Statistics for hackers
Timothée Poisot
13

Hi Tim,

I’d question your assumption that any stuff is going to be “picked up by codebros to do quantitative things without bothering to look into the basics of how it works”. I guess that a lot of “hackers” are creative tinkerers, obsessive minds, restless learners, experimenters, exactness lovers, hard-working nerds — that’s why they end up as hackers, and that is why they probably won’t consume stuff without thinking about its underpinnings. Thinking about hackers as dumb pragmatic codebros seems strange and too radical.

Second, the fundamental underpinnings of statistics are not necessarily complex, and I see some didactic sense in the intellectual position that it is all just a for loop — because in a way, technically, it is. When you bootstrap, do randomization tests, MCMC, or when you write a Gibbs sampler, it is all just a for loop, or more precisely, an awesome system of nested for loops. And if hackers get the work done by thinking in for loops, rather then in symbolic math of probability density functions, more power to them.

Oh, and can you prove that reality itself is not just a giant and complex for loop?

Show your support

Clapping shows how much you appreciated Petr Keil’s story.