That's the problem with counterexample proofs, though. If something is only true for a large-but-finite range of numbers, there might not be a satisfying reason why.