Oh, wait. Check this out: WASHINGTON (Reuters) - Democratic Sen. John Kerry moved into a one-point lead over President Bush three days before the presidential election, according to a Reuters/Zogby poll released on Saturday. Kerry led Bush 47-46 percent, well within the margin of error, in the latest three-day national tracking poll. Bush and Kerry were tied at 47 percent on Friday. http://www.democraticunderground.co...sg&forum=132&topic_id=1228694&mesg_id=1228694 Gee I wonder how many young people took time out ON HALLOWEEN to chat with the folks at Zogby International?
You aren't accounting for the November Surprise the democrats pulled. Cheap shot. 14 year old Bush DUI was made public, and that changed the polls.
I'm just relaying what I read. Shouldn't you lob an ad hominem attack at me? It seems to be your specialty.
We're winning the election fellas, we're about to appoint a true patriot to the white house. No more bickering and pointing fingers in the capital.
Yeah, the exit polls had Kerry winning at 3pm PST in 2004. It's over. Don't worry about voting, Minstrel!
Exit polls projecting a narrow victory and pre-election / exit polls showing a massive lead are rather different things. It does seem over, but I'll still vote. Enthusiasm to vote among Democrats is still very high. A report I saw on the evening news tonight said that turnout in this state (California) projects to be one of the highest ever, on a percentage basis. And that's in a state that isn't even in play.
What's the surprise? That on Nov 5th, you and CK and a few others will be posting non-stop about how the election was stolen, and how Americans are sheep, and how President-elect Obama is going cause various disasters? That won't be much of a surprise, I'm afraid. barfo
What was reported in election day exit polls in 2004 varied depending on which source you checked. But it ranged from narrow Kerry win to narrow Bush win. And the polls going into the election all had Bush ahead by a couple of points.
I wonder what country they will move to? They certainly won't stay here and support the Socialist States of America!
Minstrel, really there's no reason to study all those detailed polling numbers when Matt Drudge is picking out the important one for us. The other numbers are extraneous. It's better to let Drudge do the heavy thinking. After all, he's got a website. barfo
Just a note on polling. During the primaries, Barack overperformed his pre-election polls, while he also underperformed his early exit poll numbers (the numbers you see as the polls close...the one's that remain on sites like CNN that you can look at today are weighted to be representative of the electorate). I think this is the Bradley effect. With the ushering in of the internet, people have no problem telling the truth in something like a telephone survey. But when you are seeing a pollster face to face, which they do in exit polling, they are more apt to lie.
Drudge cuts straight to the heart of the matter, slicing away all the bias. You know, I read a recent report that illustrates polls are biased to the Democrats. This was proven by the shocking statistic that virtually all of them paint Obama in a better light than McCain, approximately 6% more positive on Obama than McCain, taken as a whole. Since we know that "no bias" means no difference in portrayal, this is pretty iron clad proof. Drudge understands this, and isolated the only fair and balanced poll, the one that shows them basically equal.
I'm quite unconvinced of this. I'm sure plenty of people would lie about why they didn't vote for Obama, if it was racism. But they have no reason to lie about who they voted for. There are any number of justifications they can use: he's too inexperienced, he's too liberal, I fear his terrorist buddies, etc. I think it is more likely Obama underperforming his exit polling was statistical noise. One race isn't sufficient to show such an effect. 538.com looked at several recent elections involving a black candidate and didn't find compelling evidence for a Bradley Effect.
That was their look at the pre-election polling. There is no Bradley effect in the pre-election polling, just the Obama effect, which is blacks lying that they are not supporting Obama, perhaps just to mess with the pollsters. I looked at the early exit polls (from when the polls closed) for every primary / caucus (when they had them, they didn't always have them), and time and time again, they overestimated Obama's support. Lets just say, Obama would have had a lot of big state popular vote wins that would have forced Hillary Clinton to get out earlier if the exit polls were not suffering from a Bradley effect. I think the Bradley Effect will be very real in some of the exit polling. There are people who will not want to vote for Obama because he's black deep down, but who know saying things like, "he's too liberal" and "not experienced" could be perceived as code words for racist motives, so they will just avoid that altogether by saying they voted for the guy. Like in telephone surveys, only about 20-25% of the people they call actually respond in telephone surveys. These are people who don't mind sharing. Obviously a lot of people want to avoid telling them with that high of a non-respondent rate. But people will feel compelled when an exit poller approaches them, to talk to them...and they will lie to look socially acceptable. Of course nothing is certain, we will see on election day. My guess is that Obama will overpoll the exit polls in Ohio, poll even in Indiana, and will probably underperform in the exit poll in North Carolina/Virginia. So an exit poll that shows him winning by 2 points in Ohio probably means he lost, while an exit poll in North Carolina/Virginia showing an even race probably means he won (especially considering the early voting returns being so heavily in his favor). But we obviously haven't been in this position, electing a black guy for president, so this is all uncharted waters. We have the primaries to look at, but that is a completely different demographic than a general election electorate.
The problem with this reasoning is why would it be limited to actual racists? If "he's inexperienced" and "he's too liberal" are seen as code words for racism, wouldn't people who are voting for McCain for non-racist reasons also fear saying those things, in case they are unfairly considered racist? We should then expect virtually every exit poll to look like a landslide, as many McCain voters claim they voted for Obama just to avoid seeming like a racist. It wouldn't just be limited to actual racists. The reality, in my opinion, is that many people will legitimately be voting for McCain and saying things like that, and that will give racists a safe way to go. I agree with this. I am not saying there is definitely not going to be a Bradley effect. I'm just not convinced that there will be. It seems very likely to me that all people who don't want Obama elected, racists and non-racists, will be reporting that they voted for McCain. They just might hide their reasons why with a lie.
