What's happening to Trump's popularity? Parsing the polls with Nate Silver

Originally published at: http://boingboing.net/2017/02/22/lies-damned-lies-and-statistic.html

I thought polls are just fake news now. Wait, is it the right or the left that says that? Urghh!!


Because Silver only gave Trump a 28.6% chance of winning, and yet he did win, so everyone knows that proves his prediction was completely wrong. P.S. we are adults who understand basic probability. :rolling_eyes:


I want to sit down at a poker table with anyone who says Nate Silver’s forecast was “wrong”.


Yeah, I actually thought he was the only one who got it right. Everyone else said it was Hillary by over 90%.


Only when you don’t like the outcome, of course. That’s how you can tell if the election is rigged, too.

@LearnedCoward: No no no he was only 71.4% wrong.

1 Like


Also interesting to note that the ‘how could Trump win’ pre-election scenario laid out was played almost to a T.


So the analysis of Trump’s popularity is “more light than heat”, but still warrants a lengthy breakdown? I’m confused. Do you agree with it or not?

Also, Silver fully admitted that 538 got it wrong, in the sense that their model still ascribed a better chance to Hillary winning. There are assumptions made with the polling data used to make predictions, and Silver was one of the few to honestly ask " What if our assumptions are wrong?"


If you play Russian Roulette you have an 83.3% of survival. Sign me up!

Put two bullets in the chamber. That’s actually a little closer to Silver’s odds.


If you looked at the polls-plus forecast, that reduced Clinton’s odds even more.

Silver’s model was more or less given that certain assumptions are true. If they’re not, well, that will most assuredly affect the accuracy of the model. It’s like estimating the probability that at least one coin flip in three will turn up tails, but not knowing that 1) two of the flips have already happened, and 2) they were both heads.


“Change” is what got Obama elected. “Make America Great Again” also has change as a central theme. When I look at it, the hollowing out of the middle class that Reagan began, and each subsequent president has perpetuated has set the table for this situation. Mao, Castro, Hitler, Mussolini, Stalin, et. al. all came to power not because of a radical minority, but because the average citizen was mobilized for change to correct the injustices of a hierarchical society where the ruling class stopped caring.


by doing land line only, they do more than exclude 50%. They exclude people with it enough to realize land lines are unnecessary. That pretty much leaves you with the average Fox viewer.


After you get the first half dozen calls during dinner, you are also inclined to slam down the phone on the next hundred pollsters. Either that, or deliberately juice the data.


P.S. we are adults who understand basic probability.

Yeah, but there’s only about 6 of you.


So we are talking about a 50-50 chance right?


I believe more light than heat is meant to be a good thing–more steak, less sizzle.

Also, having read Cory’s blog post I’m sure he has less of an understanding of Nate Silver’s methods than the average commenter so far.

1 Like

Land lines are unnecessary if you’re in parts of the country with decent cell coverage. If you’re somewhere like rural West Virginia or interior Maine, you are unlikely to have good cell coverage.


That having been said, CNN:

While he does enjoy solid support from his base, it’s hard to envision a scenario (Reichstag fires aside) where he wins many people over, especially given his wholesale ineptitude at actually governing. The left, meanwhile, is already mobilized. I’m highly skeptical of an older, whiter midterm electorate in 2018.


Or have a land line but are tech savvy enough to use NoMoRoBo.

1 Like