A Simple Proof of Optimal Approximations