Paul GrahamVerified account

I hope she bets against 538 if she's so confident. It's easy to say stuff, harder to put skin in the game. (Note also that 538 had the highest % for trump that I saw in 2016, and is still time and again pointing out that 17% chances happen all the time)

"All models are wrong, some are useful."

Competent? Yes. Exceptional? No way

Everyone who keeps saying Nate Silver has no "skin in the game" seems not to understand that if your professional career is based on accurate predictions, that might motivate you a little more than ten bucks on predictit

Nate is beast. A lot of startup data guys really have to learn from him. The methodology is as important as the final numbers. A lot of people focus on the second part and ignore the first

There will be a big market for a new conservatism. There will also be a lot of conservatives who have potential to create the new conservatism wasting a lot of time whining about supposed leftists, when in fact there are hardly any actual leftists in the US.

The same arguments get used for “data storytelling,” “climate science,” and “scientific polling.” More often, because humans are involved, the reality is to start with the answer and massage the numbers/theory (not science) into something sellable. Unconscious bias projection.