Poll dance

I'm hoping the three-degree doctor will weigh in on this issue soon, since it's his area of expertise. But it's pretty clear the wildly different poll results we're seeing in the presidential race are the result of wildly different methodologies - some of which seem hard to justify. Here's the gist of it, from the Left Coaster (via Atrios, naturally).

My contribution here is just to add to the general paranoia by suggesting a possible motivation for these screwy methodologies (beyond the we're-desperate-to-make-Bush-look-good rationale). To wit: The more confusion surrounding actual voter preference before the election, the harder it will be to notice, or prove, vote fraud. Get it?

Scared yet?