If there was a "Manifesto for Empirical Software Development", what might it advocate?

- Data over anecdotes?
- Experiments over debates?
- Skepticism over social proof?
- Hypotheses over opinions?
- Weight of evidence over follower counts?
- Efficacy over "That's how we've always done it"?

@jasongorman I often try to meet “best practices” with “best for whom, in what context?”
@wlonk @jasongorman i use sensible default instead which sounds more like that it is a good choice but not the one and only dogmatic choice
(I think i read that term in a Martin fowler blogpost)