A common feature of cyber security and resiliency assessment methodologies is to elicit semi-quantitative information from subject matter experts (SMEs). This information is frequently based on expert knowledge of the capabilities and motivations of hybrid threats. SMEs typically provide ratings of various system aspects on an ordinal (e.g., 1–5) scale which is then aggregated to create a prioritized rank order. Crucial system information may be hidden or lost during such assessments. Here we present an approach which is cognizant of multiple sources of complexity that exist in SME-driven cyber resiliency assessment methodologies.