Thanks for pulling this thread, I hadn’t seen it. By MIT Chris not only commenting on that thread, but inferring how significant these awards are it’s basically the mic drop on this issue.
2 Likes
If I am to summarize correctly, winning international math, science, engineering etc. competitions is the “straight ahead” or “direct” way of gaining admissions to MIT rather than the often cited “sideways” admission. My guess is that the sideways admission odds fall into @hebegebe’s 3% bucket.
1 Like
That’s fundamentally it. But because MIT doesn’t consider legacy, their hooked pool is smaller, and their “sideways” pool is larger and competing for more remaining spots.
7.5% of 20000 is 1500 
And a yield of about 67% results in 1000 students to fill the class.
1 Like
Ah, I see, thanks. I think it makes sense to add that to the original post.
Modified as suggested for clarity.
1 Like
It sounds like you are guessing about the percentages. However, we don’t need to guess for Harvard since the numbers are listed in the lawsuit. Harvard has 4 core ratings categories – athletic, EC, personal, and athletic. A 1 rating is the highest possible rating in these categories. A 1 EC generally corresponds to the types of national level ECs we have been discussing . Some may also fall in to the 1 academic category, if supported by a similarly strong academic profile, including in the classrooom. A 1 athletic generally refers to recruited athletes. The specific percentages of applicants and admit rate for these ratings are below:
Only 0.3% of applicants and 2% of admits had the maximum 1 EC rating, corresponding to national level ECs (not just awards/competitions). While this group had a relatively high admit rate, they certainly were not the typical admits. Instead a much larger 35% of admits were more well rounded kids who had three 2 ratings, one 3/4 rating, and no 1 ratings. This group of well rounded three 2 rating kids had a 43% admit rate, which was quite similar to the 48% admit rate for 1 EC kids. I don’t doubt that the few IMO winners would have a higher admit rate than the 48% overall for 1 EC or 68% for 1 academic, but applicants with IMO level awards are not the typical applicants or admits. They are extreme rare outliers that are not particularly relevant for typical applicants.
Percent of Harvard Applicants Receiving Ratings and Corresponding Admit Rate
1 Academic – 0.4% of applicants, 68% admit rate, 4% of admits
1 EC – 0.3% of applicants, 48% admit rate, 2% of admits
1 Personal – <0.1% of applicants, 66% admit rate, <1% of admits
1 Athletic – 0.9% of applicants, 88% admit rate, 10.5% of admitsTwo 2 ratings, + two 3 ratings – NA% of applicants, NA% admit rate, 31% of admits
Three 2 ratings, + one 3/4 rating – 6% of applicants, 43% admit rate, 35% of admits
Four 2 ratings – 0.4% of applicants, 68% admit rate, 4% of admitsZero 1-2 ratings – 37% of applicants, 0.1% admit rate, <1% of admits
The lawsuit also includes a regression analysis that reviews the calculated change in chance of admission if a single rating is changed and everything else remains the same – same other ratings, same hook status, same planned concentration, etc. A summary of change in chance of admission (odds ratio) compared to a reference 3 is below. A 1 EC rating indeed had a high 67x increased chance of admission compared to a reference 3. However, the far more common combination of 2 academic, 2 EC, and 2 personal had a much higher 6x7x11 = 462x increased chance of admission. The latter group are typical admits, not the 1 EC kids.
Harvard Lawsuit Regression Analysis – Full Controls, Baseline (unhooked) Sample
1 Academic – 160x increased chance of admission
1 EC – 67x increased chance of admission
1 Personal – 37x increased chance of admission
1 Athletic – >2000x increased chance of admission2 Academic – 6x increased chance of admission
2 EC – 7x increased chance of admission
2 Personal – 11x increased chance of admission
2 Athletic – 4x increased chance of admission
Harvard’s rep also did an analysis reviewing how large a portion of decisions were explained by the ratings in isolation, as well as how large a portion of decisions would change with ratings removed. His results are below. EC ratings in isolation only explained 9% of variance in Harvard admission decisions compared to 20% for personal. Removing EC ratings only decreased variance explained by 9 percentage points . While removing personal decreased variance explained by 12 percentage points. Again it doesn’t point to national level ECs being important for typical admits, nor does it suggest ECs as a whole are more influential for than character/personal qualities for typical applicants.
Variance Explained in Admission Decisions by Harvard Rating Categories
Personal Rating Alone – Explains 20% of variance in decisions
School Support Ratings Alone (LORs) – Explains 19% of variance in decisions
Interview Ratings Alone – Explains 13% of variance in decisions
Academic Rating Alone – Explains 9% of variance in decisions
EC Rating Alone – Explains 9% of variance in decisions
Athletic Rating Alone – Explains 8% of variance in decisionsFull Model – Explains 64% of variance in decisions
Full Model without EC Rating – Explains 55% of variance
Full Model without Academic Rating – Explains 53% of variance
Full Model without Personal Rating – Explains 52% of variance
Full Model without Teacher/Alumni LOR Ratings – Explains 32% of variance
2 Likes
I don’t understand why you are posting this analysis about Harvard in response to hebegebe’s cited post citing only MIT? MIT and Harvard admissions processes are quite different.
Regardless, one important fact that hasn’t been discussed is we don’t know whether Harvard admissions considers these awards as part of the academic rating or ECs. Why are you assuming these competitions and awards only influence the EC score?
Those applicants who advance/place/win these types of awards should enter this info in the HONORS section of the Common App, which by definition is academic awards only (no sports MVP and such)…my point being is that these USAMO et al awards could very well influence the academic scores that Harvard AOs assign the applicant.
1 Like
The post explains this. We are guessing numbers for MIT, but don’t need to guess for Harvard. The comments that generated this tangent did not imply that MIT was the only college that claims personal/character measures are more influential than ECs for typical applicants. In fact, that post mentioned Harvard specifically as an example.
I’m not. The post you replied to states, “Some may also fall in to the 1 academic category, if supported by a similarly strong academic profile, including in the classrooom.”
The post you replied explicitly discusses this and states something similar. The point is IMO level awards are a minuscule portion of admits. Even the combination of 1 academic + 1 EC was only 6% of admits. Those 6% of admits are not particularly relevant for the typical applicants and admits. The other 94% of admits without 1 academic/ECs are far more relevant.
3 Likes
Even using Harvard, that would mean 75 kids in the admitted class had academic ratings of 1, likely national award winners or finalists in similar contests
While IMO level kids may receive a 1 academic, there are many, many other possible reasons for a 1 academic rating besides national awards/contests. For example, the lawsuit mentions 1 academic kids often have near perfect stats + academic works reviewed by faculty in the relevant department.
3 Likes
Do you know that Harvard’s application review and admissions rating process is the same as it was 7+ years ago? Without that certainty these numbers lack current relevance.
1 Like
Thousands of people have perfect stats now. That means little. Published work which is reviewed by known academics probably qualified if it gets cited sufficiently
1 Like
By, “near perfect stats + academic works reviewed by faculty in the relevant department”, I mean the combination – not either in isolation.
In my best Jan Brady voice:
(Quote has been heavily modified) 
4 Likes
We don’t have ANY data from ANY years for MIT. The Harvard data has to have a little more relevance than that, right?
2 Likes
No problem! I can’t imagine what is going through the Class of 2022’s heads after this unprecedented year in college admissions. Hope this thread offers some guidance and comfort, and best of luck to your daughter this coming admissions cycle!
1 Like
I agree with your point, but we do have some college matriculation data for some years of these top award winners.
My point on the relevance of the Harvard lawsuit data is broader than just this example…if their process is different (and spoiler alert…many schools’ admissions processes are different than they were 7+ years ago) these data don’t tell us as much as some posters infer.
Thank you @PerformingDude . going through marketing materials, info sessions, all the data out there has felt more like reading tea leaves / gazing into crystal ball at times . One major take away I got from the discussions so far is that 1)safeties are very key 2) the mix of safeties/matches/reaches is key ( going with 2:6:2). In fact we are starting our campus visits ( in person) with safety colleges and moving to high/low matches etc. Really appreciate all the input from the forum.
3 Likes