It occurs to me that some people create codes to maximize the entropy of a system in order to predict equilibrium concentrations. However, others minimize the Gibbs function. I understand the relationship, so why the two methods? Aren't these results always the same? Is there a practical reason for this?