PPP has been GREAT this primary season! Don't count them out!

[adoraz]

To start my post, I still believe Clinton will win by 10% in PA.

HOWEVER-

Check out how good PPP has been.

What I am going to do is post every single Democratic poll they did right before the primary, and compare it to the actual result and the RCP average.

Remember- this is EVERY final poll.

South Carolina:
PPP- Obama +20
RCP Avg- Obama +11.6
Final- Obama +28.9
NOTE: Survey USA had Obama up 13 the same time (night before). NOBODY else had it at 20.

Georgia:
ppp- Obama +16
RCP Avg- Obama +18.0
Final- Obama +35.3
NOTE: Everyone got it wrong here, but they were consistent with the other polls.

Tennessee;
PPP- Clinton +22
RCP Avg- Clinton +20***
Final- Clinton +13.3
***- On the website the avg is lower, however the website has so few polls that it includes polls when Edwards was still in the running. As for the polls taken right before, her average was +20.
NOTE: Steady with the other polls. Clinton's average from PPP was almost exactly the same, what happened was most undecideds went to Obama.

New York:
PPP- Clinton +19
RCP Avg- Clinton +17.2
Final- Clinton +17.5
NOTE: Even though the RCP average was closer, most polls were further off than PPP was. PPP was only 1.5% off, while most the others were about 6% off in either direction. Survey USA had a 1 point difference.


So at this point, they look pretty creditable.

Every poll they have done has been pretty close to accurate.

They were also the closest in SC, while Survey had it off.


Now, their best has been their most recent. They have done 3 states since Super Tuesday (aside from PA):

Wisconsin
PPP- Obama +13
RCP Avg- Obama +4.3
Final- Obama +17.4
NOTE: They were the ONLY ones close to Wisconsin's final. Yes, this was a 10 point difference from the average and they got it correct.

Ohio
PPP- Clinton +9
RCP Avg- Clinton +7.1
Final- Clinton +10.1
NOTE: They outperformed most other polling, with only SurveyUSA doing a better job (they did 10%). Only a 1% differences though.

Texas
PPP- Clinton +6
RCP Avg- Clinton +1.7
Final- Clinton +3.5
NOTE: SurveyUSA did their final poll the same exact time period as PPP, and USA had Obama winning by 1%.





Now, what can we get from all this?

I think its obvious that PPP has never in the past OVERPOLLED Obama, they have only done that in favor of Clinton. Their polling is very respectable and they have done better than SurveyUSA and every single other poll with a few, such as in Wisconsin and South Carolina (in both cases, nearly double digits off from the other polling).

The question is, what are they doing differently?

Why is their polling so much different from the others in PA?

Well, for one they seem to be polling a lot more people than anyone else:

Rasmussen 722 LV
InsiderAdvantage 747 LV
Zogby Tracking 602 LV
Suffolk 600 LV
PPP (D) 2338 LV
Strategic Vision (R) LV
Quinnipiac 1027 LV
SurveyUSA 710 LV
Mason-Dixon 625 LV


This means they have the lowest MoE, but of course since every other poll is so different, they have to be doing something different from the others. All their past PA polls have been consistent with their newest one, so this isn't a fluke for themselves.

http://www.publicpolicypolling.com/pdf/PPP_Release_042108.pdf

I read over it, and this may be why:
White ...76%
African American ...18%

Whites seem to be a little lower than other polls and blacks seem to be a little higher. Overall though, it just seems Obama gets more support from PPP.




Point of this post:


YES, I believe Clinton will win by 10% still, HOWEVER, do not count PPP out. :)