Gotta love treating best fit lines based on 12 observations as predictive.
Seriously, it's like he doesn't understand math. Dukakis and Clinton are not in the relevant range. Primaries have changed since then. Gore and Bush are virtually irrelevant as their primary season operated entirely differently. They were both inevitable. On top of that, he's taking Trump's favorability on a down swing almost intentionally. It was fine a week ago. Like many Atlas posters, he can't understand this is not a static value?
Also, might be just as useful to run a regression on favorables of unsuccessful candidates. Hint: They ain't what they used to be. Favorabilities are way down. Carson is a rare exception. Does Silver want to say CARSON is the nominee?
I'm not saying Trump will be the nominee. In fact, I just sold almost all my shares today as I was happy with my profit for now. I am saying Nate Silver should be required to give back any college degree he ever got because the CLOWN has NO CLUE. This would never be published by a serious person. It pains me how stupid he is. No concept of time or human feelings. Tried to compare it to just 2012 with Romney and fads too, but likewise, that failed miserably because it is a completely different primary season.