The polls were wrong. Models are worthless. Here’s how Dems blew it by trusting metrics rather than taking the common sense step of doing things that intuitively appeal to me, a college educated political obsessive living in a large metro area that Dems won easily.

[1/127]

Joking aside I do not have secret insights about the soul of the electorate in small town Iowa, northern Maine, the Rio Grande Valley, or Cuban neighborhoods of Miami.

But people who reach very different conclusions from you probably think differently than you.

That’s why surveys are good.

Surveys are also hard to do properly and evidently it’s getting harder. But the alternative would be like “deep ethnographic study” not “the one thing that most troubles me and my friends about Democrats is coincidentally the key to winning Laredo.”

But as I said yesterday to me it’s telling that a minimum wage ballot measure won big in Florida while an affirmative action ballot measure lost big in California.

You want to align with and raise the salience of ideas that win, not ones that lose.