The bias in multiplication comes from losing information of the single factors. For example, a value of 12 can root from a multiplication of the factors 2 and 6, or 6 and 2, or 3 and 4 or 4 and 3. And here is the first step towards an unbiased risk level matrix. Thus, a severity of 6 might have more impact on an organization than a severity of 2. Likewise, the occurrence rates of 2 and 6 cipher for different needed measures to reduce the risk.