Maxing Out Entropy: A Conditioning Approach