Criteria for evaluating fuzzy ranking methods