Posted on 09/26/2012 12:18:38 PM PDT by smoothsailing
The poll also could have called more Republicans, but they got weeded out as unlikely voters. Considering that the Democrat base adores Obama and he's their dream candidate, and Romney is more disliked by the base than any Republican nominee in living memory, the polls are not hard to believe. Even the establishment that handpicked Romney did it because they said he would appeal to "independents." He might, but they assumed the base would come out no matter what just to vote against Obama. But many of them are sick and tired of the battered wife syndrome with the GOP and won't do it.
You decide what your total univese is ~ in politics that'd be Americans who vote ~ and then you randomly sample among them.
If you have unlimited resources you can set up an elaborate sample in some sort of matrix, or stratification ~ virtually none of these polls we see dragged in front of us on FR derive from a stratified sample selection process ~ they're just small polls where random sampling was used.
No one has ever had to jigger the data to come up with more Democrats and that's because there are always more Democrats. More recently they've added Independents but that's just garbage because there are no independents ~ and if you dig deep enough you can categorize them as Rep or Dem based on their own reported prior voting history.
Massively skewed polls will help make massive Dem vote-fraud-influenced results more credible.
Yep, we can even tag Romney as the "Comeback Kid" - that would really make some liberal heads explode.
I agree with you, in the end the pollsters will want to protect themselves. It will be entertaining to watch them try to explain it! LOL!
The magic is gone. There’s no way in hell the demholes have the same enthusiasm and turnout they had in ‘08. In their wildest dreams the demholes know it wont be a 2008-like turnout for their side. Conversely, everything that produced the Tea Party / conservative / republican enthusiasim & turnout in 2010 is still there, with a foreign policy cluster-f*#% to boot. I’ve seen some opinions that a reliable poll would need to split the turnout model difference 1/2 way between 2004 and 2008. I’d argue that half way between ‘08 and 2010 would be a better way to go.
I would laugh, but lies that jeopardize the commie’s defeat aren’t really funny.
“Recognize that in 2008 the Democrats turned out 15% more Democrat voters than the Republicans turned out Republican voters.”
Not even close. Where do you get your facts, Mother Jones?
You are wrong. Turnout is not random it’s as simple as that. Determining who will make up the voters (LV’s) each years is what separates good pollsters from bad ones.
Furthermore, according to some pollsters (ie Rasmussen) there are more self described Republicans than Democrats.
I had long suspected you were just a disruptor on here. This thread removes any doubt.
Your facts are not only completely incorrect, they are an exact replica of the Democrat talking points regarding the polls.
Nice work, Moby.
You Think??? How about Ohio? or Pennsylvania??
why not try a poll with 1/3 D 1/3 R and 1/3 I??
Not true, at least according to Dick Morris. DM claims Ras has a +2.5% Dem skew.
Simply dividing the difference in the two vote totals by the total number of votes don't tell you anything about the difference!
Either the Democrat vote was 14% larger than the Republican vote, or the Republican vote was 16% smaller than the Democrat vote.,P>Also, remember, the Democrat totals do not vary with the Republican totals. Most voters would prefer to NOT vote for the other brand every single time.
Here is what Rasmussen says are the current numbers and trends in party affilitation in the U.S.:
Republican 37.6 Democrat 33.3 Other 29.2 August 2012
Republican 37.0 Democrat 33.7 Other 29.3 November 2010
Republican 33.8 Democrat 41.4 Other 24.7 November 2008
I find it hard to believe that Democrats will have a higher turnout than Republicans six weeks from now.
You missed the R in Skrewed.
True. However, depending on the random digits involved (just the last four numbers, or were prefixes involved in the randomness?), you might be able to randomly call people in an area with a demographic most likely to achieve the desired result. Harlem vs. Manhattan, for instance...
So much depends on the finer points of the technique, and we don't have those to analyze.
Absolutely correct.
Chicago voter registration down compared with 2008, election officials say
When Chicago registration is down from '08, that means it's Democrat registration that is down. When it's down in that bastion of Democrat voters, it strongly suggests a national trend.
Id argue that half way between 08 and 2010 would be a better way to go
It makes sense, and is probably similar to the Rasmussen model.
You cannot sit down and randomly call just Democrats or just Republicans unless you have lists composed exclusively of such folks and all you want is a random sample of current Democrat or Republican thought.
Stop and think about this a moment. How is registration down in Chicago? They never remove anybody up there ~ so what happened?
He might use D+2 or D+2.5 for his LV turnout model but that isn’t what he has determine the true party ID percentages to be. The fact that he doesn’t use party ID as his only guidance for a LV model as some idiots argue supports my contention that is ISN’T random.
http://m.rasmussenreports.com/public_content/politics/mood_of_america/partisan_trends
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.