A completely fascinating election year polling comparison is up at Pollster, written by Charles Franklin.
The two maps are dynamically generated and track relative average national poll position, and undecided rate.
Undecided rate in Presidential Election Polling, '08(Blue) '04(Pink) '00(Grey)
Presidential Election Polling average, '08(Blue) '04(Pink) '00(Grey)
They say much much more about the race than any talking head could.
Undecideds?
First, there is some indication that people WERE more undecided, but now the population is snapping back to norm in this regard. There is much truth to the issue of Democratic unity, but it appears that the message is largely sold, and I expect it to narrow sharply as people divide on the character attacks and economic messaging.
10% is still quite high, but is reasonable for a non-incumbent year.
What kind of race is this?
Secondly, this seems to be unlike either Kerry or Gore, but an odd amalgam of the two. Its clear that, in general, Obama has cut far deeper, through the "fat" of the electorate. Whether this is new registration, a gain of many Reagan Dems, or a lack of cohesiveness in the core social and economic conservatives, is unclear at this point.
What is clear, is that Obama has significantly wider support than either of his predecessors ever did. As to the depth, that is what will be tested in the last 30 days.
One point about the structure and fundamentals of the race: the Palin/RNC swing was half the size of the economic swing to date. Its unclear whether that movement is complete. I'd be surprised if Obama moves much deeper, since he's already in uncharted territory. Any remaining movement will need to come from new engaged voters and this, of course is possible, given the shock that the country has been through, but unlikely in my opinion.
How much have polls moved late?
You should notice that the last 4 weeks of the last 2 elections had relatively stable polling, within 2 or so points. Minds were mostly set at that point (although Gore was underpolled by 3%, likely to be last minute deciders.)
The Swift boat attacks in 04 happened earlier, and public opinion turned against Kerry at a far more pedestrian pace than the current Financial realignment of the race. It seems very late to be introducing this line of attack now. I wonder about the traction of Ayers et al, we will see given the stability of the electorate in these prior Octobers.
Past performance might suggest that the die may already be cast, although, of course, this is no time to slack off. I'm volunteering, and have no plans to get less involved in this process.
On that note. If you're not a volunteer, go volunteer. Find out how you can help, while fitting the work into your schedule. We need thousands of volunteers for the final stretch in Colorado. We need to turn the map blue, not only for Obama, but for those crucial down ticket races, that the effort also benefits.