Pro-Obama Polling

[Shady]

For some reason, the polls lately have skewed significantly in favour of President Obama. I don't know if it's just accidental, or if it's a conerted effort by the mainstream media, but the sampling in the polls since just before the DNC convention has been highly suspect.

First we'll start with CNN's latest garbage:

CNN released a poll showing Obama 52 Romney 46. At first glance, anyone would assume that it's legit. But when you look at their math, and compare it to 4 years ago, it doesn't make any sense at all. They have Romney winning independents by 14 points, but somehow trailing Obama by 6 points overall. That's some pretty fuzzy math considering Independents made up 30% of the electorate in '08 (Dems made up 39% and Reps made up 32%.

CNN doesn't publish it's sampling percentages, so it's impossible to tell for sure how they're polling.

Next we'll look at the recent PPP poll showing Obama up 5 in Ohio:

Their sampling is ridiculous. They sampled 41% Dem, 37% Rep, and only 22% Independent. In 2010, the actual split was 36% Dem, 37% Rep and 28% Independent. Back in 2008, at the height of Hope n Change, the split was 39% Dem, 32% Rep, and 30% Independent. So apparently they're predicting a record turnout in Ohio for Obama, even bigger than in '08, despite Ohio being a traditionally red state. Not only that, but in '08, Obama won Independents by 8 points, and the state itself by 5 points overall. However, this poll shows Romney leading Independents by 2 points. That's a 10 points swing from Obama in '08, yet he's supposidly winning the state by the same margin? Riiiight.

Independents composed 30% of the electorate in both '08 and '10, yet for some reason PPP thinks that Independents will only make up 22% this time around. Riiiight.

Lastly, let's look at the newest poll, the Washington Post/ABC News poll:

They have Obama leading Romney by 1 percent, despite outsampling Dems 33% to Reps 23%. That's a 10 point difference! Never in the history of elections has there ever been that kind of Dem advantage. Like I said, even at the height of Hope n Change, the Dems only had a 6 point advantage. And even at the height of Hope n Change, Republicans still composed 32% of the electorate. Not 23%. This poll is essentially media malpractice.

In conclusion, a poll is only as good and as accurate as the sampling it's based on. So make sure you look at the sampling data before believing any of the results.

[punked]

Lastly, let's look at the newest poll, the Washington Post/ABC News poll:

They have Obama leading Romney by 1 percent, despite outsampling Dems 33% to Reps 23%. That's a 10 point difference! Never in the history of elections has there ever been that kind of Dem advantage. Like I said, even at the height of Hope n Change, the Dems only had a 6 point advantage. And even at the height of Hope n Change, Republicans still composed 32% of the electorate. Not 23%. This poll is essentially media malpractice.

In conclusion, a poll is only as good and as accurate as the sampling it's based on. So make sure you look at the sampling data before believing any of the results.

We can talk about all your other errors later. The sample doesn't really matter when they weigh the poll Shady. Just because they can only find 23% republicans when they call (which is a bad sign for your guys not Obama) only means something in the RV numbers. They weigh the polls after to reflect the LV numbers based on who is saying they are going to vote.

Which is why your claim on this poll is BULL SHIT. This poll has Obama leading RVs by 6% only when they re weigh the sample (to whatever they think but no one else thinks the electorate is going to look like) does this race become closer. Sorry buddy polls don't work the way you want them to just cause.

CNN has Obama up 8. I suspect the truth lies somewhere in the middle where gallup, ppp, and Ras all lie which is Obama up 5-6 points and killing in the swing sates.

BTW 2010 is a bad number to look at. 2010 is an off year. Know who votes less in an off year by a lot? Youth, Blacks, and Hispanics.

If you want some bad numbers I suggest you look to polls that think the electorate is going to look like 2010. This isn't 2010 the Democrat running for senate in ARIZONA, and MONTANA are either tied or winning right now.

Edited September 11, 2012 by punked

[BubberMiley]

Random sampling is just that.

But why didn't you mention Rasmussen, which sampled Republicans heavily and still came up with a strong Obama lead?

[Shady]

Just because they can only find 23% republicans when they call

Polls aren't done with random sampling. They're done based on models. You're obviously completely clueless. No real surprise.

Random sampling is just that.

Once again, polls aren't done with random sampling. Are you seriously that dense?

[BubberMiley]

I love debating with you because it's so easy.

Here's the very first line of the poll:

This Washington Post-ABC News poll was conducted by telephone September 7-9, 2012, among a random national sample of 1,002 adults, including landline and cell phone-only respondents.

http://www.washingtonpost.com/wp-srv/politics/polls/postabcpoll_20120909.html

If you were a person of integrity, you would say "Sorry. My mistake."

[Shady]

I love debating with you because it's so easy.

Here's the very first line of the poll:

http://www.washingtonpost.com/wp-srv/politics/polls/postabcpoll_20120909.html

If you were a person of integrity, you would say "Sorry. My mistake."

Yes, I read that. But do you think they just mash their hands against a keypad to decide which locations to call?

[Shady]

A 100% random poll is worthless. Polls are based on a statistical model. That's why you get the margin of errors built in. Why am I having to explain common knowledge to you people?

[BubberMiley]

You don't know what random sampling means, do you?

[punked]

Polls aren't done with random sampling. They're done based on models. You're obviously completely clueless. No real surprise.

If you are right why did the RV numbers go from Obama 50 Romney 44 to a LV model then?

From the Poll:

This Washington Post-ABC News poll was conducted by telephone September 7-9, 2012, among a LOOK RIGHT HERE SHADY random national sample of 1,002 adults, including landline and cell phone-only respondents. Results for the full sample have a margin of sampling error of 3.5 percentage points. The error margin is four points for the sample of 826 registered voters, and 4.5 points for the sample of 710 likely voters. Sampling, data collection and tabulation by Abt-SRBI of New York.

http://www.washingtonpost.com/wp-srv/politics/polls/postabcpoll_20120909.html

[punked]

A 100% random poll is worthless. Polls are based on a statistical model. That's why you get the margin of errors built in. Why am I having to explain common knowledge to you people?

NO Likely Voters which is the number you quoted is the statistical model Shady. Which is why I said you were lying because you were. BTW ABC does not tell you what their weighing is for the LV number it could be R:50% D:20% I:30% we just don't know. They just weigh them from the response with their model from the RV pool.

Edited September 12, 2012 by punked

[Shady]

IF YOU HAVE BASIC METHODOLOGICAL QUESTIONS, PLEASE E-MAIL

[email protected], OR CONSULT THE FINAL PARAGRAPH

OF THE PRESS RELEASE

PPP

What methodology, right punked and Bubber? It's all completely random!

/facepalm

[punked]

What methodology, right punked and Bubber? It's all completely random!

/facepalm

OMG Shady you have no idea what you are talking about. Go read up on how polls are done.

[GostHacked]

Polls don't matter.

[Shady]

You don't know what random sampling means, do you?

You don't know how the Washingto Post polling firm conducts its polls, do you?

Methodologically, in all or nearly all cases we require a probability sample, with high levels of coverage of a credible sampling frame. Self-selected or so-called convenience samples, including internet, e-mail, blast fax, call-in, street intercept, and non-probability mail-in samples do not meet our standards for validity and reliability, and we recommend against reporting them.

Washington Post/ABC

Bubber and punked say, what's a probability sample? What's a credible sampling frame?

It's 100% random right!

Edit for /facepalm

Edited September 12, 2012 by Shady

[punked]

You don't know how the Washingto Post polling firm conducts its polls, do you?

Bubber and punked say, what's a probability sample? What's a credible sampling frame?

It's 100% random right!

Edit for /facepalm

Your link doesn't work. Although I don't think you know what you Quoted means honestly I believe you have no idea what you are talking about so you think you are right. This is the very first case where I feel you don't know you are wrong when you clearly are and aren't just pretending to be thick.

Here I think I know how I can show you are wrong.

Pollsters always have to post the order they ask the questions in. This is because if you are running a push poll you will ask questions like "Do you think Mitt Romney eats Babies" before they ask the question of who you will vote for.

Ok so go to the poll. The last very last question asked is:

901. Generally speaking, do you usually think of yourself as (a Democrat), (a Republican), an independent or what?

That is because this is completely random sample. They weigh the sample in LVs for their head line poll number which is why you will find only one question breaks it into LVs.

This is also From the poll Shady:

The error margin is four points for the sample of 826 registered voters, and 4.5 points for the sample of 710 likely voters.

You see how they lost more then 15% of their sample from the RVs to the LVs. That is so they can reweigh the poll to be 50% male 50% female, probably by race, and to make it so that the Ds, Rs, and Is are somewhere between 2008 and 2010 make up.

Edited September 12, 2012 by punked

[TheNewTeddy]

Polls aren't done with random sampling. They're done based on models. You're obviously completely clueless. No real surprise.

Once again, polls aren't done with random sampling. Are you seriously that dense?

models are applied after random sampling. Stop trying to mislead people.

[JerrySeinfeld]

Ohio won't vote for this loser twice, The whole "you black folks need our government cheques" thing might work in cleveland, but I'm inclined to think the people who actually work for a living might come out in droves this time out.

[punked]

PPPoll coming out tonight I hear Obama is leading big in Virgina and Warren in Mass.

First one in.

We find Barack Obama leading 51-46 in VA, same margin as our August poll. D+3 sample, down from D+6 in 2008. Full poll up by 9

Edited September 17, 2012 by punked

[BubberMiley]

PPPoll coming out tonight I hear Obama is leading big in Virgina and Warren in Mass.

I'm surprised he isn't leading by more. The Republican campaign has been a complete joke this time around. I wonder whose idea it was to base their convention around a misrepresentation of Obama's words. I think they'll soon realize that compulsive lying isn't the best campaign strategy.

"Three hours, fifty-something minutes!"

Edited September 17, 2012 by BubberMiley

[punked]

I'm surprised he isn't leading by more. The Republican campaign has been a complete joke this time around. I wonder whose idea it was to base their convention around a misrepresentation of Obama's words. I think they'll soon realize that compulsive lying isn't the best campaign strategy.

"Three hours, fifty-something minutes!"

Virgina is a Confederate state. Romney needs to flip it.

[Shady]

So far this is hands down the best piece of propaganda polling I've seen yet...

Poll: Obama has foreign policy advantage

CBS

Yep. Obama has an advantage when polling 13% more Democrats than Republicans. Gee, what a suprise.

The CBS poll is weighted the following: 35% Dems, 22% Reps. Let's remember that at the height of Hope n Change back in the election of '08, Democrats held a 7 point advantage. Apparently CBS thinks its gonna be double this year.

[punked]

So far this is hands down the best piece of propaganda polling I've seen yet...

Yep. Obama has an advantage when polling 13% more Democrats than Republicans. Gee, what a suprise.

The CBS poll is weighted the following: 35% Dems, 22% Reps. Let's remember that at the height of Hope n Change back in the election of '08, Democrats held a 7 point advantage. Apparently CBS thinks its gonna be double this year.

Again they have a likely voter model Shady. How you think polls work and how they actually work are different. You would have a point if this was the only a RV model.

Here is what the poll actually is:

Total Likely Voters 1,162* 999

Republican likely voters 289 (29%)

Democratic likely voters 351 (35%)

Independent likely voters 359 (36%)

Edited September 18, 2012 by punked

[kimmy]

Man, Shady's really beyond help. There's no point trying to explain it to him, and frankly it's far more fun just to watch him shrieking about Liberal Media Bias Conspiracies.

I actually wish the polls showed the Republicans leading, though. The worst thing that could happen is if Democrat voters start thinking the election is in the bag and don't bother voting.

-k

[Shady]

Well, the latest sampling of the polls are out, and Obama's increased his lead. Not much of a surprise when you consider how they weighted the polls. It's borderline criminal.

Here's an example. In Florida in 2008, the party ID split amoung voters was 37% Democrat, 34% Republican, and 29% Independent. The New York Times samples their polls in an unrealistic 36% Democrat, 27% Republican and 33% Independent. They've subjectively made Florida a D+9 state instead of a D+3. Totally ignoring the Republican turnout advantage of 2010. That's why Obama's "increased his lead." Because they're titling the polling field and sampling more Democrats.

Not only that, but in the battleground polls of Ohio and Florida, Romney's winning amoung Independent voters, yet somehow still losing each state. FYI, if Romney wins Republicans and Independents on election day, Obama ain't winning Ohio or Florida. It's mathematically impossible. This is just more mainstream media propaganda.

[punked]

Well, the latest sampling of the polls are out, and Obama's increased his lead. Not much of a surprise when you consider how they weighted the polls. It's borderline criminal.

Here's an example. In Florida in 2008, the party ID split amoung voters was 37% Democrat, 34% Republican, and 29% Independent. The New York Times samples their polls in an unrealistic 36% Democrat, 27% Republican and 33% Independent.

Poor Shady.

Florida

2008 exits - 37 D 34 R 29 I

NYT/Quinnipiac - 36 D 27 R 33 I

So what are we seeing well all pollsters are saying this election is looking more and more like 2008 no matter how much Shady wants it to be 2010 they just can't see it. Sorry buddy. As for it being unrealistic it is just the same sample as 2008. That is realistic. Will happen maybe, but pretending it is impossible when it has happened in the last 4 years isn't out of touch.

BTW Gallup has Obama up 6. That is what Obama beat McCain by in 2008.

Edited September 26, 2012 by punked