Talk Elections

New Poll: Maine President by Survey USA on 2008-09-24 (https://uselectionatlas.org/POLLS/PRESIDENT/2008/pollsa.php?action=indpoll&id=2320080923019)

Summary: D: 49%, R: 44%, I: 4%, U: 3%

Poll Source URL: Full Poll Details (http://www.surveyusa.com/client/PollReport.aspx?g=3ea1b43d-0776-48f7-b545-6515576915bc)

McCain 44%

Obama 49%

Other 4%

Undecided 3%

seems like a tad low for Obama

Why the heck does SUSA release stuff so late?

second straight poll like this, no?

Good news is both CD's seem to be tracking about the same, so a small Obama victory would probably still lead him to winning both the districts.

They classified it as "North" and "South" - and they were about the same.  I think 1% off, but both for Obama.

I'm still skeptical about it being this close.

BANGOR, Maine - A Maine TV news anchor who bears a resemblance to the Republican vice presidential nominee says she’s been getting “hate mail and nasty phone calls” from viewers who think she’s trying to copy Sarah Palin’s signature style. (http://www.bostonherald.com/news/national/northeast/view/2008_09_24_News_flash:_Maine_anchor_channels_Sarah_Palin/)

Among the 7% who said they have already voted, it's Obama by 53-41.

Huge MoE, not relevant.

Also probably liars

I said this about the SUSA poll that showed Obama +11 in IA.


It seems SUSA might be using a methodology that prizes...ah.. excitement.. over accuracy at this point in the game?  They never seem to produce a "blah" poll, everything is either showing WA/ME/whatever within the margin of error or swing states blasting away from competitiveness

They use an exciting methodology called a sample size of 675, the MOE is 4%. An Obama lead of 12% in ME and 4% in IA would be in their 95% confidence interval.

675 isn't that bad for a state poll.

I don't have anything concrete, but obviously more people will want to hire you if your polls create exciting news stories, and this far out it doesn't really matter what your poll says since you can't get proven wrong.

SUSA just never has a boring poll.  If you have to always go to the outer reaches of the MoE to get a reasonable result, it's suspicious.  It's not like the "true result" is equally likely to fall anywhere within the MoE, it's just a 95% confidence interval.

It's not just this poll, but also NC (M+20), VA (O+6), IA (O+11), WA (O+4), and a bunch of 2006 races I looked into where SUSA would consistently be 5% off every time.

To let their clients release the poll first on their 6 or 11PM newscasts.  SUSA's clients are usually TV stations.

A closer Maine makes more sense, logically, than a blown out Maine, but I still doubt that "closer" means four points.

Maine's not that red* a state by New England standards.  Kerry won it by 9 even though he's from the region.  Gore won it by 5.  McCain trailing by 4-10 points is certainly plausible.

*-Atlas red/real world blue.

That's a common misconception. Population does not matter in determining how many should be polled.

I saw a woman with a Palin updo on the T this morning. I can't imagine what that's like... either she had it before Palin became famous, in which case she's probably getting lots of weird stares, or she adopted it after seeing it on TV, which is much worse.

Republican 34% (27%): McCain 80% (76%); Obama 16% (17%)

Democrat 35% (40%): McCain 12% (16%); Obama 84% (75%)

Independent 28% (29%): McCain 44% (35%); Obama 49% (56%)

Conservative 27% (25%): McCain 81% (72%); Obama 15% (23%)

Moderate 40% (41%): McCain 39% (35%); Obama 54% (58%)

Liberal 15% (19%): McCain 7% (12%); Obama 90% (83%)

(denotes SUSA, Feb. 26-28, 2008)

 Quite a marked shift in party ID. A Democratic advantage of 13-points reduced to one

Dave