You can check out gallup or rasmussen or any number of different pollsters on their websites. I like Realclearpolitics the best because it's a non-partisan site that just gives you a compendium of pollsters, the "poll-of-polls" numbers, and they throw up a great variety of columns by writers from all over the spectrum.

The useful pollsters tell you about their sampling data. Rasmussen is conservative in their estimates because they take a more cautious sample of voters. For instance, the number of registered Dems in the country is like 39%, to the Republicans 29%, with Indie voters around 30%. Some pollsters go with that average in their samples, but Rasmussen and Survey USA tend to have about a 5 or 6 point spread between Dems and Repubs instead of something like 9 or 10 points that many of the others are using.

This is why you see a little more of a conservative and cautious result with the FOX poll and with Rasmussen, who have Obama up by 7%, and why Newsweek and Gallup are at 11%.

If I had to make any predictions, I'd have to say that Rasmussen is probably right around where the truth is.

What's been clear is that other than a two week stretch where Obama lost his lead and then things got tied up, he's been gaining about a point a week or more on average since then in all the polls.

We've seen dramatic change in places like Pennsylvania and Michigan, which were air tight for so long. We saw a complete sea change in Florida, where Obama was down by an average of 5 for the longest time, and then we saw an 8 to 10 point swing in just two and a half weeks or so.

Watch out for West Virginia now. Obama was behind by 7 or 8 the whole campaign, and the latest polls have shown him ahead by 2 and then by 8.

In the RCP poll-of-polls just before the election in 2004, Bush lead by 1.5%. He took it by nearly 3%, so the polls were dead on. In Florida and Ohio, Bush was up on average by 2 in Florida, and he took it by 5, and Ohio was basically tied, with Bush taking it by 2. So the polls are actually right way more than they're wrong, and even though Ohio didn't go Kerry's way, it was within the margin of error.

One of the stories that has been neat to follow in the last few months are polling methodologies. Like, did you know that many high-profiled pollsters are trying to offset the Bradley effect by posing certain questions to people? They're taking into account a small percentage of people they think might be lying to pollsters when they slip in questions like, "Have you been to a dinner party where there's a black person?".

The polls tell us so many things, but they don't tell us other things as well. Like in Georgia, where the black vote represented 25% of the vote in 2004. This year, the pollsters have increased likely black turnout by just 2%. In over 400,000 early ballots cast already in that state, 39% is coming from blacks. Will the numbers stay that high on election day? Or are blacks just getting a leg up on voting now in order to avoid long lines? It's hard to say until we see all the results. But it's so fascinating.

Somebody should make an exciting movie about pollsters and pollstering.

__________________
“The America I know and love is not one in which my parents or my baby with Down syndrome will have to stand in front of Obama’s ‘death panel’ so his bureaucrats can decide, based on a subjective judgment of their ‘level of productivity in society,’ whether they are worthy of health care. Such a system is downright evil."-Sarah Palin, not having a clue once again about what she is talking about.