Other polls reflect this last-day surge for Obama.

Keep in mind, most polls actually exclude young, low income, ethnic minorities and people who have never voted before- typically filed as not 'Likely Voters'. Most research says that these people are breaking massively for Obama, and intend to vote in droves. Furthermore, polls rarely reflect those without land-line phones (IE cell phone-only users). These typically are young or low income individuals (who have to chose one or the other, but not both), and tech-savvy educated younger professionals- all demographics which will break massively for Obama.

So while the 'traditional voter' will go Obama by as much as 7-10, I suspect there is an unreflected 'surge' of new and non-traditional voters that are not reflected in polls that will break to Obama by as much as 15-20 or more. If these voters vote party line, it could really impact the House, Senate and Local races.

As far as I know, only a few presidential national polls have put out the effort to try to tabulate the change in demographics and voters- and those that do show advantages to Obama by as much as 13 points, wheras the more traditional polls and polling methods show 7-10 points for Obama. Extend these results to tight senate races, we could see some serious upsets... Unlikely, but still possible. This election may surprise people like some in the past- where demographic changes invalidated polling methodologies (Example? Reagan/Carter!)

Yeti