I never got anywhere trying to get into Princeton as an undergraduate, with average SAT scores over 700, but did get accepted in the math (or statistics) department as a graduate student. I'd read lots of John Tukey's work -- uniformity in topology (in the back of Kelley's 'General Topology'), Tukey's lemma in axiomatic set theory, stepwise regression, power spectral estimation (Blackman and Tukey), the FFT (Cooley and Tukey), and 'Exploratory Data Analysis', had been working in applied math and computing, wrote G. Watson, and got accepted. I also got accepted in the Division of Applied Math at Brown and in Operations Research at Cornell. I went to Hopkins.