Monte Carlo Analysis - Simulation in Project Management

For those of you unfamiliar with Monte Carlo Analysis, it’s a type of simulation analysis used to determine the outcome of multiple variables through the use of random numbers and statistics. As you may have guessed from the name, Monte Carlo Analysis was first developed in the gambling capital of the world—Monte Carlo. The method was created as a way to help predict the odds of winning at roulette.

What is Monte Carlo Analysis?

Monte Carlo analysis is a simulation technique that can be used in many different fields. In project management, Monte Carlo analysis is most commonly used to determine how likely it is that an event or objective will happen by given dates. In other words a Monte Carlo simulation is a type of analysis that helps you to determine how your project is likely to perform under various circumstances.  It’s not intended to be an exact science, but it can help give you a better picture of what’s likely to happen and show you areas where your schedule may need some adjusting. 

Monte Carlo Analysis is usually used as a part of;

When I talk about Monte Carlo analysis, there are two main concepts I want to cover: simple simulations and more complex simulations.  Complex simulations go into much greater detail than simply looking at individual tasks and their durations. Unlike other simulation methods, Monte Carlo does not require you to make up values for everything, it requires that you only come up with expected values (since these are all variables).

If we were going to run a real-world Monte Carlo analysis on our schedule from before, we'd start by getting actual task duration estimates from everyone involved in creating them (or do research based on past projects if none of us have experience doing what we're planning). Then, each time someone completes his or her task, he or she would estimate when he thinks he'll complete it -- maybe 25% estimate that they'll finish early and 50% guess late. You'd then sum those individual estimates up and come up with one overall duration for each task based on its probability distributions.

Why use Monte Carlo Simulation?

Using Monte Carlo Simulation for forecasting is advantageous for several reasons. First, Monte Carlo simulation does not require a specific distribution and has no restrictions on assumptions about data or prior distributions; all that is required are probability distributions of each parameter. Second, it accounts for variability through repeated runs instead of single-run averages like other techniques such as Three Stage Least Squares. Finally, Monte Carlo simulation is fast compared to alternatives such as actual stock market simulations. 

To account for these benefits and more, we can apply Monte Carlo simulation techniques to project management applications in four primary areas: Estimating project parameters (e.g., cash flow timing), forecasting project completion date uncertainty, managing risk exposure (e.g., at-risk value estimation), and designing risk strategies (e.g., value at risk). Understanding how one technique can be applied across different problem domains shows how versatile Monte Carlo analysis can be even when dealing with various applications involving inherently subjective factors like decision-making under uncertainty.

How to Use Monte Carlo Simulation?

Using a Monte Carlo simulation, you can identify specific risks or areas where performance could be improved. Then, after conducting a Monte Carlo simulation, you can adjust your project plan and budget accordingly. The following steps will guide you through an example of how to conduct a Monte Carlo simulation for your own project

  • Define variables; 
  • Develop probability distributions; 
  • Identify risk factors; 
  • Conduct a sensitivity analysis; and 
  • Discuss results with stakeholders. 

These are just some of the many possible applications for Monte Carlo simulation in project management. With Monte Carlo simulations, you have full control over what scenarios to test and how many iterations to run before drawing conclusions from your data. This means you have greater insight into both positive and negative outcomes over time as opposed to typical financial analyses that only consider average outcomes.

See also:

Probability and Impact Matrix

Multicriteria Decision Analysis

Decision Tree Analysis

Regression Analysis

What-If Scenario Analysis

Regression Analysis