You are a project manager and have to make decisions on the fly. And this can be particularly difficult when estimating the cost and the timeline of complex, volatile projects. In this article, I will share what you need to know about Monte Carlo risk analysis, how it works, and an overview of when it should be used.
Whenever we manage projects, we rely heavily on assumptions. Assumptions typically have some level of uncertainty and risk. For example, we may assume that resources will be available in the future for certain project activities.
Project sponsors and other leaders want to know when a project will be completed. Some project managers make the mistake of giving a discrete date. Estimating just one value/date for a complex, risky project is not a good idea and can lead to significant issues.
One way to manage expectations better is to run a Monte Carlo simulation resulting in a range of probable outcomes. Project managers can illustrate what may happen (e.g., timeline and cost) and the probability that it will happen.
What is Monte Carlo Analysis?
The Monte Carlo Analysis is a powerful mathematical tool used to analyze complex projects quantitatively. Project managers may use it to simulate project tasks and estimate the risks associated with various outcomes.
The simulation is a way of predicting how things might turn out in the future. Imagine you are playing a game like Monopoly. To get ready for the game, you can think about different ways the game could go and what might happen if each of these happens. That's like Monte Carlo Risk Analysis! It helps us figure out what might happen based on different scenarios. It's like taking an educated guess about what could happen in the future.
When Should It Be Used?
For many projects, a project manager may perform qualitative risk analysis only. It’s a quick way to prioritize risks. However, it’s subjective.
Sometimes, we need more detailed, numeric analysis. There may be several significant risks that make it difficult to forecast the project. If these risks occur, they will impact the duration and cost of the activities. This is when we use tools like the Monte Carlo simulation.
The most important factor when deciding if Monte Carlo analysis should be used is the level of uncertainty associated with a project. When there are uncertain estimates (e.g., timeline), then Monte Carlo simulations can reduce the level of uncertainty through a series of simulations, resulting in a continuous range of possibilities (e.g., there is a 50% chance of completing the project by June 1, a 70% chance of completing the project by July 18 ). Follow this same process to create a cost distribution curve.Three-Point Estimates
Perhaps, you are familiar with three-point estimating. Rather than asking for a single estimate from a subject matter expert, the project manager asks for three estimates of effort for a project activity:
- Best-Case Scenario Estimate
- Most-Likely Scenario Estimate
- Worst-Case Scenario Estimate
Estimate | Effort (Hours) |
---|---|
Best-Case Scenario | 40 |
Most-Likely Scenario | 55 |
Worst-Case Scenario | 65 |
Three-Point Estimate | 53.3 |
Weighted Three-Point Estimate | 54.2 |
Think of these estimates as a continuous range of possibilities rather than only three values. Wider ranges (i.e., standard deviations) have greater uncertainty.
When should a project manager ask for three estimates rather than one? If the subject matter expert can provide a single reliable estimate, that’s great. How is this possible? The expert has lots of experience with the activity. However, if the expert has little to no experience in performing the activity, the project manager may ask for the best-case, most likely, and worst-case estimates.
Monte Carlo Analysis and the Three-Point Estimates
The Monte Carlo Simulation combines probability theory, statistics, and computer software to simulate the outcome of a project. The schedule simulation uses the duration estimates and a network diagram. The cost analysis uses cost estimates.
- Assign a random value to a variable such as the effort for a best-case scenario activity.
- Perform the calculation with this random value.
- Save the result.
- Change the random value.
- Recalculate.
- Repeat until the analysis is complete.
The computer simulation is run hundreds or thousands of times, resulting in a probability distribution. Armed with the results of this analysis, a project manager can illustrate the continuous nature of timeline and cost risks. Rather than giving a single date and budget figure, the project manager can present a continuous curve with the likelihood of completing a project by a certain date and another curve for cost.
Conclusion: Key Concepts
In conclusion, project managers should determine if a project merits the effort to perform quantitative risk analysis using tools such as the Monte Carlo Analysis. The results may be used to make decisions such as go/no go decisions at key milestones.
At its core, the Monte Carlo risk analysis utilizes probability distributions and statistical sampling techniques to generate many scenarios that could arise from input parameters such as three-point estimates. Doing so can provide a better assessment of the project’s probable outcomes and an understanding of how certain risks may affect each outcome. Furthermore, this analysis considers both upside possibilities (opportunities) and downside risks (threats), thus providing a comprehensive view of the project's overall performance over time.