Just to clarify: regardless of the outcome, you can't really say that Nate Silver is wrong. Right now he is saying that based on the data and simulations he has run there is an 80% chance that Obama wins. If Romney wins it does not prove that Silver was wrong. I would certainly take a 1 in 5 chance of winning PowerBall. In essence, Silver does the same thing that WIS does when they run a WS, Final 4, Super Bowl "prediction." It's just based on their models and simulations.
Additional point regarding polling...if a poll says Ohio is 50-47 for Obama with a 3% margin of error, that does not mean that it is virtually tied. It means the result with 95% certainty (unless stated otherwise) is anywhere from 53-44 Obama to 50-47 Romney. All of the people nattering on about MOE and it being a virtual tie, are only applying the MOE in one direction...not realizing it could go the other way as well. Anyway, we will all know the result soon enough...hopefully not like 2000.