Privacy guaranteed - Your email is not shared with anyone.
Separate names with a comma.
Discussion in 'Political Issues' started by PocketProtector, Oct 17, 2012.
Looks like Romney is the perceived winner from last night.
I wonder what DOC44 will have to say after posting this just 5 days ago:
Has the conspiracy suddenly ended?
Don't you get it? Bias exists if and only if a poll shows Obama in the lead.
Yeah, but you can't poll dead people, so it may not be that bad for Obama.
... & they polled D+9 .....
With a 9 point oversampling of democrats.
just think how much obama would be behind if the polling technque had not been changed.... we will never know.
Think how debate number 2 would have turned out if obama had not been in bed with candy.....
You need a lesson from -4.5 to +4.5 that another poster can give.
Bias is still there. If it is D+8 on this one, then the reality is closer to 55 vs 41.
Done. Time for the halo to Go find a professorship in Chicago.
What's wrong with sampling more Democrats than Republicans if more Americans identify with Democrats than Republicans?
What's wrong with conducting a true random poll of likely voters?
That is perfectly fine if it a representation of the actual numbers. You OBVIOUSLY don't understand what "oversampling" is, but thanks for playing
Well, Rasmussen has it 49/48, how does that fit into this whole thing?
The first clue is the first sentence in the 3rd paragraph, "reported as 3 day rolling ave".
Now there you go again......
reading the details and embarrasing the people that post just the headlines
Shame on you.
Ahhh, just current details.
In the near future though, BHO would follow the footsteps of Jimmy, a scorned 1-term ex prez who would chime in at every opportunity for attention.
He'd go from the "anointed one" to the "annoying one" very soon.
I'm going to embroider a little on countrygun's response.
Party ID doesn't drive turnout models. You can point to a Party ID poll showing Ds with a 10 point advantage, but it is irrelevant if they aren't likely voters and they don't show up on Election Day. No, TURNOUT drives turnout models. I'm going to use Ohio as an example because a) the numbers are easier to work with in the microclimate and b) Ohio polling is highly relevant right now.
Back in 2008 exit polling showed a D/R/I split of 39/31/30. In that election, Obama won the state 51.5 to 46.9. Take the exit polling and apply it to the vote by party you end up with a 52.8 to 45.6 total. In other words, exit polling was off somewhat, which is expected. Anyway, applying the exit polls against the share of the vote by party and comes up with a calculated D/R/I split of 37.5/32.5/30...that's what gets you a vote of 51.6 to 46.9.
All of this math just means that while the 2008 exit polls showed D+8, in reality it was around D+5. And that means a current poll in Ohio with a sample of D+5 or greater is saying that D turnout will be at or above 2008 levels.
If you look at the polls making up the Real Clear Politics average (which I hate, but whatever), the average split is D+5.4, which assumes a 2008-level turnout. I can't think of any persuasive arguments as to how Obama will match his 2008 turnout, but I can think of some arguments as to how he won't For example, we aren't dealing with a McCain here. Romney has the resources to keep Ohio competitive all the way to November 6. This is evident in the large crowds Romney has been drawing there. The enthusiasm the Rs are showing this time around go a long way to rebutting a 2008 turnout model.
Then there are the independents. In 2008 Obama won independents by 8 points. But in the RCP averages, Romney is winning independents by about 5 points. This is another nail in the coffin for Obama's reelection chances...you can't have a 13 point loss in independents without getting your D turnout even higher than in 2008.
That Romney is showing up as close (average polling I've seen has him down by 2) in polls that are predicting a repeat of the Democrat Tsunami of 2008 is a pretty strong indicator that, come election night, a lot of turnout models are going to be shown as faulty. There is nothing hypocritical about pointing out a poll that's stacked in favor of Obama is still showing Romney competitive.
Of course, this has some implications. For example, polling outfits have been trumpeting these D+5 or greater polls for weeks now. When the actual results come in and, hypothetically, Romney wins Ohio by 52 to 46, what's the predictable response from the Left? Yep, vote fraud allegations.
That is a factor, but so is the lack of enthusiasm on the Obama side. As I've said all along, what people say when they answer a question from someone calling them on the phone, and what (if any) action they take on election day are two different things.
Do not underestimate the lack of enthusiasm for Obama. This is a big plus for Romney.
So capsule summary: Dem + formula is BS, but that's their story and they're stickin' to it. (Until reality overruns them and makes them look ridiculous ....
That. . .is. . .exactly. . .what Gallup. . .did. . .
Yes, I OBVIOUSLY don't understand what "oversampling" is because you wrote OBVIOUSLY really big. Great argument.