One limitation of MC simulation is that if it resamples the random numbers each time you run it you get variation in the results from the sampling - this is effectively noise in the results. The downside is that you can’t easily see the impact of changes in your plan on outcomes - is it just sampling variability or an actual difference. A simple way to get around this is to have the option of setting a random number seed - this way PL will consistently use the same set of random numbers for the MC simulation and the results will be stable unless some part of the plan is changed, making scenario analyses with the MC simulation more useful.