Hacker News Reader

Comment by galaxyLogic | Hacker News Reader

I think that makes sense. If you cannot be very sure of actual evidence why rely on it anyway? But randomness might mean that every time you use the result you learn something, at least you would learn whether your rule-based default value for the control variable is systematically wrong and tilted to one direction or another.

View on HN

I think companies could test their interview process this way. Fill some positions by choosing from the candidate pool at random (there could be a baseline filter to ensure a minimum level of competence) and compare the random hires with the regular hires to see how much of a difference the interview process really makes.

Fascinating idea. Do you know of a company that’s done this? Or maybe planning to do this?

No, it's just a half-baked idea I've had after reading people's arguments for and against various interviewing approaches.

We’re more or less hiring at “pass the baseline level of competence” and I don’t think we’re alone.

I've long advocated for building a candidate pool for both hires and promos from a low baseline (with the option to opt out for promo), then hire/promote randomly out of that pool. Then fire/demote based on demonstrated incompetence.

The important thing isn't that you promote the best worker; the key is that it isn't possible for the ambitious-but-incompetent to game their way up.

Wasn’t there some recent research that showed that hiring randomly actually works better on average?

Edit: Wait, that was promoting.

That goes along with my thinking. Some things are obvious but when you move away from the top topic or two you’ll always have a long list of things where it’s not so clear which should be done next, and there’s plenty which is always important to do but never the most important thing. A bit of randomness injected into that process whether you’re dealing with software tickets, national legislation, or household chores might just lead to better outcomes than relying on the judgement of a few people without enough information to make nonarbitrary decisions and might have bad biases which lead to things getting overprioritized or neglected.

Reply on news.ycombinator.com