The present disclosure relates to a method to determine, from a measured result and an assumed adjustment set, the probability that a permutation to the assumed adjustment set was implemented or other error, such as a sensor error, is occurring. The determination of probability of a permutation of the adjustment set may be based on the error between the predicted response and the measured actual response for the assumed adjustment set and permutations of the assumed adjustment set. A number of statistical comparisons may be performed between the assumed adjustment set and the permutation sets to determine which are closer to the measured actual response. A probability value may be assigned to each adjustment set as to the likelihood of their having been implemented.