Hacker News Reader

Comment by alexbeloi | Hacker News Reader

I agree that this is a bad way of measuring humans accurately, but I would say that _accuracy_ is not exactly the correct or optimal goal for hiring.

Google (and other's) strategy is to optimize for precision rather than recall or accuracy, which is the optimal strategy when you receive several orders more candidates than you are capable or willing to hire.

To maximize precision, you need to minimize Type 1 errors (False Positives), and you don't really care about Type 2 errors at all (False Negatives, e.g. throwing the baby out with the bath water).

This is of course if you treat people like data points, which seems more and more common (and explicit) with larger companies.

View on HN

That would be great if nobody ever tried to emulate other companies.

But they do. Look at all the people trying to pretend like they have problems that require Google's infrastructure instead of just three machines with a shitload of memory and some redundant networking hardware.

If you don't react to the fact that other people are copying your actions then your moral compass is broken.

Well but is that really Google's responsibility?

And if company A can have the luxury of copying Google and ignoring perfectly capable candidates because they had a C+ in one exam, what is the issue? There is then company B that cannot afford not to have that candidate recruited and she might turn out great.

At the end of the day, GPA is an imperfect ordering of candidates by quality, but it is such an ordering, and companies with the sweetest offer (money, prestige) will be able to get the best candidates, be that they look at GPA or not.

And it is not as if someone with a C+ is banished from the workforce - just from Google and the other companies that can afford to be so exquisite.

Just to clarify, Google doesn't actually hire candidates based strictly on their academic performance. If you've had any side projects or prior jobs before applying, I'd say that's weighed more heavily.

Sure, I was just commenting on that specific dimension of the decision space.

This assumes a false positive is such a bad thing. I've heard a bad hire costs $X, where X is some surprisingly high number, but why must it be this way? And are these big companies even avoiding bad hires in the first place?

The other problem is sameness bias. I posit that those false negatives are disproportionately folks that are underrepresented, demographically. Therefore, this approach has concerning externalities.

>I've heard a bad hire costs $X, where X is some surprisingly high number, but why must it be this way?

Because most companies don't (and generally shouldn't) operate like pro football teams. "You've had a couple bad games; we're cutting you. Sorry it's just business." Most (though of course not all) people think that once you've hired someone, you should really try to make things work. Both because of the costs associated with someone getting up to speed at a company and because of the personal cost to the person being fired. Different companies have different philosophies of course.

I think there are humane ways to part ways with an employee who just didn't work out. Give generous severance and some warm intros to places that might be a better fit.

If you really screwed up, make that choice soon. In the more common borderline case, that person is still giving you decent value, so it's not a total loss, even if you invest effort in trying to coach them up to your high bar.

I like the sentiment, but I would rather a potential hire just apply cold than if a company told me to look at a candidate they just let go (even if it was google).

Warm intros from an entity that just rejected you would be a terrible signal.

Being precisely inaccurate isn't optimal. There needs to be a balance between precision and accuracy for a measure that's close to reality. Of course, there's going to be tradeoffs between the two, as well.

Reply on news.ycombinator.com