Monte Carlo Simulation Technique
If we could perform a process a large number of times, we could build up a histogram to represent the probability distribution from the individual outcomes of each trial.
Of course, we cannot really do this with a project. In fact, we cannot even perform it once without defeating a principal purpose of risk analysis (namely, to decide whether or not the project is feasible in the first place).
We can, however, pretend to do so using Monte Carlo simulation (named for the casinos that make that European city famous). Instead of doing time analysis once based on fixed duration lengths for each activity, we simulate the project several hundred or several thousand times. Each time, we use a different set of data for durations sampled from the probability distributions of these uncertain values.
The sampling is done so that the probability of selecting a particular duration in the simulation is the same as our subjective estimate of the probability of that value actually occurring. With definite values for activity duration thus supplied, we can calculate such things as project completion date for each trial.
As enough simulations are performed, a definite picture of the risks inherent in the project begins to emerge.