My observation from doing hundreds of interviews at career fairs was the fancy schools (Stanford) had students that were way better at basic questions. This isn’t a perfect metric, but it’s signal.

If your peers in undergrad were your friends, you were probably selecting a non-representative group (smart people you liked) and your job as a TA shows you a more random sample (or even skewed negative, poor students grubbing for grades).

At least that’s my guess (selection bias).

Of course if I’m wrong there’s an even bigger opportunity in undervalued students, but I’ve done enough interviews to believe that’s not true.

This word choice surprised me as well. The average Big 10 tuition for nonresidents is approaching $40,000. It is better for in-state tuition, but locking students into wherever they happened to grow up also seems unfair.

I just meant things like couldn’t afford tuition, don’t know how or can’t get loans, or couldn’t move from family, or w/e - someone who was accepted but didn’t go due to stuff unrelated to getting in (my dad was also in this category). Some large percentage of people think they’re middle class so it’s hard to know what that means without specifics.

A lot of elite schools do cover tuition now if your parents are below a certain income, but it’s imperfect and doesn’t scale with number of siblings etc.

Yeah, this is one reason why I think the outliers at state schools are undervalued. These heuristics are indirect and lead to inefficient selection, competition over the same scarce resources.

If you’re a startup there’s a strategic advantage in trying to look for talent where others aren’t and leveraging your own subject matter expertise to do so.

In moneyball the scouts selected on stupid stuff rather than on base percentage - if you know what to actually look for you can find stuff mispriced.