2016 Postmortem
Related: About this forumPolling Averages Show Obama Near Victory
Greg Sargent runs through the latest polling averages and finds President Obama leading in Ohio, Wisconsin, Nevada, and Iowa while Mitt Romney leads in Virginia, North Carolina and Florida. The race is essentially tied in Colorado and New Hampshire.
"For the sake of argument, let's give the tied states to Romney. Here's the basic state of things: If you give Romney all the states where he is leading or tied in the averages -- Florida, Virginia, North Carolina, Colorado, New Hampshire -- he is still short of 270. Meanwhile, if you give Obama just the states where he leads in the averages, he wins reelection."
http://politicalwire.com/archives/2012/10/27/polling_averages_show_obama_near_victory.html
How the 'Swing States looked on this date in 2008 vs. Today:
Pennsylvania:
NBC/Mason-Dixon 10/27 - 10/28 625 LV 4.0 47 43 Obama +4
(2012 RCP average: Obama +4.7)
Ohio:
Mason-Dixon 10/29 - 10/30 625 LV 4.0 45 47 McCain +2
SurveyUSA 10/26 - 10/27 648 LV 3.9 49 45 Obama +4
(2012: RCP Average Obama +2.3)
Iowa:
SurveyUSA 10/28 - 10/29 658 LV 3.9 55 40 Obama +15 (Obama won by 9.5%)
(2012 RCP average Obama +2.3)
Colorado:
Marist 10/27 - 10/28 682 LV 4.0 51 45 Obama +6
FOX News/Rasmussen 10/26 - 10/26 1000 LV 3.0 50 46 Obama +4
(2012 RCP average tie at 47.8 for both Obama & Romney)
Nevada:
Rasmussen 10/27 - 10/27 700 LV 4.0 50 46 Obama +4
Reno Gazette-Journal 10/25 - 10/28 600 LV 4.0 50 45 Obama +5
(2012 RCP average Obama +2.5)
Florida:
Quinnipiac 10/27 - 11/2 1773 LV 2.3 47 45 Obama +2
ARG 10/29 - 10/31 600 LV 4.0 50 46 Obama +4
Mason-Dixon 10/29 - 10/30 625 LV 4.0 49 47 Obama +2
(2012 RCP average Romney +1.8)
Virginia:
Mason-Dixon 10/29 - 10/30 625 LV 4.0 47 44 Obama +3
Marist 10/26 - 10/27 671 LV 4.0 51 47 Obama +4
FOX News/Rasmussen 10/26 - 10/26 1000 LV 3.0 51 47 Obama +4
(2012 RCP average Romney +1.2)
North Carolina:
Mason-Dixon 10/29 - 10/30 625 LV 4.0 49 46 McCain +3
Research 2000 10/28 - 10/30 600 LV 4.0 45 47 Obama +2
Politico/InAdv 10/29 - 10/29 641 LV 3.7 48 48 Tie
Rasmussen 10/29 - 10/29 1000 LV 3.0 48 50 Obama +2
(2012 RCP average Romney +3.8--inflated by that Gravis poll of +8 for Romney)
Wisconsin:
Research 2000 10/27 - 10/28 600 LV 4.0 53 42 Obama +11
U. of WI 10/21 - 10/28 359 LV 5.2 52 42 Obama +10
(2012 RCP average Obama +2.3)
New Hampshire
ARG 10/28 - 10/30 600 LV 4.0 56 41 Obama +15
Suffolk 10/27 - 10/29 600 LV 4.0 53 40 Obama +13
Strategic Vision (R) 10/27 - 10/29 800 LV 3.0 50 41 Obama +9
(2012 RCP average Obama +1.1)
woolldog
(8,791 posts)I know its all about the electoral college.
But this seems to me to be unprecedented where you have 2 (respected) pollsters saying it's a comfortable Romney win in the popular vote and other pollsters like Rand, IBD, Reuterus, ABC/Wapo saying it's an Obama win or tied. Add to that Nate Silver, Wong etc looking at the polling and claiming an Obama popular vote win. I've never seen anything like it. Someone is going to be VERY wrong. Even in 2000 there weren't pollsters saying Bush had 50+ of the vote like Gallup and Ras are saying.
geek tragedy
(68,868 posts)in the popular vote.
National polling will either be very right or very wrong.
woolldog
(8,791 posts)IBD and Rand saying Obama is ahead. Gallup and Ras saying Romney is comfortably ahead.
Someone is certain to be very wrong.
geek tragedy
(68,868 posts)The one that worries me is the ABC/WaPo poll.
woolldog
(8,791 posts)geek tragedy
(68,868 posts)Was 50-47 Rmoney earlier in the week.
woolldog
(8,791 posts)Romney, or something like that, haven't checked it today
geek tragedy
(68,868 posts)graham4anything
(11,464 posts)it don't take into account the massive Hispanic growth
and they are oversampling the south
add 3 or 4 other rightwing polls and it skewers the aggreggate total.
woolldog
(8,791 posts)Maybe they shouldn't be, but they are. And I'm not talking about the averages.
I'm saying roughly half the polls are saying Obama wins, half are saying Romney. Never seen anything like this in my lifetime.
graham4anything
(11,464 posts)Maximumnegro
(1,134 posts)Anyone who's been following the way Ras operates has to come to the conclusion that he is narrative first, accuracy second. Wherever there is a good Dem poll, lo and behold look who pops up there right after. All year long. Or driving the narrative like he is doing w/ WI as a 'tie' in order to validate Romney's needing to go there because OH is lost. At this point there are no excuses, there is no I dunno who's wrong. Gallup has been shredded by the campaign and has piss poor history.
So anyone who actually feel doubt and fear is wasting time and energy.
Jennicut
(25,415 posts)Except Rand, Rasmussen and Gallup. Rand measures the same people over and over and I look at it as measuring if the same people will stay with Obama or not, not necessarily a national poll. Some people seemed to flirt with voting for Romney but then have moved back to Obama.
Rasmussen and Gallup have two separate issues going on. Rasmussen has a voter id screen in only changes monthly (!) and if it is on the low end for Dems then his polls will naturally show Repubs gaining in them. Gallup has issues with it's voter ID screen. 5 points between LV and RV is really, really high. That would mean Dem enthusiasm is really low. And early voting is challenging that assumption. Also, Gallup's data suggest a white turnout of nearly 80 %. It was about 74% in 2008 and is predicted by most who look at the demos as 73% white. http://www.dailykos.com/story/2012/10/21/1147769/-Gallup-s-LV-model-has-Non-Hispanic-white-vote-at-80