Monday, October 29, 2012

You Know It's Over When
Even Fantasy Polling Doesn't 
Give Obama A Greater Than
A One Point Lead At This Stage

Plenty of pollsters are still using a 2008 turnout model in their weighting of their poll results of D+7.

The results the DMC is now getting to their D+7 fantasy polling must be terrifying them.  

Here's what today's Real Clear Politics National polling of the Presidential race looks like today on Oct 29:

http://www.realclearpolitics.com/epolls/2012/president/us/general_election_romney_vs_obama-1171.html



First, let's ignore all the polls that give Romney the lead for now and just look at the polls saying Obama has a 1 point lead or a tie at this stage with about 9 days to go.  All the post-third Presidential debate polls are from the 21st onward, so let's look at those.  

First up: the   IBD/TIPP poll taken from Oct 22nd to the 27th that found a 1 point lead for Obama: 



IBD/TIPP doesn't play games like many pollsters do and force you to do the math to figure out what their partisan split was.  CBS does that all the time.  IBD/TIPP shows they polled 7% more Democrats than Republicans for this poll and got a one point Obama lead, well within their margin of error of +/- 3.5%.

So. D+7 turnout predicted, a slim one point Obama lead.  On we go! 

Next is the Politico/GWU/Battleground poll taken from Oct. 22nd to the 25th that also found a one point lead for Barack Obama. 



Politico/GWU/BG  is a Reg. Voter survey that plays the cute game of never telling you exactly how their sample was weighted, they just give you the results.  Here it's 35% Democrat [43% with leaners] and 31% Republican [41% with leaners].  So that's a D+4 without leaners and a D+2 with.  Not too bad and pretty damn far from the D+7 or > fantasy polling we've seen for much of this year.  

So D+4 here gets Obama a 1 point lead.  

Next up is ABC News/Washington Post, a poll which was done from Oct. 25 to Oct. 28.  


UPDATE!!!! While I was making this post, from the time I took the screenshot of the RCP national poll results an hour ago, ABC News/WaPo has been changed from a one point Romney lead to a TIE.  It now looks like this: 


Here's the partisan split for that poll: 

Among likely voters that's a D+7 advantage.  And as I noted above, a while ago this poll claimed a one point Romney lead and now has changed that to a tie.  They must have adjusted the results. 

And finally: Pew Research's poll taken from Oct. 24-28, which also claims the race is currently a tie: 


Since they got caught putting out a D+19 poll in early August, Pew has changed how they reveal their sample and how they weight it.  In this latest poll they show you their unweighted sample size and composition, and then simply never get around to telling you if they weighted the results and if so, how.  570 R's and 560 D's is an almost 50/50 partisan split, but WITHOUT knowing how they weighted this poll, claiming a tie doesn't really tell us anything.  

OK those are the polls claiming Obama has a lead or a tie with Romney at the moment. 

All based on either a weighting of D+7 or hiding their weighting from the readers.  

Takeaway point: Even with D+7 fantasy polling they can't drag Obama higher than a tie or a 1 point lead.  

And their BS fantasy polling was always that: BS fantasy polling.  

With Republican voter ID now leading Democrats by one point 36/35 according to Gallup, and Republican voter enthusiasm higher than that of Democrats by over 10 pts, we're looking at a R+ turnout this election. 




The poll of 1,446 adults, taken Monday through Thursday, has a margin of error of +/- 3 percentage points.
Republicans have opened a big enthusiasm gap: 64% say they are more enthusiastic than usual about voting, compared to 48% of Democrats. 

So what about all those polls out there still plaintively trying to sell their 2008 turnout model of D+7?  

I feel sorry for them.  Almost.  

Here's what happens to their D+7 fantasy weighting of their polls if this is indeed a R+1 or greater turnout this election.  

That IBD/TIPP poll that finds Obama holding a 1 pt lead? Here's what that would look like with R+1 turnout on Nov. 6: 

D+7 = 1 pt Obama lead
D+6 = TIE
D+5 = 1 pt Romney lead
D+4 = 2 pt Romney lead
D+3 = 3 pt Romney lead
D+2 = 4 pt Romney lead
D+1 = 5 pt Romney lead
D+0/R+0 = 6 pt Romney lead
R+1 = 7 pt Romney lead. 

Oh hey look, I'm where Gallup has had Romney for more than 2 weeks, with a +/- 3.5 pt margin of error:  Somewhere between a 4 pt and a 7 pt lead over Obama.  

The DMC knows if this turnout on Nov. 6 isn't AT LEAST D+5 Obama is toast.  Their OWN POLLS ARE TELLING THEM THIS.  A D+5 or greater turnout for Democrats predicted and they STILL can't get Obama above a 1 pt lead or a tie with Romney.  

But watch Liberals across the fruited plain sail on in blissful unawareness, still thinking Obama's got this in the bag somehow when he's actually trailing right now by 4 points or more.  

Obama must have it in the bag somehow because NATE SILVER SAYS SO! 



My prediction for months has been a turnout on Nov. 6 between R+1 and R+3 and I'm sticking with it.  

1 comment: