Program Evaluation and Review Technique — Convert uncertain task durations into reliable metrics. Compute Expected Time (TE), Standard Deviation (SD), Variance, and project completion probability with interactive normal distribution visualization.
The Beta distribution PERT model was developed by the US Navy in 1958 for the Polaris project. It gives more weight to the most likely estimate, producing a realistic expected duration for uncertain tasks.
Define all tasks that form the project's critical path. Total expected duration = sum of individual TE, total variance = sum of variances (assuming independence). Evaluate probability of meeting a target deadline.
| Task name | Optimistic (O) | Most Likely (M) | Pessimistic (P) | TE | σ |
|---|
The Program Evaluation and Review Technique (PERT) is a statistical tool used in project management to analyze and represent the tasks involved in completing a given project. Unlike CPM (Critical Path Method) which assumes known durations, PERT accommodates uncertainty by using three time estimates: Optimistic (O), Most Likely (M), and Pessimistic (P). Developed by Booz Allen Hamilton for the U.S. Navy's Polaris missile project, it reduced project duration by two years and is now a globally adopted standard (PMBOK® Guide).
? Weighted Average Formula (Beta Distribution):
E = (O + 4M + P)/6 Variance = ((P - O)/6)²
The factor 4 gives the 'most likely' estimate quadruple weight, approximating a beta distribution shape, and the denominator 6 reflects 6 standard deviations (P-O ≈ 6σ).
A consumer electronics firm applied PERT to their firmware development cycle. Three tasks were critical: hardware validation (O=8, M=12, P=22 days), software integration (O=10, M=15, P=26), and compliance testing (O=4, M=6, P=10). Using this calculator, TEproj = (12.0+15.0+6.33) = 33.33 days, σproj = sqrt(5.44+7.11+1.0) ≈ sqrt(13.55)=3.68 days. The probability to complete within 38 days = Φ((38-33.33)/3.68) ≈ Φ(1.27) ≈ 89.8%. Management used this to confidently promise a delivery date, reduced contingency by 20%, and improved client trust.
Triangular distribution uses unweighted average (O+M+P)/3, whereas PERT emphasizes the most likely scenario. PERT tends to be more robust for skewed data (common in projects where delays are more probable than early finishes). Research by the Project Management Institute (PMI) suggests PERT outperforms triangular estimates for schedule risk analysis.