It’s lots of fun to see so many people treating Duke – one of the top universities in the country in almost every ranking – as an inferior school. Usually, it’s my alma mater, Cornell, that gets that sort of treatment.
A very large number of students would sacrifice major body parts to go to Duke. Or Cornell. We’re not talking about East Middle of Nowhere State College here.
No kidding, @Marian, particularly when you consider the fine distinctions people make between schools. See Sayre’s Law (h/t Wikipedia):
Sayre’s law states, in a formulation quoted by Charles Philip Issawi: “In any dispute the intensity of feeling is inversely proportional to the value of the issues at stake.” By way of corollary, it adds: “That is why academic politics are so bitter.” Sayre’s law is named after Wallace Stanley Sayre (1905–1972), U.S. political scientist and professor at Columbia University.
Marian – That’s a great comment for someone educated at a state school in cow country!!! : )
Once you get into this conversation, you always wind up slicing it really thin. Google up the threads on ranking “prestigiosity” using “milliHarvards” as the scale. Harvard is 1000 mH. So is Yale or Stanford 999 or 998? That turns into a nasty knife fight real fast.
“I like your four-way EA idea, btw, but I think it would be simpler and have a greater chance of putting a lid on things to cap total apps from any individual at, say, eight.”
My idea reflects my most recent (and quite reasonable) experience with my third kid. The two early favorites came down to ED Brown and EA Notre Dame (which is a somewhat weird combo). After much back and forth, ND had a very slight lead and so we necessarily headed down the EA path. Applied to two EA reaches (ND, UVA), one EA match and one EA safety by 11/1.
Between 11/1 and 12/15, applied to Brown and few other schools RD. Got into ND and the match and safety (both with merit money) by 12/15 (with no boost coming from the EA). Still waiting on UVA. So at this point, don’t feel much need to flood the zone with a ton of lottery ticket RD apps, although my kid is free to do so if the kid wants to write more essays.
If the feedback from the EA reaches had been negative, then we’d be more active right now doing apps. But we’d probably be shooting at more matchy schools than going trophy hunting. And we get to see how the RD apps and the merit money and the scholarship apps roll in between now and May 1. TBD which school will appear most dreamy as of May 1.
A very reasonable, phased experience. Which we would not have had if (i) Brown had been the clubhouse leader as of 11/1 or (ii) the 11/1 clubhouse leader ND was an ED school.
That sounds like a great place to be, @northwesty - congratulations. In a funny way, though, everyone’s lives would have been easier if ND had been ED. Your kid would have applied and been admitted to their first choice, and today would be happily anticipating the next four years in South Bend. Instead, they’ll be re-evaluating over the next several months and if they get into Brown, UVA and wherever else was a reach, there may be agonizing and second-guessing. This is, of course, assuming that aid (if any was requested) would be similar at all the most desired options.
You recognize, of course, that you are (or your kid is) contributing to the surplus-of-apps problem you criticize 
They could, if I’m understanding the situation correctly, pull every app but Brown at this point.
If it had been ED Brown vs ED ND, you are right that we probably would have picked just one and run with it. Since that’s what the game rules tell you is the smart play. The kid would have been happy if the ED app at either place hit the target and wouldn’t look back. And if the ED bullet missed the mark, RD lottery ticket apps would now be getting produced at a furious rate.
But my kid in reality was 50.1 vs. 49.9 between the two. Doing ED really wasn’t a fit for my kid. But the game rules (by design) incent you to act like you are 100/0 when you really are not. If the system gives you one silver bullet to shoot, you feel like a chump if you leave that bullet in the gun. But that’s a first world problem for sure.
I like the idea of a draft, but not the one proposed by @Zinhead . There is at least one clear model out there already and works well from all accounts–teaching hospitals selecting medical residents. I’m not a doctor, but as I understand it, the med students apply to a certain number of residency programs and state a rank preference based on geography and specialty or whatever, and the hospitals then do their ranking and “draft” the med students they want. (I think the actual picking is done by a neutral party operating a computerized database that matches the student to the hospital.) I don’t believe that there is much “choice” after the draft (i.e., the student is given one offer and has to take it or leave it). It’s probably much harder to pull off on a larger scale given that undergraduate programs are so much larger than med schools, but it seems that given enough computing power and AI, a college draft system is not too far off from being technically possible. But it would need a bunch of top schools to agree to participate.
I’m not sure it would really present an antitrust problem if structured correctly (or maybe you have to get an exemption from Congress).
I think this proposal would also be preferable to Zinhead’s idea of a school drafting a whole slate in order of perceived rank. That gives the first schools a huge advantage (and it basically sets in stone the current ranking of schools). A blind ranking system would allow both schools and students to state their preferences and then get matched.
This idea comes up frequently, but is just not the same as undergrad admissions. Residents get paid. Undergrads have to pay the net cost of attendance (after any Finaid)…
Also, we’re talking about medical school graduates who’ve chosen a specialty, getting sorted into hospitals with residency programs looking for fixed numbers of doctors in those specialties. Probably a substantial majority of incoming college freshmen major in something other than they expected going in - how would you match them?
Presumably, the variation in medical residents’ pay is also much less than the net prices that a given prospective undergraduate may see from various colleges that admitted him/her and may have offered financial aid and/or scholarships.
Well, I haven’t thought too much about it, but I suppose the student ranking could be as simple as just ranking their top schools 1 through n. 1.Stanford, 2. Michigan, 3.Duke, 4. Harvard 5.Wharton 6. Penn 7.Creighton, etc. The students wouldn’t have to rank what their area of study is or anything like that. It’s a straight preference. Bob may have a list as stated above, while Teri has a list of only Boston-area schools. It would be the schools that do the heavy lifting of determining what field of study the student have a stated interest in, their GPA, SAT, recs, etc. and then they give a score to that student application. The schools would be free to give higher scores to URMs or top trombonists, good interviews, 800 math SATs, or rich alumni children or whatever criteria they choose to select for. The computer would then match the highest scoring students with their highest stated preference. Bob, the student above with Stanford as his number one, whose ranked score by Stanford is in the school’s top tier of scores would be a match, and the computers would sort out every single student until they are matched with the highest preferred school where their score gets them in and the school still has an available seat. I don’t know what algorithms that the medical residency program uses, but it doesn’t have to be much more complicated than that, does it?
The tough issue is the financial aid, and I suppose what would have to happen is that every school that participates would have to guarantee aid sufficient to attend the school, and students would be allowed to negotiate higher aid with their match school. But that isn’t much different than the residency programs where the residents are basically trusting that they will be paid enough by their match hospital that they will be able to live off it.
My understanding of how the medical residency matching algorithm works is that hospital rank matters, and it basically sets in stone a particular ranking of hospitals, although there’s some movement over time. The effect of ranking isn’t as radical as it would be in Zinhead’s proposal, but it’s there. There are 50 other reasons why that model can’t or shouldn’t be extended to undergraduate admissions, but it doesn’t get you out of the ranking problem, either.
I think the difficulty is that college admissions as practiced by the elites is based on a school accepting a cohort of students that it theoretically likes roughly equally, for different reasons, resulting in a class that is a mix of people who satisfy various institutional needs.
In this sense, a class is like a football team or an orchestra where you have a limited roster and you need the right number of people for every position - it’s no good if you end up with five quarterbacks and no kickers, or five tympanists and no bassoonists. Sure, there are some students everyone wants (I’m looking at you, Malia Obama) and some who are only going to get into one school at a given level (certain legacies), but, for the most part, I think the schools would struggle to rank the students in the classes they admit - everyone’s there for different and valuable reasons, and there were many others with similar attributes who were passed over because the slots they could have been a match for were filled by people the school thought were a little better and wanted a little more.
In that kind of situation, you need either to be able to make separate deals with everyone individually, which is effectively how it’s done now by colleges and orchestras, or you need a professional sports-type draft model, where there’s a known, fixed talent pool available to all the teams, which take turns selecting from it and are therefore able to fill all their positions eventually.
Princeton is well-regarded in engineering, but also has very strong humanities departments. More men than women go into engineering. More women than men major in certain humanities disciplines. Princeton wants, as an institution, to have a roughly equal mix of men and women, and, depending on the resources in its various departments, some reasonable idea of how many students it’s going to get who are interested in engineering vs. humanities. Princeton also has a general sense for what’s an optimal mix of students by race and geography. It has sports teams and an orchestra to staff appropriately, etc. It wants to have a significant but not too large number of legacies in the class. There are some celebrity and development admits that it feels like it might be in a good position to get. And so on…
How would the Princeton admissions office rank-order close to 30,000 applicants to ensure that the matching algorithm didn’t stick them with, say, 60% men from the New York area who want to be engineers, negligible numbers of African-Americans and students from the Mountain time zone, 20% legacies and no development admits or quarterbacks? I’m not sure it’s possible.
Ironically, the algorithm might get better results if it were used for the slices, e.g., male engineers. That would, however, require every applicant to be classified in an unacceptably reductive and potentially offensive way (the notion of universities rank-ordering a cohort of “African-American women” or “Asian STEM men” so they could be sorted by the algorithm comes to mind). With holistic admissions, which we have today, at least schools can consider the whole person.
Huh? If you don’t see the difference between a school like NYU and GW that has lotsa (city) love from applicants that does not meet full need…(not to mention that MD’s have “guaranteed” jobs waiting for them somewhere).
Except some/many of those schools admit by College: Engineering, Business, A&S…and by Major within college: Civ Eng, EE,…
I’m just spitballing here, so no need to get so worked up. For some schools that have a history of not meeting need, they would either have to step up the fin/aid game or not participate in the program. Alternatively, the participating schools could all abandon the need-blind system and instead go to a collective system where the algorithm also determines fin/aid for all students so each student has a “number” which says how much they will cost the school to accept. That could go into the school’s ratings.
But as I stated above, the medical residency program has to deal with this in some manner because some residents will be living in NYC and others will be living in Omaha Nebraska, so there’s got to be a certain amount of “trust me” when a resident’s compensation is set. The resident is stuck with single match and they are hoping that they are not being thrust into abject poverty for the next two or three years. Maybe the medical residency program has some compensation formula based on cost of living of each city that each hospital has to meet. I don’t know, but given the wealth of the top 50 or 100 schools in the US fin/aid doesn’t seem like an insurmountable problem for such a match program.
Regarding medical resident pay, http://www.medscape.com/features/slideshow/public/residents-salary-and-debt-report-2016 lists averages, but not variances within category (region, specialty, etc.). It is not obvious how pay rates are set. It does note that 40% of medical residents have over $200,000 in medical school debt.
But perhaps the relevant question is, are medical students informed of the resident pay levels for each hospital and specialty before they apply? If they are, then they can write their preferences with that in mind. For high school seniors looking at colleges, that would be like having each college give an actual financial aid offer before students apply into a theoretical matching system modeled on medical residency matching.
I disagree with a lot of @northwesty 's logic:
Eliminating ED will not make it markedly different for Johnny to get admitted to Penn. The difference between RD and overall admittance rates is usually around 3 percentage points…I take this to mean that 97% of kids will get the same decision if ED were eliminated. Yes 3% of borderline applicants will now be accepted, bumping an equal number of students (formerly admitted under ED) to a ‘lower’ school. Maybe some think this is more fair, but I don’t see it materially changing the big picture. And it is certainly not enough to change anyone’s mind into thinking, “now I only need to submit half as many apps.”
Definitely eliminating ED will increase the total number of apps: all those former ED admits now have to submit 10 more. And these are very strong candidates who will be accepted at their high choice, only they no longer have a way of knowing that in advance.
I also question whether ED is a rankings game when USNWR says admission rate factors in at only 1.25% by their algorithm.
That said, I actually like @northwesty 's proposal of EA everywhere, with a max of 4 EA apps per student. Colleges still get some level of candidate-seriousness, they can still adjust thresholds for their desired EA admittance percentage, and families can merit-shop if they choose. I’d guess a large number of students will accept one of their EA admits, thereby reducing the number of match and safety apps later. That could be a big benefit for second tier colleges who are probably flooded with safety apps from Ivy-aspiring students.
Thinking about it, I believe @northwesty’s proposal of EA everywhere with multiple apps permitted is unlikely to happen, because it represents unilateral disarmament for the SCEA schools (particularly HYPS - the situation’s different in certain respects at M). When these schools lose admits, it’s very likely to be to each other, and right now, when one of them gets an early app, they know with a high degree of certainty that they’re that kid’s first choice, because she’s chosen to use her single bullet there rather than face 3-4% odds in the RD round. The last thing Harvard and Stanford want is for a lot of the top kids to be able to apply early to both of them, with no implied commitment either way.
Here’s an example of a matching process, as outlined by the organization that runs the medical residency match program. Sorry the format doesn’t transfer, so you’ll have to look at the site to see how the tables look.
http://www.nrmp.org/wp-content/uploads/2014/05/Run-A-Match.pdf
[quote]
RUN A MATCH
Five applicants are applying to three programs. After considering the relative desirability of each program, the
applicants submit the following ROLs to the NRMP:
Applicants’ Rank Order Lists
Anderson Chen Ford Davis Eastman
• Applicant Anderson makes only a single choice, City, because he believes that he will be ranked highly at City and
has assured the program director that he would rank City number one.
• Applicant Chen ranks City, which she prefers, and Mercy. She believes Mercy will rank her first, and so she reasons
that there is no risk of her being left unmatched, even if she does not rank additional programs.
• Applicant Ford would be pleased to end up at Mercy, where he had a good clerkship, and believes they will rank
him high on their list. Although, he does not think he has much of a chance, he prefers City and General and so
ranks them higher than Mercy.
• Applicants Davis and Eastman have interviewed at the same programs. Like the other applicants, they desire a
position at City or Mercy and rank those programs either first or second, depending on preference. In addition to
those desirable programs, those applicants also list General lower on their rank order lists.
Two positions are available at each program, and they submit the following ROLs to the NRMP:
Programs’ Rank Order Lists
Mercy City General
Applicant Rank Program Status Match
Anderson 1. City City has 2 unfilled positions Tentatively match Anderson with City
Chen 1. City City has 1 unfilled position Tentatively match Chen with City
Ford 1. City City has no unfilled positions and
tentatively has matched with
more preferred applicants
2. General General has 2 unfilled positions Tentatively match Ford with General
Davis 1. Mercy Mercy did not rank Davis
2. City City has no unfilled positions and
tentatively has matched with
more preferred applicants
3. General General has 1 unfilled position Tentatively match Davis with General
Eastman 1. City City already has 2 tentative
matches but most prefers Eastman
Chen is removed from City to make room
for Eastman; tentatively match Eastman
with City
Chen 2. Mercy Mercy has 1 unfilled position Tentatively match Chen with Mercy. Mercy
has 1 unfilled position.
The process is now complete: each applicant has either been tentatively matched to the most preferred choice
possible, or all choices submitted by the applicant have been considered. Tentative matches are now final.
Results:
• City matched to applicants Anderson and Eastman.
• Mercy ranked only two applicants and was left with one unfilled position.
• General, which ranked four out of five applicants, filled all its positions.
Mercy City General
Another thought: HYPS could establish the “Premier League” for admissions purposes, and agree that students could apply to multiple schools EA but only one from the Premier League. Then they could dare all the ED schools to move to unrestricted EA but with a cap on the number of EA applications.
That might go a long way toward solving the overall problem, but I doubt Penn and Columbia would agree to it, because it would mean that they would no longer be able to use ED to narrow the distance between them and HYPS in terms of selectivity/yield while trying to leave Duke, Northwest, Tufts and Vandy behind. They’d be permanently in the group they’re desperately trying to stay/get out of.
It would also, in a funny way, look like HYP were formally admitting S to the club, which I’m not sure HYP want to do. Even though S is now clearly tops in selectivity/yield, it’s a different animal from HYP. Much younger, much more about sports and tech, much less about the arts and humanities (despite excellent resources there), and located in California, which has its own attractions and subjects S to much less local competition than the Ivies face. This all makes S a potent nationwide threat, particularly to H, which in many ways has the most to lose if S can make a credible claim to being #1 overall.