The stakeholders involved in a flood risk management plan will need to generate an initial set of evaluation criteria which will later be refined during the MCA process. It is better to start with too many, being comprehensive, than too few.
Any criteria which do not discriminate between the options can be removed in the course of the analysis. In particular, it can be useful separate out ‘desirable’ and ‘undesirable’ consequences of the same type. For example, to have a separate criterion for positive environmental consequences and another for undesirable consequences. Particularly at the coast, there is often an environmental trade-off to be made in deciding which course of action to adopt. On the coast, managed retreat as a FRM measure can prevent the loss of coastal mudflats and wetlands but at the cost of the loss of the habitats behind the existing flood defence line.
The first risk from having many criterion is that of double-counting: more than one criterion measuring the same thing. The second is that humans are limited in the number of factors that we can consider simultaneously; in a classic paper G A Miller (1956) demonstrated that most people can cope with no more than seven. One purpose of the statistical tools used in MCA (see below) is to address these problems by eliminating criteria when one duplicates another.