Here's a fun little back-of-the-envelope analysis of Romney's standing in the polls. As you may have heard, some people are having a hard time believing the polls lately. So let's give Romney the benefit the entire margin of error of every poll released over the last ten days. This is how it works:
Rasmussen's latest poll had Romney at 47% and a margin of error of 3.* So if we gave him the entire margin of error, he is at 50%. That means, according to Rasmussen, there is only a 2.5% chance that Romney's support is higher than 50%. So I did this with all the polls I could find that were released in the last 10 days:
POLLSTER ROMNEY MOE
Rassumsen 47% +3
Gallup 46% +2
AP/GfK 46% +4.3
NBC/WSJ 45% +3.6
Monmouth 45% +2.7
Fox News 43% +3
Reuters/Ipsos 43% +2.9
The average of Romney's standing in these polls with the full standard of error in his favor is 47.7. The lowest percentage for a winning candidate without a strong third party challenger was George W. Bush in 2000, with ... 47.9%. So even with the entire Margin of Error, Romney would do worse today than the worst winning candidate in history.
So even if the pools were all systematically skewed towards Obama ... and Romney had a once in a century electoral college map in his favor ... and Ralph Nader re-emerged to steal a slice of the Democrat vote ... and the Supreme Court intervened to give him the election ... he might have a chance. Otherwise, he's fucked.
* As I understand, most polls have a 95% confidence interval