Last summer, the United Kingdom voted to leave the European Union in what bettors, financial markets and the London-based media regarded as a colossal upset. Reporters and pundits were quick to blame the polls for the unexpected result. But the polls had been fine, more or less: In the closing days of the Brexit campaign, they’d shown an almost-even race, and Leave’s narrow victory (by a margin just under 4 percentage points) was about as consistent with them as it was with anything else. The failure was not so much with the polls but with the people who were analyzing them.
The U.S. presidential election, as I’ve argued, was something of a similar case. CONT.
Nate Silver, FiveThirtyEight