For the purpose of my post (#7), I took the athletic participantion figure directly off the website of a top 20 (USNWR) LAC, where it states, “In any given season, 35 to 40 percent of the student body participates in a varsity sport.”
In a natural distribution it is impossible to “select” a bottom quartile without shifting the demarcation points of the other quartiles. (Though I think I understand your point on how aspects of this could, in theory, be manipulated, in a sense, artificially.)
“Overall score” is simply a metric that relates more to absolute differences between schools than the less statistically relevant rank. If anything of value is being measured at all, then score is a more valuable indicator than rank.
“The thing I don’t want to see at Emory is an ‘underachieving’ student body that doesn’t garner the amount of recognition it does now but comes in with higher scores than ever. It isn’t good for the institution.” (#8)
This relates to the risk of an over-consideration of rankings and is a comment I concur with in its general viewpoint.