Surprise in Harvard ALDC admissions

It’s true that recruited athletes often get a pre-screen, which can act as a filter such that the recruited athletes who do apply are the ones who are extremely likely to be admitted. However, it’s also true that athletes who are admitted average substantially lower academic qualifications than non-athletes, including LDC hooked non-athletes.

This fits with AI index comments above. Under Ivy League athletic conference rules, athletes must have a AI index that averages no more than 1 SD less than the student body as a whole. AI is based on stats are well correlated with the academic rating that was posted earlier.

If athletes average 1 SD lower academic index than non-athletes, then under a normal distribution we expect that the top ~16% of athletes should receive the average academic rating for non-athletes or better. The actual numbers posted earlier in the thread were:

Average non-athlete = 2 academic, Top 25% of athletes receive or 2 or better.

We also expect that bottom ~16% of non-athletes should receive the average academic rating for athletes or worse. The actual numbers posted earlier in the thread were:

Average athlete = 3 academic, Bottom ~19% of non-athletes receive 3 or worse

This seems reasonably consistent with expectations, given the lack of precision with the 1-5 scale.

The 4 academic rating is also relevant to the discussion. A 4 academic rating is essentially an auto-reject for non-ALDC kids. The admit rates was 3 / 18162 during the Harvard lawsuit sample. ~0% of non-ALDC admits have this rating and less than 1% of ALDC admit have this rating. However, a far more substantial ~15% of athletes have a 4 or worse academic rating. Some athletes are admitted with stats in the near auto-reject range for non-ALDC. Harvard also admitted a 5 academic athlete most years, which is the lowest possible rating of applicants Harvard’s scale.

As mentioned above, I believe the specific distribution is largely driven by the Ivy League athletic conference rules of requiring athletes AI to be within 1 SD of the overall student body. Different colleges and different conferences have different policies about athlete admissions. For example, LACs playing in the DIII conference are likely to have a much larger portion of students being varsity athletes, but are likely to give a much smaller degree of boost on average, with a larger potion of athletes having similar admission stats to non-athletes. There are also DI schools that give far larger boost to athletes than the Ivy League conference, particularly for revenue-earning sports, like football and men’s basketball.

1 Like