[twelfthnight]

This is great statistics, but it avoids the problem that moving to online polling has made it very difficult to get representative populations, so the data itself is biased in ways that cannot be counteracted by methods (only by assumptions, priors, etc). Which makes this misleading because it gives the forecast air of confidence that is unjustified.

[JumpCrisscross]

> makes this misleading because it gives the forecast air of confidence

The entire first half of the article is about why polling is not only unreliable at this point, but becoming increasingly unreliable in biased ways.

[twelfthnight]

True, I suppose my disagreement is that I believe it doesn't go far enough to explain how big of a deal it is, and how there _aren't_ ways to deal with it without substantial, subjective intervention from the forecasters.

I've worked on weighting code for online polls, they literally rely on dozens of hand picked decisions to stay "reasonable". These decisions aren't factored into the error bars, making them appear smaller.

And as far as the fundamental style predictions, how can you use a single GDP number when Fox tells its viewers one number and MSNBC tells its viewers another?

This article does describe a faithful statistical effort, but to me it doesn't emphasize the risk of a "black swan" event enough.

[whakim]

Absolutely. But we’ve been past the “golden age” of polling using live callers on landlines for more than 20 years now. We now have a reasonable corpus of polling data that we can use to evaluate how good pollsters are at making the corrections (often educated guesses) that they use to adjust their polls.

[AlexCoventry]

The justification for Bayesian inference, that the posterior will eventually converge to the true distribution, breaks down unless your prior has support for a good approximation to the data-generating model. So without a good model of how polling results map to actual voter distributions, the Economist model is guesswork, to a large extent.