6 ms·
That seems orthogonal to me. Say I have a list and I'm taking some number of items off one end. I want to maximize the value of that chunk. A larger chunk is be
by munificent 13y ago
That seems orthogonal to me. Say I have a list and I'm taking some number of items off one end. I want to maximize the value of that chunk. A larger chunk is better, of course, but sorting the list so that bigger items are on one also helps.
That's what the author is arguing: we're incorrectly ranking people because of a gender bias. Having a larger pool of poorly ranked people won't solve that.
- raganwald 13y agoI don't follow what you're saying about a list, but if you're saying that either or both help, the key thing to remember is that it isn't a dichotomy. In the context of HN, what startup isn't trying to hire all the smart and capable people it can? If it comes down to two people, and there is a perception that A is better than B but both A and B are excellent candidates, a great company hires both.
- munificent 13y agoRight, but by saying "comes down to two people" that implies you've already done some filtering: you've ordered the set of all possible people and then taken the first two. My point is that if the ordering relation is wrong (which is what the article suggests) then it doesn't matter what subset you take: you're still not getting the best possible set of people. You have to get the right people. If your ordering is off, being more inclusive lets in as many more bad candidates as it does good ones.