National Tracking Poll Thread

[afleitch]

Gallup:

LV - Romney 51(+1)/45(-1)

RV - Romney 48(+1)/46(-1)

Dear lord, how strict is their LV screen? Gallup is getting pretty close to jumping the shark.

Apparently Obama leads in all their 'regions' except the South where Romney has a something like a 22 point lead. As a result Romney leads nationally despite trailing everywhere else.

Here: http://www.gallup.com/poll/158048/romney-obama-among-likely-voters.aspx

[afleitch]

So.

RAND - Obama +4
Reuters - Obama  +3
TIPP - Obama +2
Rass - Romney +1
Gallup - Romney +6

Quite a spread.

[afleitch]

So.

RAND - Obama +4
Reuters - Obama  +3
TIPP - Obama +2
Rass - Romney +1
Gallup - Romney +6

Quite a spread.

So would the race basically be O+2, at the moment?

That is not what that averages out to. It'd be Obama +0.4. Which essentially means a tie.

I'd be inclined to think it is fair to drop the two outliers (RAND and Gallup) which would give Obama a 1.3% lead. Throwing YouGov in there which was O+1 would give the same result. Given Rassmussens favourability to Romney if you're partisan and chuck that out then Obama's lead is 2%

Speaking of potential outliers here's RAND!

Obama 50.03 (+0.96)
Romney 44.34 (-0.82)

[afleitch]

PPP launched their own 3 day tracker today

http://www.publicpolicypolling.com/main/2012/10/obama-romney-tied-nationally.html

Obama 48
Romney 48

[afleitch]

Once again, if you want to believe the Gallup poll that Mitt Romney is leading by 7% nationally, then you're going to have to not believe basically every single poll of every single state we've gotten in the past week. There is literally no evidence, other than the Gallup tracking poll, that Mitt Romney is leading by 7% or anything close to it.

And PPP's 4 point lead (based solely on a really good day for Romney) has now been superseded by their new poll. So we are left with outliers which if they look like outliers, usually are.

[afleitch]

PPP

Obama 48
Romney 47 (-1)

[afleitch]

Have to play catch-up post down time. PPP is tied again. 48-48

[afleitch]

UPI/C-Voter poll - Obama leads 49-47

[afleitch]

IBD/TIPP

Obama 47
Romney 44

Obama up 3 (was +2 yesterday)

[afleitch]

It was recorded as 50-46 yesterday

[afleitch]

Gallup's likely voter make up is a little confusing. What Gallup are comparing are their own figures from both 2004 and 2008 based on their pre-election polls, not the exit polls. Whites made up 82% of the voting electorate in 2004 and 78% in 2008 according to Gallup yet the exit polls had them 77% and 74% respectively.

The exit polls showed the following breakdown in 2008:

White 74%
Black 13%
Hispanic 9%
Asian 2%
Others 2%

Therefore if Gallup are not seeing a huge movement from 2008 then we would expect 2012 to look like this

White 74%
Black 12%
Hispanic 10%
Asian 2%
Others 2%

[afleitch]

Gallup will not be releasing polls again until Monday.

[afleitch]

Rasmussen have it tied at 48-48

[afleitch]

Rasmussen have it tied at 48-48

I'm not surprised. I think that there is a little obama bump due to the storm but I think that the bump will be over monday.

Possibly, but bear in mind that the movement in most tracker polls, national polls and the state polls began before the storm if you look at the survey dates.

[afleitch]

Rasmussen have it tied at 48-48

I'm not surprised. I think that there is a little obama bump due to the storm but I think that the bump will be over monday.

Possibly, but bear in mind that the movement in most tracker polls, national polls and the state polls began before the storm if you look at the survey dates.

not really in the ras and ppp polls.

PPP has a poll less than a week old for all swing states except Colorado and Nevada (no one seems to want to poll Nevada). Rassy's Ohio poll is old, and there are no polls within the past week from him for Florida, Virginia, North Carolina or Nevada.  Everyone of it's current swing state polls when compared to other pollsters are the outlier poll in the state excepting Wenzell which dumped a -3 poll in Ohio. He has to get his last poll's out in these states soon.

[afleitch]

PPP

Obama 50
Romney 48

http://www.publicpolicypolling.com/pdf/2011/112113114TrackingResults.pdf

This also shows 16% crossover Dems to Romney, 12% for Republicans to Obama (he'll be lucky to get 5%), great news for Romney.  Oh yeah and turnout is going to be D+5 , at least they are coming down to earth a little.

So let me get this straight, 12% of Republicans voting for Obama will never happen but 16% of Democrats going for Romney makes total sense? Okay then, hack.

Both of those numbers seem high, compared to the rest of the polls. 

Their own party I.D model; i.e who they allocate to 'Dem' 'Rep' etc is more sensitive than most polls. For example they had Dems at 41 and Reps at 39 which is higher than most other surveys who when pulled on average have the Dems at about 34 and the Reps at 27 (going by Pollsters data). Therefore their Dem and Rep identifiers contain a large number of people who would be labelled 'Independent' in other polls. As a result PPP's sub-samples tend to show larger numbers of Dems and Reps who would vote for the other side than you find elsewhere.

[afleitch]

RCP are saying that Rassmussen has Romney +1 - 49-48

[afleitch]

Romney seems to be re-gaining a tad in national polling.

On what basis? He's up 1 in Rasmussen. The last national poll released that had Romney ahead was Rasmussens poll of Romney up 2.

[afleitch]

Reuters/Ipsos (final?) poll says Obama leads 48-46.

He was up 1 in their previous poll so that is good.