Hacker News Reader

Comment by ksenzee | Hacker News Reader

This very common assumption -- which I used to hold -- has it exactly backwards. Let's assume tall people and short people are equally good at cracking eggs. Mathematically, if 90% of your egg-crackers are short, you've already hired under-performing egg-crackers. You've passed over some really good tall people in favor of the short people you hired. If you want the best egg-crackers, you need to interview a bunch of tall people.

View on HN

This analysis has its own flaw. This assumes that their other opportunities are also equal. But tall people may make better apple pickers. And apple picking pays more than egg cracking. If you want 50% tall people, you are going to need to compete directly against the apple picking market as well. You need to account not just for your own selection criteria, but that of everybody else as well. (Not to say this actually explains tech demographics; it's a point about the stats in your example.)

If 90% of all egg-crackers are short, and you want to hire the best egg-crackers, it's very likely that 90% of your egg crackers are going to be short.

If 90% of all egg-crackers are short, it's likely that there's a huge problem driving tall egg-crackers out of the field.

That's a bit specious. 90% of nurses are female but there's nothing driving men out of the field.

Do you know any male nurses?

There actually are things driving men out of the field. Look at how many head nurses are male. Men in nursing is a problem as well. But this isn't NurseNews, it's HackerNews.

Let's assume tall people and short people are equally good at cracking eggs.

Umm, in the actual example we are talking about, the evidence is on the side of that assumption being false: https://news.ycombinator.com/item?id=9977751 https://jaymans.wordpress.com/jaymans-race-inheritance-and-i...

> Let's assume tall people and short people are equally good at cracking eggs. Mathematically, if 90% of your egg-crackers are short, you've already hired under-performing egg-crackers.

In fact, this depends on the underlying population sizes of tall vs. short people, so this is a faulty argument.

More germane to the situation, however, is your initial assumption, whose analogue is dubious. So much so, in fact, that it is taboo to even allude to it. cf. pg's "What you can't say".

Reply on news.ycombinator.com