@nimh,
I am pinning my hopes on the undecideds, and the hope that the polls are mostly skewed. Judging from my personal experience, everyone I know lets their answering machines pick up, and then they delete all the pollsters and political pep talks delivered. So I am wondering how many people have to be called in order to get a statistical sample? Does the sampling even represent 50%, 30%, or 10% of all the people called, and of the people that don't answer the pollsters, what would their mindset tend to translate into - in terms of party affiliation or political preference? Would Obama voters tend to be more vocal, etc. etc. etc.?
So, I think the risk of polling being outright wrong, or at least skewed considerably, may be higher this election cycle? I could be totally wrong.
The other factor, election fraud, who really knows, and the lawyers are lining up. The election process has been really corrupted, and one of the big reasons is early voting. Early voting should be eliminated, except for legitimate absentee balloting. Regardless of party affiliation, it looks like we could all agree on conducting a good and tight election, honest, but sadly that does not seem to be the case even. No shame on the part of some people.