Last edited Wed Jun 29, 2016, 04:40 PM - Edit history (1)
***I'm sorry, still learning this system. I tried to attribute this reply directly to the first comment (I think), claiming that Nate Silver is an incompetent pollster. Instead my reply seems to have landed wayyyy down here.*****
Really? Based on what? I know you stated "IMHO" but, really? Silver ain't no hack, brother.
Fivethirtyeight isn't a pollster, at all. The true genius, aside from Nate and his team, lies in their commitment to a distinctly scientific approach specifically harnessed to break through all the noise and hype (and trolling) by distilling the qualified data down to meaningful insight.
Understanding their process and methodology should easily flip even the most skeptical naysayer, pronto. The analyst site is unique in the actual process they've created to arrive at their highly accurate forecasts. Based on your flip, uninformed comment, I can only assume you haven't spent much time researching the organization.
The 538 team almost always includes detailed explanations of how they arrived at the predicted outcome. The thinking process and the way they process data includes a very important, self-correcting mechanism.
Nate Silver explains;
"That means observing the world, formulating hypotheses about it, and making those hypotheses falsifiable. (Falsifiability is one of the big reasons we make predictions. When those hypotheses fail, you should re-evaluate the evidence before moving on to the next subject. The distinguishing feature of the scientific method is not that it always gets the answer right, but that it fails forward by learning from its mistakes."
They distinctly work with a fully vetted and qualified set of data sources (garbage in, garbage out). For instance, they create a weighted list of qualified pollsters. This means any given pollster is assigned a value based on historical performance results. While this isn't THE sole correlated variable, it is certainly a critical factor. Naturally, Nate's formula isn't public but he's rather forthcoming in sharing much of the thinking involved.
Your assertion of "incompetence" is likely one based on an improvised "eyes wide shut" approach at best, and simply one of selective ignorance, at worst. As far as I know, the current election cycle's entire primary outcomes were predicted accurately by 538, with only one anomaly—Michigan.
Nate's initial success emerged during a midterm contest, Pre-Obama, involving considerable attention after he correctly picked the winner, almost exclusively. No one else got it right. He followed the pollsters and evaluated the final analysis based on his proprietary formula. The 538 organization is admirably transparent, to the extent that their approach and outcome sync up predictably, the thinking is annotated and they thoroughly analyze any errors or glitches they might have experienced. So, as far as I know, no one has debunked his work or even come close.
The best thing is, his work is respected across all political boundaries and demarcations, except perhaps the occasional, disgruntled odd statistics analyst club.
I challenge you to come back with substantive facts that demonstrate and confirm fivethirtyeight.com's INCOMPETENCE. Certainly a fair request, especially in this volatile moment in this historically and bizarre election cycle. Nate's work is amazing and more than encouraging within the scope of what typically seems like completely arbitrary guess work.
As a matter of interest in trivia, which "pollster" are you willing to identify as being COMPETENT?
= new reply since forum marked as read
