Fuzzy numbers cannot be easily ordered as ordinary real numbers. Several proposals have addressed this problem, each with some drawbacks and limitations. This paper renounces to the idea of finding a universal ordering method for fuzzy numbers and suggests to compute the degree by which an arbitrary permutation of fuzzy numbers is ordered. Thus, ordering becomes itself fuzzy. This is useful in the development of decision support systems as a higher degree of accurateness is achieved, with respect to existing methods. However, accuracy is counterbalanced by a higher computation time. The paper address this issue, by proposing an efficient numeric solution.