After identifying the random variables, you will need to assign them to appropriate probability distributions.  There are two classes of distributions: parametric distributions and outcome table distributions, with each requiring different declarations. Declaring parametric distributions involves declaring a particular type of probability density function (e.g., normal) and its parameters (e.g., mean and standard deviation).  Outcome table distributions may be declared using scalar values or, assuming your model makes use of sets and attributes, they may also be declared very conveniently using attribute vectors and matrices. In either case, you will need to identify all possible outcomes and their respective probabilities.