I understand why Romney supporters (and Romney) want to claim the polls are wrong. In their position, I'd probably be trying to make that case, though I'm not sure I'd actually believe it. What I don't think I'd try to claim — and what I'm pretty sure is not the case today — is that the polls were deliberately biased. In this piece, Ezra Klein points out how silly it is to think that the professional pollsters, whose careers hinge on delivering a reasonably-accurate forecast, would all get together and conspire to not report what America is telling them. I could maybe buy incompetence…or a theory that this particular election for some reason defies traditional polling methods. But not intentional fraud.
Oddly enough, the loudest folks out there insisting that Romney could not possibly be that far behind also seem to be the ones who have never really wanted him as their candidate…and I'd be very surprised if even one of them didn't think Romney's been running a very bad campaign. When they say, "He couldn't be doing that poorly," what they really mean is, "Obama couldn't be doing that well."
By the by: I keep seeing articles about the upcoming presidential debates that note how the tune-in for such contests has declined so since the days when it was Kennedy vs. Nixon. The comparisons seem pointless to me. When JFK went up against Nixon, it was a time when if you didn't see it live, you probably didn't see it at all. You couldn't record those debates for later viewing or go see replays online or even substantially on later newscasts. You got one shot at it. I probably won't watch Wednesday's Obama/Romney square dance live and might not even bother TiVoing it. It'll be on a half-dozen online sources and the highlights, assuming there are any, will be on a thousand sites. If it were a one-shot deal, I'd rearrange my life to be in front of the set. But it isn't and nothing on TV will ever be that again.