Monte Carlo Simulation is a statistical method for iteratively evaluating the cost impacts of risks given a three point estimate to highlight the uncertainty of the cost impact. A risk is defined with an Optimistic, Most Likely and Pessimistic cost that depicts the three possible cost impacts of a risk should it occur.

A simulation may span several thousand iterations or more. During a simulation, random numbers are generated from a probability distribution whose standard deviation is limited by the three point figures used to depict potential cost impacts. Each risk is iterated using it’s own distribution.

The probability distributions may be replaced with other distribution types (such as uniform or triangular) however the most common distribution type used is the Normal/Gaussian distribution whose characteristics include a bell shaped curve.

In some cases, team members enter their own probabilities against risks (called subjective probabilities) which are then used to switch a risk on or off depending on the value assigned to the risk and the random number derived. This limits the possible cost impact that is extracted by limiting the standard deviation of the distribution.

In the context of risk simulations, one would select a confidence interval (CI) which depicts the total uncertainty of the collection of risk. A value of 80% (referred to as P80) is a standard selection by many organisations to represent a scenario where there is a high level of uncertainty. Selecting a value of P80 suggests that you are 80% confident that should all risks eventuate, the total dollar impact to your project is the value that intersects that confidence interval.

Part Two: Requirements for Monte Carlo Simulations