Early Predictions are Dangerous But I'm Sticking With Mine for Now

I don't know if I've even mentioned that early prediction here on KN@PPSTER yet, but here it is (I've been making it for a couple of months in various venues):

In November, assuming he's the GOP nominee (I still think a successful national delegate rebellion is just barely possible), Donald Trump will win every state that Mitt Romney won in 2012. He will also win Michigan, Ohio, Pennsylvania and Florida, he will at least be competitive in New York and New Jersey, and he may even put California in play.

Of course, it's not just early predictions that are dangerous, but any predictions involving Donald Trump. I didn't think he had a chance in hell of winning the Republican nomination. Every time he stumbled I figured he was going down for good. So if I'm making a mistake this time, it's the opposite mistake -- he's had a bad couple of weeks, but I expect him to turn things around.

The Hill's Jonathan Easley reports that even with Trump's slump, Clinton has failed to pull convincingly ahead in the ten "battleground" states where the election will likely (per the conventional wisdom, not per my prediction) be decided. Interesting. Even more interesting, though not fully explored:

Trump and Clinton both have historically high unfavorable ratings, opening the door for two outsider candidates -- Libertarian Party nominee Gary Johnson and Green Party candidate Jill Stein -- to potentially play spoiler.

When Johnson and Stein are considered in polls, the margin between Clinton and Trump almost invariably narrows.

See what's hinted at, but not actually said, there? As Matt Welch points out at Reason, when Johnson and Stein are included in polls, their inclusion hurts Clinton more than it hurts Trump. IIRC, even some polling that included only Johnson and not Stein shows Johnson hurting Clinton more than he hurts Trump. If there is a "spoiler effect" (yes, I hate the whole flawed concept, but that's how it gets reported), that effect seems to be pro-Trump.

I will be interested to see whether Johnson and/or Stein get enough votes to arguably affect the outcomes in any states. To the extent that they might do so, I expect that that effect would militate toward an outcome like the one I predict.