Rating the Rankings, at Science News: by weighting the different factors in the US News and World Report college rankings differently, different schools can come out on top, and the rankings in general are pretty strongly perturbed, say Lior Pachter and Peter Huggins. The original paper is (I think) "Selecting universities: personal preference and rankings", arxiv:0805.1026. Basically, some schools do some things better and some schools do other things better.
It would be interesting to see this implemented as a sort of "make your own rankings" thing: a student could say how much they cared about various factors, and it wouldn't be hard to produce a ranking of schools according to which they would most like to go to.
However, to do this right, you'd have to allow nonlinear weights; most students would probably like to go to a school where they are near-average in the student population. Ranking schools according to their suitability for a given student is not the problem these rankings attempt to solve. And then the average 17-year-old doesn't really now what they want out of a college anyway... so maybe it isn't so easy. That's what guidance counselors are for, I suppose.
I've suspected that the ratings depended heavily on the weighting scheme for a while. I've even heard it somewhat cynically suggested that the weights on different factors are adjusted basically to make the rankings "look right", i. e. so that the schools that "everybody knows" are the best come out on top.