A series of posts on Nate Silver's blog about elections and polling examined Herman Cain's chances of winning the GOP nomination. Silver criticized pundits who dismiss Cain's prospects, arguing that historically, there have been few candidates analogous to Cain, who is doing well in public opinion polls but lacks the support from the GOP establishment that is generally considered necessary for victory. According to Silver, this lack of historical precedent undermines the ability of pundits to predict Cain's chances with any certainty. He argues that pundits generally fail to adequately acknowledge this uncertainty when they make predictions.
While Silver's argument is technically correct - the chances of Cain winning the nomination are not actually zero - Bernstein makes a persuasive case that the numerous obstacles in the way of a Cain nomination allow us to be fairly confident in predicting that he will not win. However, I'm not sure Bernstein responds effectively to Silver's argument about the unpredictable combination of strong public support and limited establishment support - Bernstein's explanation of why we can assume that old rules do apply didn't really seem adequate to me. It will be interesting to see, over the next few months, whether public support will be sufficient to sustain Cain's campaign.
Note (11/13) - I meant to post this 2 weeks ago when these articles were written, and found today that I left it as a draft. The accusations about Cain that have emerged since these articles were written would seem to reinforce the claim that he has little chance of winning the nomination.