With all due respect to Nate Silver, who I admire tremendously, there is some tightening in this race. The fact that three to four of the daily trackers have gotten closer is an indisputable fact. Whether their weighting is accurate or not, doesn't matter that much. R2K vs. Zogby- their partisan weighting are polar opposites, but both show a race growing tighter.
Should we be concerned? Yes. Should we worry? No.
The fact is, outside of IBD, all trackers show Obama maintaining a steady lead of 50% or above. Obama's lead has been more a result of McCain's poor performance with the inability to be over 43% in most polls over the last couple of weeks. McCain's ceiling since the crisis has been about 46%. That three to four percent delta (from 42%-46%) represents his 'soft' support - no undecideds. Since the stabilization of the race after the economic crisis, I highly doubt there has been much swapping of support between McCain and Obama. In fact, as Rasmussen points out, Obama's lead has remained above 50% for 33 straight days. No, the daily fluctuations are more of the 'soft' support variety of each candidate. These are the ones who will have made their decision, but not very comfortable with it and are swayed back to undecided when they aren't happy.
Both candidates have been static within 2-3% of their current number for hte last month. That's why you can see fluctuations in the polls up to 5-6%. When Obama is at 52% and McCain is at 42% - that's 10%. But when Obama, within his range, drops to 50% and McCain increases to 44% - the lead drops to 6%. Until either candidate breaks their barriers (Obama falls below 50% or McCan breaks 46% in several polls), there is no 'real tightening' of the popular vote. With regard to the independents, there is no break for either candidate. They both continue to swap in and out their 'soft' support to the undecided column. As a result, McCain appears to be getting the break on undecideds as his support creeps up. When McCain breaks 46%, then you can declare the independents are breaking for McCain. If Obama drops below 50% and McCain continues to rise, then you might be able to claim a shifting of the electorate.
Beyond all that, Obama has gotten some breaks. Whether they were lucky or well designed is anyone's guess. Fact - when Obama lost his lead during the RNC and Palin mania, he needed a big boost. He got it in the economic crisis. It brought the election back into focus. Fact - when Obama saw the race tightening a few weeks ago, he won the second debate and got another bounce. Fact - a week ago after the third debate, the polls were tightening. That Sunday Sec. Powell endorsed Obama and he got another bounce. Obama continues to get the pep he needs when things close in. On Wednesday he will deliver his 1/2 hour program. With any luck, this will be another bounce needed when the polls are tightening. That is a bounce that will probably take him to election day.
**Update: David Gergen just made the EXACT point (on AC360 - CNN) - down to the examples I used - to the paragraph above. So it's not my crazy observation**
I think it's important not to dilute ourselves. We think we are very smart about talking about not getting complacent and getting to the polls early. We also tend to find the flaws in nearly every poll that shows tightening. It tends to be a bit of 'group think'. As Obama supporters, we need to recognize the trends early across polls. Rather than find the flaws in the numbers (voter ID wrong, too much regional sampling, undercounting cell phone/new voters), we need to understand what they are really saying - particularly the consistency across polls (levels of support for each candidate, etc.)
A 4% lead, with Obama at 50% right now is great! a 75% break of undecideds would still mean a 51-49% win for Obama in a two person race. We tend to forget that there are two other candidates in this race, since most polls report only on the two. Barr and Nader, per Real Clear Politics, account for 5% of the vote. When they are factored into the RCP average, Obama's lead opens to 9 points. Barr will definitely pull votes from McCain (as will Ron Paul) and Nader from Obama. But the RCP net is a -1% for McCain. Factor that into my scenario and Obama will win 49% to 46%, with 5% to Barr/Nadar. 3% is a tremendous win.