I was wondering if the fact you graduated a few years ago have not caused this kind of results. Using the mean-variance model described elsewhere, certainly you are an applicant with a very high 'expected value', but this time elapsed since your graduation may have increased your variance for most schools (not so much for MIT, since you studied there). Therefore, it is possible some schools have chosen students with a lower expected value, but with lower variance (from their point of view).
PS.: This is my first post, so hi to everyone.