marketors i have lied to, but maybe it is just me.

the problem is that people can lie. i cant tell you how many telephone marketors i have lied to, but maybe it is just me.

Ultima_5

People can lie, but given how often polls get the right outcome (or have the final result within their margin of error, in the case of extremely close elections), it's pretty darn evident that people don't, in general.

When the topic involves numbers, that means you average out the numbers and come up with one final aggregate number that takes all the data into account.

GabuEx

You can also add the numbers together (as in aggregate sales) or simply analyze them. The act of bringing together bits of the data is what makes it aggregate polling, not the averaging. You should have used the term "aggregate average" poll. Just because my link wasn't an average doesn't make it non-aggregate.

I wouldn't put too much faith in aggregate averages, since it's influenced by the bias of the compiler. I think it's better to look at indivdual polls that have sound methods.

You can also add the numbers together (as in aggregate sales) or simply analyze them. The act of bringing together bits of the data is what makes it aggregate polling, not the averaging. You should have used the term "aggregate average" poll. Just because my link wasn't an average doesn't make it non-aggregate.

I wouldn't put too much faith in aggregate averages, since it's influenced by the bias of the compiler.

mysterylobster

Given that this post was a chance for you to point out the flaws in my overall statement that 1980 was not an exception to the rule, and given that you have not made any attempt to do so, I can only conclude that you agree that my assertion was correct.

[QUOTE="mysterylobster"]

You can also add the numbers together (as in aggregate sales) or simply analyze them. The act of bringing together bits of the data is what makes it aggregate polling, not the averaging. You should have used the term "aggregate average" poll. Just because my link wasn't an average doesn't make it non-aggregate.

I wouldn't put too much faith in aggregate averages, since it's influenced by the bias of the compiler.

GabuEx

Given that this post was a chance for you to point out the flaws in my overall statement that 1980 was not an exception to the rule, and given that you have not made any attempt to do so, I can only conclude that you agree that my assertion was correct.

If you define this rule as being based on an average of the aggregate polls, then yes, 1980 isn't an exception. However, I disagree with this method of analysis.

If you define this rule as being based on an average of the aggregate polls, then yes, 1980 isn't an exception. However, I disagree with this method of analysis.

mysterylobster

On what grounds?

You're basically asserting that the one single poll during the week that showed Carter +3 is better than the polls that showed Reagan ahead. Otherwise, you wouldn't see anything wrong with averaging the polls and coming up with the conclusion that Reagan was ahead on the whole. I see no logical reason to do this whatsoever.