Compensatory decision methodologies, such as AHP are the most effective way to prioritize alternatives. A rational decision is one that best achieves an individual's or group's objectives -- their wants. However, there are sometimes constraints, perhaps legal, that rule out certain alternatives. We say that these alternatives are infeasible and use "musts" to eliminate them from consideration before doing a prioritization or making a choice.
Musts can also be used to narrow down the number of alternatives being considered in order to shorten the evaluation process. For example, we might restrict our evaluation of stocks to those that are on a major exchange, have a price to earnings ratio below a specified number, and are recommended by a favored source of information. However, musts should be used somewhat sparingly because once an alternative is eliminated because it does not satisfy one or more musts, it cannot be selected no matter how desirable it is with respect to other objectives.
In addition, the specification of musts is often somewhat arbitrary. When selecting a car for example, we might say that we will consider only alternatives that cost less than $50K. This would rule out a car that costs $50,001 and might be more preferable than a car that costs $49,000 on all other objectives.
Musts can sometimes be wants in disguise, or can be both a must and a want. Musts are often useful in doing source selection, where a list of mandatory requirements is often given. After eliminating alternatives that do not meet one or more musts, some of the musts might also be considered as "wants." For example, there might be a must that each alternative must provide a specific function, because that criteria might advantage or disadvantage an alternative.
