New Colorado Poll: Obama 51% Romney 45%

[aaaaaa5a]

PPP Poll:

Obama 51%
Romney 46%



Obama was 3% points better than the last poll conducted by PPP around Labor Day. Obama is now over 50% in Colorado

http://twitter.com/ppppolls

[Esse Quam Videri]

1. Woot

[femmocrat]

2. Excellent!

[kag]

3. Yeah for us!

I really hope that this bodes well for all of the down-ballot candidates in Colorado, too!

[RockyMtnGuy]

4. Me too.. Coffman's gotta go!

[Esse Quam Videri]

5. Was just thinking that before I got down to

your post. Coffman's my rep now. Due to redistricting I think I get Gardner unless Shaffer can pull out the win.

[woolldog]

6. PPPs polls always seem too good to be true

What's their record for accuracy in the last few elections? Anyone know?

[aaaaaa5a]

7. Here is some very un-scientific data.

I took every battleground state that was polled this week. I then pulled the state's result from 2008. I then compared that result to PPP's final poll for the state just prior to the 2008 election. The results are below.


Battle-ground State-Obama 2008 Margin of Victory/PPP-Final Obama poll margin just prior to 2008 election


Florida-Obama +2.8/PPP-Obama +2............................. Obama -0.8

Ohio-Obama+4.6/PPP-Obama +2.................................. Obama -2.6

Virginia-Obama+6.3/PPP-Obama+6................................. Obama -0.3

Colorado-Obama+9.0/PPP-Obama+10...............................Obama +1

North Carolina-Obama+0.3/PPP-Obama+1............................. Obama +0.7

Nevada-Obama+12.5/PPP-Obama+4...................................... Obama -7.5

New Hampshire-Obama+9.6/PPP-NO POLL FROM 2008

Wisconisn-Obama+13.9/PPP-NO POLL FROM 2008

Iowa-Obama+9.5/PPP-NO POLL FROM 2008


For the 6 battleground states, PPP on average under valued Obama's performance in 4 of the 6 races. This includes Nevada where they undervalued his performance by nearly 8 points! The only two times they overvalued Obama's performance, they did so by margins of just +0.7 and + 1.0 points.


Not scientific I know. (Hey I'm not Nate Silver). But in looking at these results, which were not cherry picked at all and give us a very good sample, it looks like PPP is a very accurate pollster. My bet would be if you did this simple test with any other "gold standard" pollster, the results would not be any better. And in many cases they would be worse. PPP is regarded as a Democratic pollster. But to call them the (R)asmussen of the left is not true. As they point out in their Twitter account tonight, FOX news gave Obama better battleground numbers in some states this week than they have.

[WI_DEM]

9. Great, thanks for the stats! PPP has an undeserved reputation as a 'dem' pollster...

they don't cook their polls the way Ras does.

[TroyD]

12. True, they're a decent pollster

But Nat Silver says they have shown a bit of an in-house bias towards the Dems earlier this year, but they seem to be pretty accurate lately. Sometimes they've been less generous lately to the Dems than other pollsters (eg. they show smaller leads for Obama & Baldwin in Wisconsin than Marquette does). So they don't seem to be inflating Dem support the way Rasmussen does for Republicans.

[woolldog]

10. This is great work aaaaaa

good stuff, thanks for that research. makes me feel even better about the numbers.

[aaaaaa5a]

16. Thanks! nt

[WI_DEM]

8. Actually many of their polls this year have not been all that great if you look back.

[woolldog]

11. maybe ive repressed my memory of the bad ones

[aaaaaa5a]

13. Party ID breakdown of poll:

D 34
R 34
I 32

http://www.publicpolicypolling.com/pdf/2011/PPP_Release_CO_9231.pdf

[TroyD]

14. So it looks like they can't be accused of oversampling Dems

And Colorado was actually one of the few good swing states for the Dems in 2010.

The Democrats won the Governorship and the Senate seat.

So I don't see why it wouldn't be winnable in 2012.

[aaaaaa5a]

15. No, PPP didn't over sample Dems. In fact they might have been under-sampled in this poll.

Its funny because today ARG released a Florida poll showing Obama up 5. This is one point HIGHER than what PPP reported Sunday.


I've noticed that on more than one occasion, PPP has been AHEAD of the trend lines. Often they are sighted for bias early in the week, only to have other pollsters produce the same findings later in the week proving PPP's analysis correct.