Project Monte Carlo Risk
Use Monte Carlo simulation to understand the range of possible outcomes for your project timeline and budget. Get probability-based estimates using three-point estimation and discrete risk events.
Simulate Project Risk
Use Monte Carlo simulation to understand the range of possible outcomes for your project timeline and budget. Get probability-based estimates instead of single-point guesses.
Quick Start:
- Add your project tasks with three-point estimates
- Optionally add risk events with probabilities
- Set target deadline and budget (optional)
- Run the simulation to see probability distributions
What are three-point estimates?
For each task, estimate the optimistic (best case), most likely, and pessimistic (worst case) duration and cost. The simulation samples from these ranges.
Start by adding tasks above
Understanding Project Monte Carlo Risk Analysis
What is Monte Carlo Simulation for Projects?
Monte Carlo simulation is a technique that uses random sampling to understand the range of possible outcomes for a project. Instead of giving a single estimate like “the project will take 45 days,” it provides a probability distribution showing the likelihood of different outcomes.
The simulation runs thousands of hypothetical project scenarios, each time randomly sampling task durations and costs from their estimated ranges. By aggregating these scenarios, we can answer questions like “What is the probability of finishing by our deadline?”
Three-Point Estimates and the Triangular Distribution
For each task, you provide three estimates:
- Optimistic (a): The best-case scenario if everything goes well
- Most Likely (m): The most probable outcome under normal conditions
- Pessimistic (b): The worst-case scenario if problems occur
This tool samples from a triangular distribution using these three points. The triangular distribution is commonly used in project management because it is intuitive and captures asymmetric uncertainty (tasks often have more upside risk than downside).
Understanding P50, P80, and P90 Estimates
Percentile estimates tell you how confident you can be:
- P50: 50% of simulations finished by this duration/cost
- P80: 80% of simulations—a more conservative estimate
- P90: 90% of simulations—high confidence but may include contingency
Using P80 or P90 estimates for planning provides a buffer against uncertainty. If you commit to a P50 estimate, you have only a 50% chance of meeting it. Many organizations use P80 for planning and P90 for contracts or commitments.
Discrete Risk Events
Beyond task uncertainty, projects face discrete risks—events that may or may not occur. Examples include:
- Key personnel becoming unavailable (probability: 20%)
- Scope changes requiring rework (probability: 30%)
- Vendor delays (probability: 25%)
- Technical issues requiring redesign (probability: 15%)
Each risk has a probability of occurring and an impact (additional duration and/or cost) if it does. The simulation randomly determines whether each risk occurs in each run, incorporating their potential effects into the overall distribution.
PERT Expected Value vs Monte Carlo
The traditional PERT formula calculates a single expected value:
While useful, PERT gives only one number and doesn't show the range of possibilities. Monte Carlo simulation provides the full distribution, helping you understand not just the expected value but also the uncertainty and tail risks.
Duration-Cost Correlation
The correlation coefficient (r) between duration and cost tells you how strongly they move together:
- r close to 1: Strong positive correlation—longer projects cost more
- r close to 0: Little relationship between duration and cost
- r close to -1: Negative correlation (rare in projects)
Limitations of This Model
This tool uses a simplified model with important limitations:
- Sequential tasks: Assumes all tasks are on the critical path (no parallel execution)
- Independent tasks: Doesn't model correlations between task durations
- No resource constraints: Doesn't consider limited resources or resource leveling
- Triangular distribution: May not fit all task types (log-normal or beta might be better for some)
- Static risks: Risk probabilities don't change based on project state
Practical Tips for Better Estimates
- Use historical data: Base estimates on similar past tasks when possible
- Consider multiple estimators: Get input from different team members to reduce bias
- Don't anchor on the most likely: Ensure optimistic and pessimistic estimates are genuinely different
- Include all work: Don't forget testing, reviews, meetings, and ramp-up time
- Revisit estimates: Update the simulation as you learn more during the project
Frequently Asked Questions
For most purposes, 5,000 simulations provides a good balance of accuracy and speed. The percentile estimates (P50, P80, P90) stabilize around 1,000-2,000 simulations.
Use 10,000-20,000 simulations if you need very precise tail estimates or are presenting to stakeholders who expect high precision. For quick explorations while adjusting inputs, 1,000 simulations is fine.
Related Tools
Monte Carlo Simulator
General-purpose Monte Carlo simulation for analyzing uncertainty in any quantitative model.
ROI, NPV & IRR Calculator
Evaluate the financial return of your project using time value of money analysis.
Safety Stock Calculator
Calculate inventory safety stock levels using statistical methods for supply chain planning.
Queueing Theory Calculator
Model waiting times and service capacity to understand resource allocation tradeoffs.
Subscription Cohort Revenue Decay
Visualize how subscription revenue decays across cohorts over time and project future revenue.
Franchise Location Profitability Model
Model profitability for franchise or branch locations with revenue and cost projections.
Explore More Operations & Planning Tools
Build essential skills in operations research, project management, and data-driven decision making
Explore All Data Science & Operations Tools