Jay Cost's "Note on the Polls" is one of the most cogent explanations of the statistics involved in poll interpretation that this words-before-numbers person has yet seen.

Jay notes the unusually large variations in the assorted presidential polls, and notes that it's attributable to the different samples the pollsters are using.

The upshot is this: Obama is ahead, but no one -- no one --  really knows by how much.  Ultimately, the only way to guarantee that it's by a lot is for Republicans to give up and give in.  And wouldn't that be a shame, if it turned out that the margin is much, much smaller than most pollsters are forecasting?