2016 Postmortem
Related: About this forum10/2/ NC Update. Obama's lead continues to grow. 2 Polls show race tied.
While many are shouting the praises of Nate Silver, North Carolina shows just how the volume of right wing polls have undermined his model and made the race seem closer than it really is.
http://fivethirtyeight.blogs.nytimes.com/author/nate-silver/
Gravis has Romney up by 8 and Silver still gives the discredited fax/phone con operation 4 bars (and Ras only 3 ). And he gives the President only a 18% chance of winning NC.
Two new polls shows the race tied.
Elon University, who as far as we know doesn't derive most of its income from conning people into buying bogus trips to Florida, has the race tied. Actually the President won by exactly one respondent.
http://www.elon.edu/e-web/elonpoll/102912.xhtml
Obama is also polling much stronger among those who have voted early. The Elon Poll found the president leading Romney by a margin of 55-37. Among those who have not already voted, the challenger holds an advantage over Obama (48-43 percent).
PPP also shows a tied race.
http://www.publicpolicypolling.com/main/2012/10/presidential-race-all-square-in-north-carolina.html
The thing is either it is a tied race or Romney is 6-8 points ahead per Ras and Gravis. You cannot average facts with idiocy and come out with a reasonable 'middle'. As Axelrod says, we will see whether Mr. Silver was in the ball park in North Carolina or there was a whole bunch of dung being thrown about.
Now for the update from MattTX at DailyKos
http://www.dailykos.com/story/2012/10/29/1151889/-North-Carolina-Early-Voting-Day-11-30-of-NC-Af-Ams-Age-18-Have-Voted-New-Elon-Poll-Has-NC-Tied
30% of African Americans in North Carolina age 18+ have already voted.
That's 440,941 voters - 30% of North Carolina's entire 1,480,769 person African American voting age population counted by the Census Bureau in 2010.
That's not 30% of African American registered voters. That's 30% of all African American adults. Because, for one more week in North Carolina, you can register to vote and actually cast your ballot at the same time, all in one stop. So even if a would-be voter is not registered to vote right now, they can still register and vote.
But that 30% number will go much higher.
Because there's still another week of early voting left to go.
At this time in 2008, half of the African Americans who voted early had already voted. And so if we keep up at the pace we're going, another 440,000 African Americans in North Carolina will have voted by the end of the early voting period, bringing turnout up to about 60% of the North Carolina's total African American adult population.
Plus, then there's Election Day on top of that
. . .
So far, African American turnout is up from 357,160 to 440,941 votes, a 23.5% increase over this same time in 2008.
If that is not an amazing statistic, then I'm afraid you are just not the sort of person who can possibly be amazed by statistics.
Yes Romney may win North Carolina.
But he won't win it by much, if he does.
Nate Silver's projection of 82% is based on Silvers' bout of Gravisnesia, the acceptance of any 'polling result' without any knowledge if the numbers come from pollsters or 'pollsters'.
And here is the kicker.
If we win North Carolina then it is going to be an Electoral College landslide.
And here is the kicker X2
If Romney just barely wins NC then it is logical to assume that he is not going to win VA, OH and possibly FL.
Romney has to flip a bunch of states and NC is supposed to be the easy one (after Indiana).
If he can't do it here he will be thrown to the dust bin of history which he so richly deserves.
Maximumnegro
(1,134 posts)after the election when it comes to throwing any ol pollster into the mix.
ItsTheMediaStupid
(2,800 posts)He may miss this one, but he's one of the good guys in the business
grantcart
(53,061 posts)He can be rolled by simply producing a large number of right wing polls. In 2008 his model assumed that the left and right wing biases even each other out.;
This time there are a dozen right wing polls; Gravis, Ras, ARG, Purple Strategies, Etc.
There are no left wing polls. Zogby is basically gone.
PPP is a professional pollster, not a left wing one.
In fact Nate hasn't had a problem with deciding good from bad in the past:
Nate Silver, statistician of FiveThirtyEight.com, called Zogby International's online polling division, Zogby Interactive, "The worst pollster in the world." ranking it last in his Pollster Rankings. He also ranked Zogby's Telephone polls 53rd of 64.[7] He notes the pollster relies on one day polling which causes large fluctuations in data. Silver uses their misleading questions, seemingly to create a bias, as evidence for their status.[8]
this from Nate despite
In a review of the performance of polling firms during the 2008 elections, the Wall Street Journal noted that "Zogby International polled in eight states in the last week, including six of the closest races, and missed the final margin by an average of less than two points -- as accurate as the poll aggregators such as Pollster.com."[9]
I am not saying that Nate is a bad guy I am saying he has raging inconsistencies in his standards of pollster, in fact he doesn't have an objective standard at all or someone like Gravis who only started polling 90 days ago wouldn't have standing.
My take is that he is a liberal, but one that is so afraid of being accused of a liberal bias that he is willing to be tough on liberal pollsters but gives a pass from every conservative pollster to avoid criticism. I find his rating Ras over Gravis hilarious.
bigdarryl
(13,190 posts)or some bullshit like that.This seems to be his pattern
VirginiaTarheel
(823 posts)this was an old thread because it said "10/2".
Drunken Irishman
(34,857 posts)If North Carolina is a toss-up, or leaning Obama, then it's not hard to say the same about Ohio ... where Obama has a narrower lead than Romney in most polls. The best Obama has been able to do is tie, same, pretty much, with Romney in Ohio. Until these polls show Obama with the lead, I have no problem predicting it'll swing Romney ... just as I expect Ohio to swing Obama. But we can't have it both ways. Silver is assuming that Ohio will ultimately go Obama, even though the polling margins are pretty tight (less than 2 points).
So, if NC is moved to toss-up, then Nate probably has to move Ohio to toss-up as well.
ItsTheMediaStupid
(2,800 posts)He won't be calling that a tossup.
Drunken Irishman
(34,857 posts)Obama's advantage in Ohio is similar to Romney's in North Carolina. The best he can do is tie. The only poll that show Obama up in NC, at least recently, is from the recent partisan Grove Insight ... and even with that poll, Obama still trails the overall average by a wider margin than Romney in Ohio.
My point is that if North Carolina is a toss-up, then Ohio should be too. Some are looking at NC similar to how Republicans are looking at OH ... they focus on the good, positive polls and ignore the bad ones.
Well you can't do that.
Dawgs
(14,755 posts)Although, I don't think either state should be considered a toss-up.
Obama has ALWAYS had the lead in Ohio. That is not the case in North Carolina. Romney has a big lead in some polls now, and he's tied in others, but he has never had a consistent lead like Obama has had in Ohio. History matters considering Obama has lead there more than Romney. Just look at the history. The ONLY time Romney has had a consistent lead in NC has been the month of October. Obama can make a come back in NC where he has led more often than not, while Romney has NEVER led in OH.
And what bad polls in Ohio for Obama? Rasmussen (Rove's buddy) is the only one that has shown any kind of lead for Romney.
http://www.realclearpolitics.com/epolls/2012/president/oh/ohio_romney_vs_obama-1860.html#polls
http://www.realclearpolitics.com/epolls/2012/president/nc/north_carolina_romney_vs_obama-1784.html#polls
Drunken Irishman
(34,857 posts)I see no evidence Obama can win North Carolina. It's wishful thinking at this point. Yes, it could switch, but you don't count a state as a toss-up until it actually does change into one. Florida? Probably a toss-up. Virginia? Probably a toss-up. North Carolina? Probably a narrow Romney win ... similar to a narrow win for Obama in Ohio.
I just don't think Obama will win North Carolina and until I see the evidence he can, or he takes the lead in some other polls, I'm with Nate here ... North Carolina is likely to go to Romney.
grantcart
(53,061 posts)All of the polls that you cite make a distinction between likely and registered voters.
In NC there is no distinction, just people who vote and those that don't.
Show me a poll that shows the NC population favors Romney.
So far 30% of the African American NC population has voted.
Not 30% of registered AAs but of all AAs.
If this continues then the turnout of AA in NC could exceed 65-70% and no poll has accomadated that type of turn out in its model.
VirginiaTarheel
(823 posts)Which candidate shares your values more closely?
Obama: 47 percent
Romney: 45 percent
Which candidate is more honest and trustworthy?
Obama: 46 percent
Romney: 38 percent
Which candidate has the clearer plan for addressing the countrys problems?
Obama: 45 percent
Romney: 41 percent
Which candidate would be more likely to raise taxes on the middle class?
Romney: 49 percent
Obama: 37 percent
Dawgs
(14,755 posts)Not good news for Romney, especially when the base is pissed about new voter laws.
VirginiaTarheel
(823 posts)dsc
(52,162 posts)Elon was way, way off on Issue 2 which was the measure that banned marriage equality. They had us winning when we lost 61 to 39.
VirginiaTarheel
(823 posts)They did poll on voter sentiment on gay marriage, civil unions, and domestic partnerships, which is not the same as a poll directly on Question One. The poll exposed that many people who supported same sex unions ended voting for AMendment One because the message was not clearly communicated that Amendment One would also ban domestic partnerships and civil unions.
Elon Poll surveyed North Carolina adults not likely voters, not registered voters, but potential voters nonetheless about same-sex unions. About 6 in 10 favor some form of legal recognition, though of that group, not all are in favor of calling it marriage.
Just 1 in 3 are against any form of legal recognition.
This is, of course, what Amendment One would do if passed by the voters May 8. It would ban all forms of legal recognition ... no same-sex marriage, no civil unions, no recognition of any kind of "domestic legal unions" except for the marriage of one man to one woman.
Still, asked a straight-up question about the amendment, yes or no, the results keep coming back in favor of the amendment. That's because people don't know that it doesn't simply put the existing definition of marriage into the state constitution, it also prevents the state from enacting a law allowing civil unions
http://www.indyweek.com/citizen/archives/2012/04/02/elon-poll-6-in-10-in-nc-favor-same-sex-marriage-unions-or-partnerships
grantcart
(53,061 posts)JiminyJominy
(340 posts)Probably a dumb question but how do they know which guy is getting the votes? Are they exit polling? Or is it completely off party registration?