leftard Silver got it wrong on the last 2 elections so I’ll just take this with a grain of salt. Did he have a dartboard this time to get that number?
He’s just not looking at the relevant data, and his models are over weighting less relevant data.
For example, his models heavily weight endorsements, because in most elections endorsements matter. Naturally, Hillary being an establishment politician with nearly 50 years tied at some level to the democratic machinery has lots more endorsements and by far more politically connected folks that a disruptive candidate like Donald Trump who just entered politics have.. etc...
As Such Nate’s models tell him, Trump should lose, even though he is clearly winning right now based on polling... Because his models can’t handle a disruptive candidate, they can’t see a disruptive candidate is better off NOT getting an endorsement from certain establishment politicians that he would be if he did.
Nate’s models aren’t able to grasp the relevant data this election, because they either don’t know they should be looking at it (ie the relevant data isn’t something that was relevant in the past) or they don’t know how to interpret it properly, (such as my endorsement example).
When a Kasich or a Bush or a Cruz doesn’t endorse or verbally belittles or refuses to come to the convention for Trump... those things appear historically to be WEAKENING to a candidate traditionally... but for a disruptive candidate with an anti-establishment/anti-status quo electorate, those are all huge positives... Nate’s models just won’t pick it up until its too late.
Trump will continue to lead in the polls and Nate’s models will keep telling him Hillary should win... that will up until an inflection point in his models, where historically polls now outweigh all those other factors... which happens as you get closer to election day.. then “SUDDENLY” Nate’s Model will inflect because all those other data points become less of a prediction than the actual polling.
We are in JULY so even though polling shows at every level Hillary is desperately in trouble with no obvious path to victory, she has the right things ticked off traditionally that “should” give her the edge... closer to election day, patterns weight toward polling strips away those other factors...and you hit that election point... July is too far out to hit that inflection point.... Nate’s model is missing and misinterpreting relevant data.