Two techniques to process conflicting subjective probabilities in individual decision making

Two techniques for resolving conflicts in elicited subjective probabilities for individual decision making are proposed. The first one adjusts the fuzzy sets that represent the subjective probabilities in order to eliminate the conflicting parts, whereas the second one accomplishes the same task by expanding those sets. While the first technique is proposed as a necessary pre-processing step when the Laplace criterion is applied, the second one extends the results from Wald and Maximax criteria and proposes even more extreme outcomes. The application of both techniques to a problem of betting on sport games with conflicting elicited probabilities shows that in contrast to the proposal, the conventional pessimistic and optimistic analysis do not cover all the possibilities implied by the decision maker uncertainty. These techniques are intended for applications where probabilities are subjectively elicited and therefore more likely to include conflicting values.