Their models don’t go to zero.

There’s always a lot of discussion about the polls. Check that, usually most of the discussion is about the polls. Unless you’re watching the PBS Newshour (I still think of it as the MacNeil-Lehrer Newshour), political coverage seems to be all about the horse race.

The public tends to look at the polls as foreshadowing, as if the polls are telling us what will happen. But two months or six-weeks out from election day is an eternity. At this point, the polls are actually more of an indication of how the candidates are winning or losing each news cycle. For example, Hillary has a great DNC while Trump gets in a fight with the Khans and Hillary takes a big lead. Alternatively, Trump doesn’t tweet anything provocative (or ridiculous) for a few days and Hillary faints on camera. Guess what? Her numbers, well, come tumbling downward.

I always wonder what the polls are not able to reveal. This year, I wonder is it going to be like 2012 when Obama had a narrow lead over Romney in the polls but ended-up winning easily or is it going to be more like 2007–2008 (when the housing market collapsed and the seemingly impossible actually became highly probable in retrospect.) In other words, could the seemingly improbable Trump victory actually happen?

Returning to the housing collapse for a moment, Wall Street models of risk in the housing and the derivatives markets did not ever anticipate a time that so many Americans, all at once, would stop paying their mortgages. Mortgages in America were thought to be a sure thing. To some degree, at this point, I’m sensing a similar high degree of certainty in the political markets that Hillary will win. It’s as if much of the public and the pros (like Wall Street prior to 2007–08) can’t imagine Hillary will lose, so the feeling is it can’t happen.

Last week, on ‘’All in with Chris Hayes,’’ the Democratic pollster Doug Shoen said very confidently that given Trump’s ‘’unqualified to be President’’ number is over 60%, Schoen doesn’t see how Trump can win. Schoen said that unless Trump’s number comes down to below 50%, Hillary winning is basically a certainty.

Similar to Schoen’s comment, the Democratic pollster Peter Hart said on David Axelrod’s ‘’Axe Files’’ podcast, that Americans are certain they don’t want Trump. Hart said Trump’s likeability and competence numbers have never reached positive territory. He says Americans have already decided they know they don’t want Trump. He said there isn’t going to be a ‘’Bradley Effect’’ type of situation. In all their polling, Hart said people haven’t been holding back. They either really like or hate the candidates. He doesn’t see a large number of people in the shadows. He doesn’t see any big surprises on election day. He thinks Hillary will win.

The ‘’bed wetting’’ episode of the ‘’Keepin it 1600'’ podcast episode on Friday with David Plouffe was another instance where I was hearing a lot of confidence from the pros. Plouffe talked a lot about Hillary’s supposed built-in advantages in certain swing states, the professionalism of her campaign and the significance of the field effort that’s been ongoing. He mentioned that at about this point in 2012 Obama only had a slim lead in the polls.

But I kept thinking, these guys don’t think their models can go to zero. They’re all assuming Americans will keep paying their mortgages.

I don’t care what the polls are showing in Pennsylvania. I don’t care that Hillary is competitive in Georgia and North Carolina. Count me as a very, very nervous Hillary voter.