Deep.
At this level the ea vs. ed distinction matters less than you suggest. Since hyps use SCEA rather than unrestricted ea like mit does.
Deciding to go scea means you don’t go ed anywhere. And ed means no scea. So the effect of scea approaches that of ed
Your point would be stronger if there was no single choice.
“I think it’s odd that H would beat M in STEM cross admits, but that Princeton wouldn’t do as well.”
Location.
Well, according to some posters on the UChicago threads, @northwesty, yield for this past year there (disclosed unofficially by admissions staff/administrators) was ~75%, which would put it above Yale and Princeton and behind only Stanford and Harvard. This, of course, is the result of admitting the vast majority of the class ED. Yet, given the choice between UChicago and either Yale or Princeton, I’m willing to bet most cross-admits would go with door #2 or #3, so I’m not ready to agree that yield is always the best proxy for revealed preference.
I also think the difference between SCEA and ED is significant. If they want, UChicago can fill virtually their entire class with a combination of ED1 applicants who love UChicago above all, ED2 applicants (many of whom may have been spooked by being deferred or denied at their HYPS first choices), a few kids they take a flyer on in the RD round and a group cherrypicked off the waitlist who get offers on the explicit condition that they’ll immediately accept them. Oh, and some Z-listers. UChicago can sculpt the class with a high degree of precision, knowing that >95% of the kids they admit are coming, because they’re basically required to matriculate unless they simply can’t pay what UChicago is asking them to cover. The price for this is that UChicago consciously restricts itself to a subset of the applicant universe that HYPS are targeting.
Contrast that with HYPS. Sure, 85%-90% of the kids they admit early will come, because it’s HYPS. But HYPS want to have a shot at as many as possible of the tippy-top candidates, not just pick from a subset of undeniably high-achieving candidates in the way UChicago seems prepared to do in the service of increased selectivity and yield. HYPS also don’t want to send a message to non-full-pay candidates that by far the easiest way in is the one where you give up the right to compare aid offers (because that will turn off some of the best candidates, leaving aside the fact that it arguably unfairly selects for full payers).
If, say, Yale admits too high a proportion early, they’ll lose more than they gain, by driving too many tippy-top candidates to conclude that they have no shot at Yale unless they apply early, and therefore to use their early bullet elsewhere and maybe not bother applying RD. RD numbers might then go down significantly. Sure, Yale could goose its yield by admitting more of the class SCEA, or even moving to ED. But it would be an illusory success, and wouldn’t say much about revealed preference relative to its peers, it seems to me.
@DeepBlue86 I agree, it is pointless to compare the Chicago yield with that of HYPS or even other top ED schools. They are admitting 75%+ of the class early, of which the majority is ED. You can’t even really compare their RD yield since their RD admit pool is so tiny now. The 75% yield they will have doesn’t mean much.
All the data is significantly influenced by scea and ed.
Even the cross-admit data. Since the cross-admit pool does not include so many of the possible applicants who got excluded by a successful ed application.
At this rarified level, the difference between scea and ed is more like oranges and tangerines rather than apples and oranges. Everyone’s yield is getting goosed up, although some a bit more than others.
If HYPS weren’t getting their numbers goosed up by scea, they’d just use unrestricted ea like MIT does. Or some kind of mildly limited ea. Why exactly is H prohibiting a kid from applying non-binding ea to YPS since it is all is so non-binding?
46% of all Harvard offers came through scea, meaning probably more than half their class comes via scea. That’s VERY ed-esque.
I think all the gaming mechanisms mostly cancel each other out, leaving yield and YTAR as pretty reliable ranking indicators as the results show.
Sorry Caltech (under 40% yield). : (
MIT is second to none when it comes to the best and brightest for STEM. Check out the following lists and see where most of these folks landed (as examples).
Physics
https://www.aapt.org/aboutaapt/2017-US-Physics-Team-Announced.cfm
https://www.aapt.org/physicsteam/2016/team.cfm
Math
https://www.maa.org/news/2016-usamo-winners-announced
https://www.maa.org/news/winners-of-the-2015-usa-mathematical-olympiad-announced
A student like Allen Liu need not even fill up his college app 
https://artofproblemsolving.com/community/c5t360f5h1272498_allen_liu_is_absolutely_unreal
https://www.washingtonpost.com/news/answer-sheet/wp/2016/07/18/u-s-students-win-prestigious-international-math-olympiad-for-second-straight-year/?utm_term=.86197292ca9c
@northwesty the mechanisms definitely do not cancel each other out. SCEA is definitely not the same as ED. HYPS really do not need to protect their yield from any other schools other than each other. Many people who get into HYPS SCEA either do not apply anywhere else because they got into their first choice or they apply to other HYPS schools. HYPS lose the vast majority of cross-admits to each other not other schools.
Comparing the yields of top ED schools with HYPS is complete apples to oranges. It doesn’t even make sense to compare RD yields between HYPS and the other top ED schools.
Overall yield is only meaningful when comparing just HYSPM to each other and RD yield is meaningful when comparing top ED schools to each other.
“Many people who get into HYPS SCEA either do not apply anywhere else because they got into their first choice.”
That sounds like ed to me.
“HYPS really do not need to protect their yield from any other schools other than each other.”
That sounds just like ed to me as well. Limiting applications to one school to increase yield. Tangerines to oranges.
Within the top 10, yield is a quite reliable ranking metric. Sure Duke and Penn are big users/abusers of ED, but they still come in below HYPS on yield.
What outcome that comes from using yield and YTAR do you disagree with?
Good thing about using yield and YTAR is that the schools publish that data. Cross-admit data is partial and self reported by the applicants. So not a totally solid data set.
I don’t think so, @northwesty, because if it really were the same as ED, HYPS would all go to ED - after all, doing so would increase their yield somewhat, since it stands to reason that if they don’t absolutely have to matriculate, some admitted candidates will find reasons not to.
HYPS don’t go to ED because, based on the fact that they know that they’re going to land the great majority of those they admit, they’re simply trying to get candidates to show their preference. They’re not worried about losing candidates in large numbers to anyone but each other (particularly since they all have similar super-generous fin aid), they know that many applicants won’t apply anywhere else after getting into one of HYPS early, and they’re willing to accept slightly lower yield and take the risk of losing a few admits in order to have a jump ball with a better pool of RD applicants (some of whom may have an SCEA offer in hand from another school), and to avoid putting low-income students into a worse position by making them apply ED.
I think there’s a correlation-causality issue at play here. HYPS don’t have such high yields primarily because they use SCEA, but because they’re HYPS. SCEA is a tool that helps them compete better with each other, not one that they’re using to compete more effectively against aspiring members of the club. That’s why I think you can’t really compare HYPS to anyone else in terms of yield and selectivity.
The question to ask is: if UChicago dropped ED and went to SCEA, how competitive with HYPS do you think their yield and selectivity would look? I think we might all agree on the answer, which is: don’t know, exactly, but not nearly as good as they look now.
Of course you can compare them all on yield.
S, H, M, Y, P, Penn, Col, Chi.
I just did.
What don’t you agree with?
Of course you can compare them all on yield.
S, H, M, Y, P, Penn, Col, Chi.
I just did.
What don’t you agree with?
HYPS use SCEA because (a) they believe ED is unfair to applicants and favors wealthier applicants, and (b) if they went to unrestricted EA (back when some of them tried unrestricted EA) their early applications quadrupled.
It’s interesting to note that, in the most comparable example elsewhere in the world, not only can you not apply early to more than one ultraselective college, you cannot apply at all to more than one ultraselective college. And Oxford and Cambridge don’t give a rat’s tushy about their yield.
@northwesty Not at all true. “Many people who get into HYPS SCEA either do not apply anywhere else because they got into their first choice.” or they apply to other HYPS schools RD - you missed that part.
This sounds nothing like ED. in ED candidates are not allowed to apply to any other schools RD. Many would jump at the opportunity to apply to HYPS having already had a top ED school acceptance. When it comes to SCEA schools people do not bother to do so because they are already in their very first choice or they actually go ahead and apply to another HYPS school RD… If they want to apply, nobody is stopping them to do so in RD. They are not really limited like they are with ED.
You can definitely not compare all the top 10 on yield. HYPSM loses cross admits almost exclusively with each other. For example, you cannot compare the Penn’s yield (67%) with Princeton (also 67%) and say that Penn is just as desirable as Princeton. Or now that Chicago has a yield of 75%+ that it is more desirable than Yale or Princeton. That is completely absurd.
You can only get meaningful insights if you compare yields amongst HYPSM only, and then RD (not overall) yields amongst the top ED schools (Columbia, Penn, Chicago, Duke, Brown, Dartmouth, Cornell, NU,JHU etc).
You have a cite for Chicago’s 75% yield?
Have they even released their 2021 data yet? I have not seen it.
You guys protest too much. If ED was an all powerful data manipulator, how come Duke and Penn still come in below HYPSM on yield? And why wouldn’t NEU go all ED to further fuel their rankings climb? There’s limits to how much ED can do, and SCEA is the tool that HYPS use to goose their numbers.
Yield is just as good a ranker as cross-admits. And the data is much more solid.
I think @JHS and I broadly agree on the following:
Personally, I very much doubt that UChicago posting a 75% yield is going to change any minds at the HYPS admissions offices about the use of SCEA vs. ED. The fact that UChicago now has a yield in the same ballpark is neither here nor there, because HYPS know how it was obtained, don’t think it’s worth it to play by those rules, and don’t expect to lose many kids to UChicago - i.e, they don’t think UChicago’s newly-lofty yield is an expression of true revealed preference. This is why I don’t think you can compare HYPS on yield to anyone but each other.
The balancing act is for each of HYPS to admit just enough applicants SCEA to achieve the desired effect without disturbing the current equilibrium and setting off an arms race wherein the other three would feel compelled to respond and the end result would be for every applicant to self-select into one of the early pools and only apply anywhere else RD if they weren’t admitted early to the school of their choice.
@northwesty - I don’t believe they’ve released a yield number or other statistics; various people have been piecing it together and trying to do the math based on statements made by admissions officers at functions for admitted students. See one such thread here: http://talk.collegeconfidential.com/university-chicago/1961894-sobering-statistic-ea-ed1-class-of-2017-university-of-chicago-p1.html
Another more recent thread (http://talk.collegeconfidential.com/university-chicago/2017063-uchicago-holds-third-place-with-yale-in-2018-usnwr-ranking-p8.html) cited a statement that put the number of students in the class at 1,780. If so, depending on the number of Z-listers, the yield could be higher than Stanford’s.
According to the Chicago Maroon (https://www.chicagomaroon.com/article/2017/4/6/uchicago-release-class-2021-acceptance-rate-spring/), the university didn’t release admissions statistics in the spring, as it had done in previous years, because “the numbers will not be final until the third week of fall quarter” (which is in mid-October). Frankly, I’ll be pleasantly surprised if they release much granular detail, because it seems very likely that this would show relatively high ED admit rates and an RD admit rate lower than Harvard or Stanford.
@northwesty all of the top ED schools Columbia, Penn, Duke, Chicago (until last year that is),NU, brown etc have yields lower than HYPSM. This is because they lose a big chunk of their admits to each other and HYPSM during RD, and they choose not to rely even further on ED to raise yield because that restricts the quality of students they can get. If they increased the % of the class they got ED even further, then they could have yields higher than HYPSM. Chicago just did that (reportedly over 75% of the class admitted early) and its yield is allegedly around 75%, higher than Yale and Princeton.
HYPS do not face the restriction of failing to attract the top talent if they rely mainly on ED. But still they choose not to do it because they do not need to. They do SCEA to limit competition from each other. A person who has gotten early into HYPSM will not bother applying RD to any non-HYPSM schools. They either apply to another HYPSM school or no school at all for RD.
Yield is for sure not a good metric when applied generally across all top schools. It is much more nuanced than this.
Not saying my kid is a top student type but after he got into Stanford REA, he was too lazy (and tired) to apply to other top schools. It’s not that he did not view other top schools (including Berkeley and UCLA) as having certain advantages over Stanford, but he felt fortunate to have been accepted into his number 1 choice and wanted to do something else with his time than spend it on working on more applications. In his and our minds, the very effort of applying and choosing between/among top schools was not worth the effort. Of course, a possibility of getting denial letters did not seem attractive either. This is not to say we found everything great about Stanford during Stanford’s Admit Weekend, but we met some parents who had already attended Admit Weekends at Harvard and other schools, and they also had some complaints about other schools. He did use additional free time he had to apply for a summer paid internship though.
I know it sounds dumb but our kid does not like the butt-freezing type of cold weather, so that made it not a difficult decision. lol I do not envy those who got into so many top schools and have to spend a lot of time and effort making a choice.
“HYPS lose the vast majority of cross-admits to each other not other schools.”
“Or now that Chicago has a yield of 75%+ that it is more desirable than Yale or Princeton. That is completely absurd.”
You guys seem especially focused on UChi and their recent addition of ED.
Did you miss that they made it to T3 in USNWR on the data that doesn’t include ED?
And Parchment says they lose 59/41 to P on cross admits. 66/34 vs. S. They seem to be doing pretty well even before ED…
@northwesty Why are you quoting me selectively…I spoke about Penn, Columbia, Chicago and other top schools, not just Chicago. Sorry but the top 3 rank doesn’t mean much. No one is really buying that. both Columbia and Penn i addition to Chicago have been in the top 5 for many years at one point or another, but people do not see them as on par with HYPSM. Same goes for Chicago now.
What does Parchment have to do with anything. Yes Chicago is doing as well as any other non-HYPSM top 10 school would. So what? it still remains that comparing its artificial 75% yield with HYPSM yields is pointless. As is the case for comparing Columbia’s or Penn’s yield with HYPSM. The yield of all these schools are a product of extensive use of ED. you just can’t compare them with HYPSM schools.