<p>Cellardweller still hasn't addressed my earlier critiques of his analysis, which lead me to believe (a) he is considerably understating the number of "academically elite" kids -- not by a factor of five, certainly (100,000 vs. 20,000), but maybe by 50-100%, and (b) he is considerably overstating the number of kids accepted at the Ivies from that group. So, although I don't disagree fundamentally with the conclusions of his analysis -- and, indeed, I have advised my own kids and those of our friends accordingly -- I think it is meaningfully over-sanguine.</p>
<p>The main issues:</p>
<p>On the number of "academically elite" kids:</p>
<p>-- There may be only 20,000 NMFs, but NMFs are determined on the basis of a single test, on a definite date, and almost all of the colleges consider and report SAT scores based on either (a) the best individual component scores of multiple tests (the overwhelming majority rule) or (b) the best single test date out of what may be multiple tests. I have never seen anyone provide real data on what this means, but since it's a one-way ratchet, and since scores do improve some with repeated testing, I think it's clear that, even if the population were limited to the population that took the PSAT, the number of seniors with NMF-equivalent SAT scores as reported by the colleges is considerably greater than 20,000.</p>
<p>-- Also, because the practice and additional maturity actually mean something, some kids may just improve. One of my kids was an NMF and had 2300 SATs, taking them twice. The other wasn't even close to being an NMF and had 2340 SATs, taking them once. And one of my daughter's impressive friends, now at Harvard, didn't come close to NMF status because, after only two years in this country, her English wasn't good enough in the fall of 11th grade. By the fall of 12th grade, her English was good enough for NMF-equivalent SATs.</p>
<p>-- I don't think taking the PSAT is universal at all. The pool of applicants is not limited to kids who were part of the pool that produced 20,000 NMFs. This absolutely includes international kids, who represent as much as 10% or more of some elite admissions.</p>
<p>-- Some kids do better on the ACT than the SAT. Colleges will report the SAT-equivalence of their ACT scores. Presumably, their PSAT scores were not as high as if they had taken a "PACT". This is just another ratchet that means that the single-test data reported by the testing services, or represented by the NMF population, understates the number of kids whose applications have those numbers.</p>
<p>-- Going the other direction, some significant number of kids with those stats take themselves out of the Ivy applicant pool voluntarily.</p>
<p>On the admission side:</p>
<p>-- Only half of the enrolled classes at HYPS etc. -- less, actually -- has NMF-equivalent SAT scores. Since the number of kids with NMF-quality SAT scores is clearly big enough to fill more spaces, that means that the colleges are accepting about half their classes (maybe a little less) from pools of kids without those score levels. </p>
<p>-- Some kids clearly get multiple admissions to elite colleges. So there aren't 16,000 kids being admitted to Ivies; it's something meaningfully less than that.</p>
<p>So, the bottom line is that there may be 30,000 "academically elite" kids, and at the Ivies they are chasing maybe 6-7,000 admission slots. I rate the chances there at 20-25%, not 50%. However, if you add Stanford, MIT, SWAMP, Chicago, Duke, etc., I think you move into 50%+ territory. </p>
<p>In other words, cellardweller's basic point -- that there is a home at elite institutions for the population of elite students -- is basically right. But you have to define "elite institutions" much, much more broadly than "the Ivies" to make it right. (It really helps if you toss Berkeley, Michigan, UVa, etc. in there, too.) And guess what? That's what the world actually looks like, as far as I can tell.</p>