Computer software can be used to simulate a wide variety of real-life phenomena. Monte Carlo Simulation in particular gets its name from the famous casino and is used when the purpose is to quantify the effects of uncertainty.
Typically there will be a number of uncertain inputs, modeled by probability distributions supplied by the user, and a number of outputs which depend on these inputs. Monte Carlo Simulation proceeds by sampling from the input distributions multiple times, calculating the outputs based on each set of samples, and accumulating histograms and other statistics representing the distributions of the outputs. Each set of samples and the accompanying calculations is called a trial, or sometimes an iteration.
Full Monte does Monte Carlo Simulation in the project scheduling context, where the inputs are generally task durations and outputs of interest typically include the project cost and completion date and the dates of important milestones.
Full Monte accumulates data for costs and for all the dates and floats which would normally be calculated by a deterministic critical path analysis. Unlike some other systems Full Monte does this for every task in the schedule, presenting results in the form of histograms and s-curves.
Monte Carlo Simulation is necessary because the durations in a typical project network combine in ways – sometimes in parallel and sometimes in series – which are too complex for the required statistics to be computed analytically. (The technique known as PERT tries to do this but simplifies the problem by considering only one path, which can lead to very misleading results.)
To get reliable results from a Monte Carlo Simulation it is important to do an adequate number of trials, which usually means many thousands, and this can be time-consuming. Full Monte is fast – over 100 times faster than competitors – making it much more practical to do a thorough analysis. If you would like to learn more, please call us at 281-971-9825.