There is no way more than 1 out of 100 arbitrarily sampled humans can win Putnam or get tenure in pure math/physics/etc. at an Ivy.