Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Solution to Harvard’s admissions problem is simple: run a lottery

qz.com

Post reply on HN
Print view View original

71–80 of 110 posts

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#71

Post by throwawaymath »

motohagiography wrote: post #58

From a statistical perspective, in a fair random lottery, would not the shape of the probability distribution of attributes of winners necessarily match that of the resulting qualified applicants? If fairness is defined in this approach as a process that does not add information to the system, and in this case actually removes both information (bias) and noise (bias) equally, all it would serve to do is further obfus…

I don't really follow your first two paragraphs. I don't know about the ultimate utility of this system, but I don't think it exhibits the statistical properties you're saying it does.

In a threshold-based admission lottery, everyone more than k sigma from the mean (for example) is collapsed into the same category, such that information distinguishing them is lost. But the implicit premise to this system is that you can't accurately measure the distinctions between those deviations anyway.

Given that premise, you're not adding noise to the system, though you are removing information. I think the claim under question is that trying to precisely measure people more than k sigma from the mean is intrinsically noisy and prone to spurious correlation with academic success. So then you'd also be removing noise under this system.

So I think your point of contention should be with the premise if you disagree with it, because I don't think we can really argue about statistical properties of the lottery distribution until we first settle on the underlying axioms.

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#72

Post by chillacy »

abnry wrote: post #12

The problem of college admissions seems similar to the problem of selling tickets to a hot event. The free market solution to selling tickets to, say, a Red Sox game is to charge as much as you possible can until people stop buying tickets. But you get a problem. You make the less wealthy fans very unhappy because they have no choice in being able to attend. This has long term consequences. So the Red Sox, for instan…

It might not matter as much to the Red Sox but if your event happens to also depend on the participation of the fans then you want to make sure your event isn’t 100% rich people too.

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#73

Post by akhilcacharya »

trowawee wrote: post #64
babyslothzoo wrote: post #55

Would it be a problem to remove all identity categories (name/legacy/gender/ethnicity/origin/etc) of applicants and simply admit the most qualified students?

"simply admit the most qualified students" is the "assume a spherical cow" of the college admissions discussion.

For the record, this is also an assumption that the "coding interviews are bogus" crowd makes too.

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#74

Post by BurningFrog »

agent008t wrote: post #51
Someone1234 wrote: post #29

Earlier quoted context omitted.

> The lottery system depicted here is unfair to those students who've gone beyond and accomplished things. Conversely it might take some of the pressure off. As it stands there's never "enough." With a system that effectively caps you, you get to your lottery tier and can then back off. And considering how much the standards have increased over time to get into a school like this, that may not be a bad thing for ment…

Would you really want a future Feynman to be playing an admission lottery?

Would you really want a future Feynman to be spending high school rushing between violin lessons, lacrosse practice, model UN and volunteering at the homeless shelter to pad up his college application resume?

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#75

Post by throwawaymath »

jihadjihad wrote: post #26
throwawaymath wrote: post #8

Earlier quoted context omitted.

I think that a lottery without a qualification threshold would contradict meritocracy. In contrast, I don't think a threshold-based lottery contradicts meritocracy. Rather it can be interpreted as a claim that meritocracy cannot be properly quantified beyond a certain precision (which in this case is the threshold). Even if you disagree with that premise, it's substantially different than the rejection of meritocracy…

But how is that threshold itself determined? Seems like it's completely arbitrary and at the discretion of the institution.

Good question! I think you can come up with different models for how to define a threshold, and I'm not done reading the paper the article is based on. But implicit in any discussion of a threshold is the idea that the threshold has to be explicit. You can't realistically hold a lottery for ostensible fairness maximization without publicly defining your baseline for lottery eligibility.

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#76

Post by subjectHarold »

trowawee wrote: post #60
agent008t wrote: post #51

Earlier quoted context omitted.

Would you really want a future Feynman to be playing an admission lottery?

...kinda, yeah. Feynman's an interesting choice of example, actually; he wanted to attend Columbia, couldn't get in because of anti-Semitism, and then attended MIT before it was "MIT, Famous Engineering School", when it was still basically seen as a vocational school for middle-class students. I think the whole higher education system would benefit from having people like that scattered throughout it, rather than con…

The problem isn't that we miss out on some specific genius. The real issue with our system is that we have "Feynmans" working as waiters and driving taxis (and some other countries don't have that system and are maximising their talent).

It is interesting to critique this idea for introducing chance. Pitch this idea to a poor person, pitch this idea to a rich person...that is your answer. It is hard to understand if you grew up with opportunity but for poor people this randomness represents a tremendous improvement. From a system that is designed to crush them, to one in which everyone has the same chance.

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#77

Post by stronglikedan »

wolfgang42 wrote: post #61
stronglikedan wrote: post #36

Earlier quoted context omitted.

I personally don't know if such a lawsuit exists. However, if it hasn't been settled in favor of the plaintiff, and no other such lawsuits have been settled in favor of the plaintiff, then they aren't doing anything illegal.

> if it hasn't been settled in favor of the plaintiff...then they aren't doing anything illegal. Wouldn't that mean that they haven't yet been determined to be doing anything illegal? If you do things against the law they're still illegal even if you haven't been caught.

Semantics aside, I subscribe to the innocent until proven guilty school of thought.

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#78

Post by throwawaymath »

dannykwells wrote: post #28
throwawaymath wrote: post #8

Earlier quoted context omitted.

I think that a lottery without a qualification threshold would contradict meritocracy. In contrast, I don't think a threshold-based lottery contradicts meritocracy. Rather it can be interpreted as a claim that meritocracy cannot be properly quantified beyond a certain precision (which in this case is the threshold). Even if you disagree with that premise, it's substantially different than the rejection of meritocracy…

I think the point is too subtle to be broadly appreciated. Any crack in the dam of meritocracy is likely to become a flood. The illusion must be preserved. wrt to point 2: I guess it's about perspective. It's fine to propose systems that in theory would be nice but without an eye towards what could actually work, what's the point?

Why is any "crack" in the "dam" of meritocracy likely to become a flood? I don't necessarily disagree, but I don't see any specific evidence justifying that claim.

Hypothetically a threshold-based admission lottery would cast more sunlight on the admissions process since:

1. you'd cease having nonpublic, ill-defined measures for admission above the baseline; and

2. you'd have to publish the threshold in any realistically accountable system.

We can argue about whether or not such a system would be (willingly) embraced by universities, but I don't see how the system itself destroys meritocracy. It seems more accurate to say it simply denies meritocracy can be meaningfully measured above a certain baseline of qualification. That only obliviates meritocratic differences above your defined baseline.

Is your contention that the baseline itself would simply shift the uncertainty "downwards" to maintain the status quo?

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#79

Post by agent008t »

Someone1234 wrote: post #59
agent008t wrote: post #51

Earlier quoted context omitted.

Would you really want a future Feynman to be playing an admission lottery?

Would a future Feynman really do enough extra curricular that have nothing to do with their area of expertise to even get into a school like Harvard? Likely not... If anything a lottery would increase their chances compared to the status quo. Harvard and similar schools have moved beyond just being smart. Everyone who applies is super smart. Now the yard stick is smart++ (meaning they want you to be "well rounded" wh…

Do they actually care about extra curriculars though? Cannot remember of a single case in Oxford CS admissions where that was a factor, let alone a deciding factor. Usually those with loads of extra curriculars were not as good as those for whom maths/CS was their true passion. They would look like they had received better admissions interview coaching perhaps, but it was our job to see through that and it was not that difficult to do so.

Re: Solution to Harvard’s admissions problem is simple: run a lottery

#80

Post by trowawee »

subjectHarold wrote: post #76
trowawee wrote: post #60

Earlier quoted context omitted.

...kinda, yeah. Feynman's an interesting choice of example, actually; he wanted to attend Columbia, couldn't get in because of anti-Semitism, and then attended MIT before it was "MIT, Famous Engineering School", when it was still basically seen as a vocational school for middle-class students. I think the whole higher education system would benefit from having people like that scattered throughout it, rather than con…

The problem isn't that we miss out on some specific genius. The real issue with our system is that we have "Feynmans" working as waiters and driving taxis (and some other countries don't have that system and are maximising their talent). It is interesting to critique this idea for introducing chance. Pitch this idea to a poor person, pitch this idea to a rich person...that is your answer. It is hard to understand if…

Absolutely agree. I don't know how much this particular proposal would help them, though; I think catching those people would require a wholesale overhaul of the educational system: universal free pre-school, standardizing school funding nationwide (and specifically decoupling it from property tax revenue), fully subsidized university education for anyone who wants it...and then lotteries to get into elite schools.

Otherwise, I suspect this would primarily benefit middle class students who can afford to pay for test prep and tutors, but not for a month-long volunteering stint in Nepal, and primarily hurt the relatively-sub-par-but-well-off legacy admits (which, who cares) and students with a lower SES that currently get a boost from consideration for that status/the follow-up effects of growing up and going to school in poor areas (which I do care deeply about).

Post reply on HN