<p>Moosemaster, There is actually an empirical basis for some of what you say in your first post. A revealed preference study conducted, I believe, by a couple of Wharton professors a few years ago concluded among other things that some schools including Yale, Princeton (the worst offender), and Stanford (but not MIT) were gaming the USNWR rankings by lowering the admit rate for students just below the very top tier. The speculation was that they were concerned too many of those students would attend Harvard, reducing the matriculation statistics to an unacceptable level. Interestingly, this phenomenon didn’t affect students at the very top, presumably because the schools had to get some of those super elite students to maintain their own super-elite status and simply had to take the risk of competing for those students head to head against Harvard. It also didn’t affect students another notch below, presumably because those students were less attractive to Harvard in the first place. BTW, that analysis also determined that 2/3 (IIR) of all Yale/Harvard cross admits choose Harvard (and notably Yale fared signficantly better than any other school against Harvard, including Princeton). I’m not sure that phenomenon still exists, though. In fact, I think SCEA may be an alternate route to achieving the same end. I think the idea in SCEA is to admit (among the unhooked) primarily students who are statistically likely to also be admitted to Harvard – with the twin ideas that (1) the school will have a few months to woo those students before the students even know whether Harvard will take them and 2) because the students have self-selected Yale, there is a relatively higher likelihood that they will prefer Yale to Harvard (as compared with students who have not so self-selected) irrespective of the wooing process. The first of these two ideas potentially explains the “single choice” feature of SCEA. At any rate, this explanation is consistent with Silverturtle’s observed phenomenon that very few CC’er below 2300 or 3.9 are admitted SCEA since only those students are likely candidates for Harvard. And, while it’s almost certainly true that CC’ers have higher stats than nonCC’ers, even within that relatively select group a cutoff of sorts seems to apply. Of course, all bets are off in RD – and I’ve never seen an analysis of those outcomes.</p>