One of the most frequent questions I get about this election is some version of, “why were the polls so wrong?”

Assessing the accuracy of polls is much more complicated than it looks.

Some were clearly off. Ann Selzer is a fine pollster, but she clearly made a mistake somewhere when she found Kamala Harris ahead in Iowa by 3 points days before the vice president lost the state by 13. That’s an easy case.

There were other, less egregious, outliers, as there always are. Outliers are in consonance with the laws of statistics.

One poll had Harris 4 points ahead nationally, though she’s losing at the current count, albeit by less than 2.

A Nevada survey had Trump up 6 points, while another in the same state found Harris ahead by 3. Trump won by there by 3. Both polls were outliers; clearly neither was on target.

But which of these polls was more accurate — Echelon Insights, which gave Trump a 6-point lead in Pennsylvania, or the........

© The Hill