58 ms·
Just rank by salary and pick the top N? There's plenty of things to bikeshed over, like how to adjust for cost of living, but I don't see how you could argue th
by jmgao 9y ago
Just rank by salary and pick the top N? There's plenty of things to bikeshed over, like how to adjust for cost of living, but I don't see how you could argue that this could be any worse than literal random chance.
- didibus 9y agoYou imply normalizing across role and geography is trivial, but it's not. Maybe you're smarter then me though, so please proceed to detail your rule set. Btw, this is the only scheme I think could be better, but I actually don't hear a lot of people bringing it up. And when I think more about it, I'm not sure it's possible to implement. Lotterie is in a way a solution to the normalization problem. Given two companies in different places, both who've interviewed a different candidate and believe they merit to get the job, and are both willing to pay a different amount for the candidate. You could say, instead of attempting to normalize and risking possibly high false positive, we will randomly distribute. That's a normalization scheme.