I repeat - Zogby has been horrible the last four years. It had Kerry winning on election day carrying Ohio, Virginia, and Colorado. He had Obama winning California, New Jersey and Massachusetts in the primary. Until he gets back on track - I would advise all partisans to ignore Zogby. His on line polls tend to demonstrate who's supporters are paying the most attention at a given time. I think the race is roughly tied, with McCain probably leading for a few days and then going back to the tie it was in most polls before the conventions begin.
probably first time in my life(born 1965) that MSM media bias liberal polls show a lead for a Repub...that says something... I mean, I think the polls had Mondale up over Reagan..
R Convention bounce. Give it a week and we will be back to even or something like Obama at 2% up. It will be a close electoral college race. Obama will have a lot of wasted votes in California and NY, so being up 2% is like being even. It really will come down to the 8 or so swing states, overall national numbers are useless. For some reason Obama is doing spectacularly well in Iowa, a state Bush won over Kerry. McCain is running stronger in some traditional D states than Bush did, but not enough to carry them. Tell me who wins Colorado, Virginia, Ohio, Nevada, New Mexico, Iowa, and New Hampshire and I will tell you who wins the election.
all i remember about the national polls is that in 200 and 2004 i was sure bush was going to lose on election day due to the polls. i hope this trend continues.......
"I have made a career of taking bungee jumps in my election calls. Sometimes I haven’t had a helmet and I have gotten a little scratched. But here is my jump for 2004: John Kerry will win the election." -- John Zogby, 2004 Zogby has odd polling methods and because of that, his results are all over the place. Sometimes he's right when everyone else is wrong. He built his name on those results. But more often, he's wrong when everyone else is right. Zogby can be accurate, but he is definitely the least consistently accurate among the pollsters. He also doesn't enjoy a good reputation among pollsters: The Link
The 2000 and 2004 elections were very tight both in the polls and in reality. People are attempting to rewrite history by saying that Kerry was dominating Bush in the polls, he wasn't. It was a close race in the polls and it was close in reality. This is all expectations-management, and we're all so used to seeing it from the right-wing ("Our candidate isn't as inept as you thought!") that it's just commonplace.
Bush had a small, but just barely significant, lead in a great majority of polls following the conventions leading up to the election. The Link Gore had a decent lead until the beginning of October, but Bush took just about every poll leading up to the election, though it trended back towards a dead heat right at the end. The Link mop, you're probably remembering the whiffs by the exit polling, especially in '04.
Gallup: McCain +3 Rasmussen: Even Both polls included 1 night before McCain's speech, and 2 nights after. Tomorrow's tracking polls will get the entire effect of the Pub convention "bounce."
i was referring to the day of the election....sorry to be confusing. was i wrong? it seems like the day of the election the impression given by the polls was that Bush was going to lose.
So if the one is showing McCain/Palin up 3, wouldn't that mean a +11 point jump, which is more than the 4-6 point jump Obama got?
Here's the situation: These pollsters know how peoeple of a particular profile will vote. They just don't know who is going to vote. If pollsters had a list of who would vote along with a complete demographic breakdown, they'd be phenomenally accurate. But they don't, and guessing that model isthe hard part. One of the biggest uncertainties is how the first time voter will act. This group heavily favors Obama. But many don't vote. The group is notorious not only for not voting but for saying they will when they won't. If a pollster knows that, for example, every election 80& of UT students say they will vote but 40% do, should the model guess that 40% will vote or are there other indicatioons that this election is different. Answer that question and a lot of fog clears.
