Pollster.com

boombatty will have an orgasm with this one. Outlier. Fail.

btw, how's Intrade lookin' today?

http://www.intrade.com/?request_operation=main&request_type=action&checkHomePage=true

You Betcha!

Looking at the poll's stated internals, it indicates that Obama is leading women by 9 points and losing men by 4 -- yet he is only 3 points ahead.

I'm not sure how this can be the case, given that women comprise a slight majority of total turnout.

In any event, the CBS News poll showing a tightening as well -- Obama has a 5-point difference.

However, on the other hand, Rasmussen still shows 8-points for Obama.

Perhaps between Palin's good debate performance and the increased scrutiny on Obama/Ayers connections are having slight payoffs?

I'm not sure this would be enough for McCain though.

both zogby and hotline undervalue the number of dems to republicans signficantly...zogbys numbers suggest the number of republicans and democrats is equal, while hotline has only a 2 point difference. its rather unlikely that there was a dramaic swift in the number of people identifying themselves as republicans over the weekend.

@Whitetower
I am not sure if the gender numbers are wrong, if the electorate was split 50/50 then Obama's lead would be the exact half way point between +9 and -4, which is +2.5 roughly where the poll is. However the party-id margin, (which according to zogby's methodology he does weigh for), appears to be at +1D, assuming around about 25% independent vote, (which Obama is winning by 7 points according to the poll). Now if the sample had happened to turn up a +1D margin then fair enoungh, but I would like to know what Zogby's reasoning for weighing for it?

Didn't Zogby also predict Bush would lose both times?

Also, in general, the only tightening is shown in the national polls, states are still swinging towards Obama by a large margin, look at the map.

The differnce in the polls is not so much on the McCain side which seems to come in consistently at 44-56%. But on the Obama side this may be an issue of hard the various pollsters are pushing for leaners and may indicate that Obama still has a little work in firming up his support.
This might be confirmed by Rasmussen showing in most states that the % of supporters who might change their mind is a bit higher for Obama tha McCain.
Still almost all polls show the trend in Obama's direction.

Yes Zogby predicted Kerry in a close election, and Gore in a landslide. lol

Of course Gore won so Zogby was right there.

It's the Palin Bounce from the Debate

note previous post should have been McCain 44-46% not 56% sorry.

Remember... Gore won the popular vote which correlates with national polls with 48.4% vote.

The year 2000-- In an Oct. 3-5 Zogby poll respondents favored Al Gore over George Bush in a head-to-head match-up by 5% (46% to 41%).

---------

NOT an "outlier." Quit abusing that statistical term to dismiss legitimate polls.

Absolutely no point in trying to dissect the internals of individual polls, unless they show something seriously enough wrong in their methodology to discard them altogether.

They all have different models, therefore house "leans", and reliability. Best to produce weighted averages of the day's grab when we have them all. Particularly since we have such a large number each day. Yesterday was +7 for Obama.

The trend continues...MACK IS ON THE RISE! YOU GO BOY!!

SARAH PALIN, ANYONE?

I agree this poll is not an outlier, it's internals are similiar to many other polls. Actually it's internals are very close to rasmussen (albeit rasmussen has slightly fewer Other/undecided voters). The difference, of course, is the party-id weighting.

I've been researching 2004 accuracy. Most of the major polls got the winner correct.

Final Presidential Polls:
2 tie
5 Kerry
12 Bush

Out of those 12, 7 under estimated Bush's win. 5 over-estimated.

The most accurate: +/- 1%
CBS / NY Times
CNN / USA TODAY / GALLUP
PEW RESEARCH
TIPP