• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

PolliGaf 2012 |OT5| Big Bird, Binders, Bayonets, Bad News and Benghazi

Status
Not open for further replies.
Seeing how some of these polls are tracking at the national level, does anyone do a poll AFTER the election asking "who did you vote for" and seeing if that matches up with the actual results?
 
"A new CNN Poll of Polls on Monday indicates the race in Ohio–perhaps the most decisive battleground this presidential cycle–is locked in a statistical dead heat."

wolf.jpg

(They are aware that aggregating polls reduces the margin of error right?)

i can`t wait to unsubcribe from CNN on Nov 7th and their bullshit horserace narrative
 
These guys are amazing.

http://nation.foxnews.com/2012-presidential-election/2012/11/05/romney-will-win

Are you kidding me. Talk about misleading your viewers.

I love it. This is great ammunition to discredit Fox "News".

Karl Rove: Rove says he believes the race will be very close, but predicts Romney gets between 279 and 286 votes in the electoral college and wins the popular vote by 2 to 2.5%.

GEORGE WILL: I forgot my exact number. I guess you have a graphic here. I guess the wild card in what I've projected is I'm projecting Minnesota to go for Romney.

Dick Morris: Morris says Romney will capture 325 electoral votes while Obama will get 213, a significant difference. . . . Morris tells van Susteren pollsters are oversampling Democrats and says a poll that claims Obama is up 3, really means Romney is winning by 4.

Michael Barone: Michael Barone explains why he believes Romney will win handily on Tuesday on FOX News' "Huckabee" program. Barone expects Romney to win Florida, Ohio, Wisconsin and Pennsylvania.

Peggy Noonan: Romney will win

Charles Krauthammer: Krauthammer Predicts Romney Win

OK, duly noted. This will be used to judge predictions you make in the future.
 

Angry Grimace

Two cannibals are eating a clown. One turns to the other and says "does something taste funny to you?"
BTW, we can't trust Gallup. Let's wait for some reliable polls.
/PD

I don't even get where the "oversampling" shit came from. Are they suggesting that the polls magically are different from every other poll since 1936? Polls as a whole that are done to standard are not subject to wild bias and the type of simple oversampling suggested here is literally a fundamental mistake at the very first level. I don't think every single pollster just forgot to adjust their polls.
 

pigeon

Banned
RCP: "Gallup? Never heard of them."

In fairness to RCP, they're very consistent about not allowing privately funded polls. This cycle it happens that most of the released privately funded polls were positive for Obama, and that's mostly because Obama's doing really well, so it looks like they're underplaying his lead. But if I were to run an aggregator I'd probably want to weight down or reject privately funded polls as well -- you know immediately that they wouldn't have been released if they didn't say what the funder wanted.
 

Averon

Member
I guess Gallup will save face after all. Their numbers were wild outliers. They were either detecting something in the electorate that every other polling firm's been missing or their model was out of whack.
 
I guess Gallup will save face after all. Their numbers were wild outliers. They were either detecting something in the electorate that every other polling firm's been missing or their model was out of whack.

Or reverse mittmentum. He had so much it backfired.
 

Angry Grimace

Two cannibals are eating a clown. One turns to the other and says "does something taste funny to you?"
Or reverse mittmentum. He had so much it backfired.

Silver's article today suggests that bounces (like the Debate 1 bounce) are just that, "bounces." They come back to earth - remember that Obama's numbers were inflated Pre-Denver due to both the DNC and the "47%" video.
 

Cloudy

Banned
In fairness to RCP, they're very consistent about not allowing privately funded polls. This cycle it happens that most of the released privately funded polls were positive for Obama, and that's mostly because Obama's doing really well, so it looks like they're underplaying his lead. But if I were to run an aggregator I'd probably want to weight down or reject privately funded polls as well -- you know immediately that they wouldn't have been released if they didn't say what the funder wanted.

I don't see why they allow robocall polls but not web polls like Ipsos and YouGov
 

Angry Grimace

Two cannibals are eating a clown. One turns to the other and says "does something taste funny to you?"
Bush 2004 and Obama 2008

Maybe I'm thinking of double digit victories
KuGsj.gif
KuGsj.gif


derp

He is. 33k thrown out already. 40k were thrown out in 2008

You are actively lying now. 33,000 ballots were not thrown out, they were sent absentee ballots a week late and you've been explained this fact multiple times. Moreover, the 40,000 you are citing are provisional ballots, not absentee ballots, meaning your comparison is nonsense and they were thrown out due to provisional balloting rules that don't exist anymore due to the 2010 Consent Decree.
 

Jackson50

Member
Obama barely outside the margin of error despite +11 dem ID lol. Fucking ridiculous poll
PhoenixDerp.
Bush bucked the trend because more Democrats identified as independents. The situation is similar to this cycle except the parties are reversed.
Seeing how some of these polls are tracking at the national level, does anyone do a poll AFTER the election asking "who did you vote for" and seeing if that matches up with the actual results?
Yes. The NEP is a consortium of media organizations that conducts exit polls which permit comparison. And the ANES conducts a panel study before and after the election.
Bush 2004 and Obama 2008
Bush 88.
 

Effect

Member
Ugh. Everything out of Ohio is making me nervous as hell. Any crap like this taking place in Virginia just in case that's needed?

Edit: Worrying about polling is one thing. Having to worry about someone trying to actively steal the election with a complete " I don't give a fuck" thought process is something else.
 

Angry Grimace

Two cannibals are eating a clown. One turns to the other and says "does something taste funny to you?"
Oh God, imagine the Freeper/Redstate brain vomit if those provisional ballots *actually* became the Florida-2000 of this election and the results hinged on a court ruling as to whether to accept those ballots: the presiding judge in charge is an African-American Clinton appointee.

"HERR DERR CONSPIRACY!!!!!"
 

Juice

Member
I love it. This is great ammunition to discredit Fox "News".

Karl Rove: Rove says he believes the race will be very close, but predicts Romney gets between 279 and 286 votes in the electoral college and wins the popular vote by 2 to 2.5%.

GEORGE WILL: I forgot my exact number. I guess you have a graphic here. I guess the wild card in what I've projected is I'm projecting Minnesota to go for Romney.

Dick Morris: Morris says Romney will capture 325 electoral votes while Obama will get 213, a significant difference. . . . Morris tells van Susteren pollsters are oversampling Democrats and says a poll that claims Obama is up 3, really means Romney is winning by 4.

Michael Barone: Michael Barone explains why he believes Romney will win handily on Tuesday on FOX News' "Huckabee" program. Barone expects Romney to win Florida, Ohio, Wisconsin and Pennsylvania.

Peggy Noonan: Romney will win

Charles Krauthammer: Krauthammer Predicts Romney Win

OK, duly noted. This will be used to judge predictions you make in the future.

Oh this will be fun.

The only intellectual defense: "we were advocating for our guy! The perception of winning increases turnout!"

Which of course means they'd rather manipulate the public than inform it.
 

pigeon

Banned
You are actively lying now. 33,000 ballots were not thrown out, they were sent absentee ballots a week late and you've been explained this fact multiple times. Moreover, the 40,000 you are citing are provisional ballots, not absentee ballots, meaning your comparison is nonsense and they were thrown out due to provisional balloting rules that don't exist anymore due to the 2010 Consent Decree.

Here's a quick thing to note about Ohio -- there were 5.7 million votes there in 2008. A 3% lead is around 170,000 votes that Husted has to magically vanish. And note that even in Obama's best counties, he's not going to win more than 70% or so* -- so throwing out ten votes indiscriminately only cuts four votes from Obama's margin. For fraud to erase that lead it needs to throw out over 400,000 votes from Cuyahoga County or similar. That's around 7% of the entire Ohio vote. Provisional ballots will probably only make up about 3%. So you could toss them all and probably not change the result -- or even get into recount territory.

This doesn't even mention, of course, that there are pollwatchers everywhere, that there are weeks before provisionals will be counted for the judge to overrule Husted's directive, etc., etc.

* Obama's tally in Alameda County, home to Oakland and Berkeley, was only 79% in 2008.
 

Tamanon

Banned
Oh this will be fun.

The only intellectual defense: "we were advocating for our guy! The perception of winning increases turnout!"

Which of course means they'd rather manipulate the public than inform it.

To be fair...that's kinda their whole job description.
 

Zzoram

Member
Here's a quick thing to note about Ohio -- there were 5.7 million votes there in 2008. A 3% lead is around 170,000 votes that Husted has to magically vanish. And note that even in Obama's best counties, he's not going to win more than 70% or so -- so throwing out ten votes indiscriminately only cuts four votes from Obama's margin. For fraud to erase that lead it needs to throw out over 400,000 votes from Cuyahoga County or similar. That's around 7% of the entire Ohio vote. Provisional ballots will probably only make up about 3%. So you could toss them all and probably not change the result -- or even get into recount territory.

This doesn't even mention, of course, that there are pollwatchers everywhere, that there are weeks before provisionals will be counted for the judge to overrule Husted's directive, etc., etc.

Maybe they'll do what Texas did and threaten the International Election Monitors and shut them out so they can do shady things in secret.
 

Angry Grimace

Two cannibals are eating a clown. One turns to the other and says "does something taste funny to you?"
Here's a quick thing to note about Ohio -- there were 5.7 million votes there in 2008. A 3% lead is around 170,000 votes that Husted has to magically vanish. And note that even in Obama's best counties, he's not going to win more than 70% or so* -- so throwing out ten votes indiscriminately only cuts four votes from Obama's margin. For fraud to erase that lead it needs to throw out over 400,000 votes from Cuyahoga County or similar. That's around 7% of the entire Ohio vote. Provisional ballots will probably only make up about 3%. So you could toss them all and probably not change the result -- or even get into recount territory.

This doesn't even mention, of course, that there are pollwatchers everywhere, that there are weeks before provisionals will be counted for the judge to overrule Husted's directive, etc., etc.

* Obama's tally in Alameda County, home to Oakland and Berkeley, was only 79% in 2008.

Moreover, most provisional ballots are filled out correctly.
 

OmniOne

Member
Did anyone else see Chuck Todd just go retarded?

He's so terrible. They're still on this Romney could win PA stuff. Not happening.
 
Status
Not open for further replies.
Top Bottom