Improving risk matrices: the advantages of logarithmically scaled axes

Abstract

Risk matrices are a common tool used throughout the public and private sector to assess and manage risk qualitatively. However, these matrices have well-documented shortcomings when used for either assessment or management that can be shown by assuming a quantitative scale for the likelihood and consequence axes. This article describes the construction of a logarithmically scaled risk assessment matrix which alleviates some of the limitations inherent in using linearly structured risk matrices. In particular, logarithmic risk matrices can better differentiate between hazards with a large dynamic range in risks and, when used in combination with a new categorization scheme, the categorization of risks is more straightforward. These properties are demonstrated using a hypothetical example. Finally, the defensibility of logarithmic matrices is examined in the context of previously proposed rules for developing risk matrices.