The point was to show that major selection can make colleges that are often thought of as less selective become highly selective. It was an extreme example to show that extremes are possible, not that they are likely.
If you want a more common, less extreme example, then replace it with a Black, not low-SES, prospective CS major is accepted to Stanford, but rejected at Berkeley and UCLA. Admission to the UC system is more dependent on major selection selection than Stanford, and the UC system is legally forbidden from considering race in admission. Both factors can contribute to Stanford admits being rejected by UC system schools, even though Stanford is generally more selective than the UC system.
There are also countless other factors in admission that different schools treat differently, beyond just major and race; which can result in admission decisions not following general selectivity well.
Over the years, many, many posters on this site have advised applicants to not take college decisions personally. The reason is that college admissions decisions are based on a huge number of factors beyond the individual applicants perceived “pure” academic potential which I’d say comes down to GPA/Rigor+Standardized Tests+ Academic EC
The non-pure academic factors (in no particular order) include:
Yield protection
Legacy
$$$ which includes need aware schools, OOS and International students paying higher rates at state schools, donations, etc.
Personality (non-academic ECs, leadership, etc.)
Geographic diversity
Race, ethnicity and religion
Athletic recruitment
etc.
Additionally, all these factors play out differently depending on the overall selectivity at each individual school.
Wait list is a whole different thing and can have many reasons-including a simple show of support.
UMich is a “lesser ranked college?”
If a college thinks they’re being used as a safety, why waitlist at all? If all else fails and the kid does become available? They wouldn’t know that. That’s not functional for the college.
Rejection is “much more widespread” than families realize. If you don’t match, don’t expect them to go gaga. We’re talking top colleges accepting a kid others reject. If you don’t match, yield is the least of it. Unfortunately, they simply may not want you in their yield.
The first problem kids have, imo, is they simply do not show in their apps/supps what the target colleges want, all of it. That has zip to do with wanting the right kids to matriculate. You may not be what they consider one of the “right” kids.
So, what yield are you protecting, if you don’t like this applicant? The bigger issue is kids who are so solid, in ALL ways, (not just stats,) but clearly gunning for Harvard, Stanford or MIT. And you aren’t one of those colleges.
What do you mean by “predictable”? They certainly are not deterministic - have these 5 numbers and activities and you’re in, without them you’re not (with few exceptions). Nor are they randomized.
There is certainly a correlation. MIT publishes acceptance rates by standardized test score tiers, and they are quite consistent over the years. Given a certain set of stats and factors, a reasonable estimation can be made on your admission chances (see “chance me” threads).
Yes, “highly likely” students will be rejected. And “low probability” students will be admitted. But in aggregate, the former will be admitted at a higher rate than the latter.
Which can be overcome by a genuine display of demonstrated interest. There is a fine line between yield protection in the perjorative sense and selecting students to be a part of your community who have proven that they really want to be there. Michigan and Tulane have no shortage of high stat kids in their applicant pools, so why would they not select those who have taken the additional steps to emphasize their enthusiasm? Who is the better choice to enhance their environment between a kid with a 36 and 4.6 who rests on those laurels and makes no additional effort and a 34 and 4.2 kid who can barely contain her excitement to get started on campus?
People here imply that you’ll have a difficult time getting into one of these “yield protection” schools with high stats because they have some sort of inferiority complex and are immediately suspicious of someone with high stats deigning to apply to them. The reality is these schools have gotten pretty good at identifying applicants who have provided little indication that they will attend if accepted (whose fault is that?) and instead choose to admit those who have shown them some love. That would seem to be an effective way to build an engaged community.
I find admissions to be highly predictable most of the time when looking at schools in like batches. For any one school under the magnifying glass, that’s a whole other story. When it comes to highly selective schools, it’s also another story. The old statistics and probability rules do not hold when you are looking at chances for admissions to a single digit accept school. But for most schools, if you apply to a critical mass of them, cull the special factors, the math will work out
There are always outliers to the process. The UCs, in particular show puzzling results where kids who apply to all or most of them get unpredictable results. But, often when you look at the situation, it’s not all that puzzling
I spoke with a mom whose DD was supposed to be an example of this UC unpredictability. The story had grown to the point that it was becoming an example of UCs colluding or at least sharing admissions info so that a highly qualified kid gets rejected because the school knows that the kid will reject the school due to better prospects.
Kid involved rejected from UCSD while accepted to Berkeley. Also accepted to a couple of ivies and other highly selective schools. Certainly more selective than UCSD. So why?
Well, she had selected a Marine Biology Major at UCSD. It was the only outlier result of her admissions process but the one heralded everywhere.
These things do happen but not at the rate where it’s a crap shoot.
Even in the highest score range the acceptance rate is still under 10%
ACT Composite 34-36
Applied: 5,848
Admitted: 569
When >9 out of every 10 of the highest scoring students are rejected, it is kind of hard to consider anyone “highly likely” (other than the usual suspects: legacy, donor, athletic recruit, etc.) at the most selective schools.
“Just want to point out that yield protection is real for some “lesser” ranked colleges, University of Michigan and Tulane come to mind. They waitlist lots of top stat students who they think are using them as a safety.”
Vanderbilt also uses the waitlist similar to UM or Tulane, but more from the standpoint of not being the first choice college as opposed to safety. The difference is that Michigan is transparent about it on their CDS, while Vanderbilt is not.
@cptofthehouse makes a great point re schools in like batches. This was the case with S a few yrs ago. He settled in on a highly selective group. What was important was to not get overly connected to one vs. the others as, individually they are unpredictable. But together they were quite predictable in terms of getting in to at least one. In his case, that like batch consisted of 7 schools. He got in to 4 of them. Too risky to gun for any specific school but very reasonable to expect to win within the cohort.
FYI - many of his friends applied to several of the same peer schools (listed as common or cross application via Fiske) and it’s interesting to see who was accepted where. No real pattern. They all got into several but in very different combinations.
Also important is Fit. These schools all had a lot in common (not everything but enough). S was quite deliberate in creating a fit list. That’s important because rejection from top choice (which happened) wasn’t a crushing blow. Wound up at No. 2 and would likely reverse them with two+ yrs under his belt.
That only indicates that you cannot predict admission to MIT well by looking at test scores alone, which is expected based on how MIT describes there admission process. It does not indicate admission is unpredictable. With enough information, I’d expect MIT admissions to be largely predictable.
True, but not really the topic here. It’s certainly possible to state, in aggregate, the likelihood of being admitted based on a set of test scores. If you have 750+ on test scores, you have about a 10% chance, in lieu of additional information. If you have 600-650 scores, your chances are dramatically lower.
This data is consistent and “predictable”. Acceptance among these two cohorts are not random, unpredictable and equivalent, back to OP’s question of “unpredictable”.
OP. Very very very few students are accepted to multiple Ivy League schools. There’s always someone who will post about “the kid they know” etc as though it is statistically probative. It’s not.
The fact is, after lots of students are moved to a pile of prescreened preferences, a school like Harvard has really less than 400 spots per gender for each class.
There are over 3mm high school graduates per year. 72000 valedictorian and salutatorians alone each year. As a reference point only. Where does everyone fit? Literally ?
This doesn’t include the international cohort.
It also doesn’t account for all of the preference categories that schools utilize to create a diverse environment, protect its rankings, alumni support, its brand and monetary considerations.
It’s a very competitive process at the very top. Think 100 schools not just the top ten. So if you fit the right profile, have a sympathetic and supportive admissions rep having a good day and are able to afford it in the end you may get the news you hope to receive.
The good news is that there are literally 100s of great, high quality schools to help you reach your goals.
Be strategic about the process. Look at it with open eyes and expectations. There is solid information relative to whether or not you are competitive statistically. Choose several you can afford, a few that feel like a stretch, a few that seem to admits students very similar to your profile and a couple where you feel 75-100 percent sure to get in.
UC needs to be more transparent about showing frosh admission stats by division or major where that affects selectivity. It does for transfer admissions as shown at https://www.universityofcalifornia.edu/infocenter/transfers-major , so it should not be all that difficult to show something similar for frosh admissions.
MIT does not have legacy preference, unlike most other super-selective schools.
Only an insider or someone well connected to the colleges in question (e.g. some prep school college counselors) would be able to estimate an applicant’s chance as possibly being “highly likely”, and then only after seeing the full application (not just test scores).
That’s what people on this site would like to think about all colleges, not just MIT. “If only we knew the secret sauce recipe…”
At Harvard, we do know the secret sauce recipe (thanks to the discrimination lawsuit) and yet there are still going to be tens of thousands of disappointed kids and parent this admissions cycle.
I’m more than a bit skeptical whether prep school college counselors are as connected as they are made out to be. I suspect that if you stripped out legacy (not at MIT, thanks for pointing that out.), donor, powerful celebrity, recruited athlete and the occasional URM from Prep school applicant results, the prep school counselors would be close to as clueless as everyone else when it comes to predicting results at the most highly selective schools for the average “unconnected” applicant from among their students.
The Harvard lawsuit is a good example of showing admission decisions to be largely predictable. Both sides of the lawsuit created models that explained the majority of variance in admissions decisions. They could correctly predict who would be admitted and rejected in the vast majority of cases, even though modeled information about the applicants was severely lacking and far less than a typical reader would consider. The admissions decisions did not appear to be largely random or whimsical. Instead there was a clear process laid out, and clear explanation for why certain applicants were accepted or rejected. There are elements of the process than are beyond a particular student’s control and elements that may vary from year to year depending on the particular applicant pool and particular demands. However, admission decisions were still largely predictable for typical unhooked applicants, with enough information.
Other insider type publications describe similar systems in which unhooked applicants who get top ratings in various combinations of both academic and non-academic categories are almost certain to be accepted, while other high scoring applicants who are lacking in various areas are almost certain to be rejected. I once tried to see if I could admission decisions among Stanford decision thread posters on this site, with the limited information in the post. As I recall, I could predict the overwhelming majority correctly by placing a strong emphasis on out of classroom activities, rather than the typical forum focus on scores and stats. Admissions decisions also appear to be largely predictable among students I’ve interviewed. I see no reason to assume MIT is different and everyone is “clueless when it comes to predicting results at the most highly selective schools for the average “unconnected” applicant…” because admissions decisions cannot be predicted well from test scores alone.
@Data10 I’m not sure whether your ability to predict whether students will be accepted to Stanford will be as useful outside the CC community. It may require modifications, and may be a lot less successful. It will still likely do better than most of the “Chances Calculators” being used out there.
Regarding less prestigious colleges rejecting applicants who were accepted by more prestigious colleges. The problem here is that people forget that many “less prestigious” colleges are also very selective, also accept holistically, and, therefore, are also rejecting many kids with top-notch profiles.
Yes, a college with a 16% acceptance rate may be accepting 60% of the top students who apply, versus 15% acceptance of these type of students by a college with lower acceptance rates. However, what that also means is that a good number of those 40% who were rejected could be in the 15% who were accepted by the more selective school.
Of course, what is a hook at one college may be less of one at another, and legacy preference is a big one there. A legacy at a very selective school may have a more difficult time being accepted at a less selective school where they aren’t a legacy.
Then there are other preferences. A kid with multiple math awards may have the advantage at CalTech or MIT, but may be pushed aside in favor of the kid with the national attention, due the their political activism, at a small liberal art college.
The applicant has one of the most impressive resumes. We all could have predicted that he would probably be accepted at a few of the most selective colleges, but which ones? No one could have at the outset. College admissions at “holistic” colleges are simply predictably unpredictable. If you claim you could, you’re either delusional or you possess some mythical power beyond math, certainly beyond some simplistic regression/correlation models as used in the Harvard lawsuit, that the rest of us don’t have.