@paulcowgill

first principles > flawed experiments there are two main ways to make decisions: 1. run experiments 2. reason from first principles the reality: perfect experiments take too long, so you're left with flawed ones. both flawed experiments and first principles with bad assumptions can lead to wrong conclusions. but here's the key difference: * flawed experiments = delayed action + unreliable data + months lost before you realize it was wrong * first principles = immediate action + real production data + forward momentum bottom line: ship decisions based on sound reasoning. you'll get better data faster from real users than from a compromised test.

2 replies

2 recasts

14 reactions