Criteriamost

Criteriamost is a metric used in multi-criteria decision analysis to identify which criteria tend to be most influential across different scenarios or weightings. The term combines “criterion” with “most” and is used in theoretical discussions and some applied contexts to summarize the robustness of criterion importance.

Formally, given a set of criteria C and a set of scenarios S, and for each scenario s a scoring function s_s(c) for each criterion c in C, criteriamost assigns to each c a value equal to the proportion of scenarios in which c attains the maximal score among all criteria. If multiple criteria tie for the maximum in a scenario, each tied criterion receives a fractional share of the scenario count. The criteriamost value ranges from 0 to 1, where higher values indicate criteria that are most frequently dominant across scenarios.

Computation typically involves normalizing scores across criteria for each scenario, identifying the top criterion or set of top criteria per scenario, tallying the occurrences, and converting counts to proportions. This yields a robust, scenario-aware picture of which criteria repeatedly drive decisions under varying weights or judgments.

Applications include comparing robustness of criteria when weights differ, informing policy prioritization, product design, or risk evaluation. Criteriamost helps communicate which criteria tend to dominate decisions regardless of specific weighting, supporting transparent decision rationale.

Limitations include dependence on the chosen scoring and scenario set, sensitivity to normalization methods, and potential ambiguity in ties. It does not convey the magnitude of lead beyond the top criterion and should be used alongside explicit sensitivity analyses.

See also: multi-criteria decision analysis, sensitivity analysis, robustness measures.