Calculate PERT (Program Evaluation and Review Technique) estimates with weighted averages. Get expected duration, standard deviation, and confidence intervals for project tasks.
The PERT Estimate Calculator uses the Program Evaluation and Review Technique to calculate weighted task duration estimates from optimistic, most likely, and pessimistic scenarios. PERT is one of the most widely used methods in project management for creating realistic schedule estimates that account for uncertainty. It is especially useful when teams need to replace a single guessed date with a defensible range.
Unlike simple averaging, PERT weights the most likely estimate four times heavier than the optimistic and pessimistic values, producing a more realistic expected value. The technique also calculates standard deviation and variance, enabling probability-based schedule analysis with confidence intervals at 68%, 95%, and 99.7% levels.
This calculator supports both individual task estimation and multi-task project analysis. Enter estimates for multiple tasks to see the total project duration with aggregate confidence intervals, critical path identification, and Monte Carlo-style risk analysis. It also supports the modified PERT formula with adjustable weighting factors for teams that prefer different emphasis on the most likely estimate.
Use this calculator when you need a schedule estimate that reflects uncertainty instead of a single-point guess. It is useful for task planning, deadline discussions, buffer setting, and combining several three-point estimates into one project-level duration range. It also gives stakeholders a clearer view of risk than a flat duration estimate.
E = (O + 4M + P) / 6. Standard deviation σ = (P − O) / 6. Variance σ² = ((P − O) / 6)². For multiple tasks: E_total = ΣEᵢ, σ_total = √(Σσᵢ²). Confidence: 68% = E ± σ, 95% = E ± 2σ, 99.7% = E ± 3σ.
Result: PERT estimate: 7.0 days, σ = 1.67 days, 95% CI: 3.7–10.3 days
E = (4 + 4×6 + 14)/6 = 42/6 = 7.0 days. σ = (14-4)/6 = 1.67 days. 95% confidence interval: 7.0 ± 2×1.67 = 3.7 to 10.3 days.
While PERT was developed for large defense projects (Polaris missile program, 1958), it's widely applicable to software estimation. Agile teams can use three-point estimates during sprint planning to size stories more accurately. The standard deviation helps set realistic sprint commitments — committing to the PERT estimate gives roughly 50% confidence, while E + σ gives ~84%.
For more sophisticated schedule analysis, PERT estimates can feed Monte Carlo simulations. Each task's duration is sampled from a beta distribution defined by its three estimates, and thousands of simulated project runs produce a probability distribution for total project duration. This shows the probability of meeting any given deadline and identifies which tasks contribute most to schedule risk.
Planning fallacy (overconfidence in best-case scenarios), anchoring (fixating on the first number mentioned), and Parkinson's law (work expanding to fill available time) all affect PERT inputs. Counter these by: using reference class forecasting (comparing to similar past projects), having different people estimate optimistic vs. pessimistic scenarios, and tracking actual vs. estimated durations to calibrate future estimates.
PERT expected value = (Optimistic + 4 × Most Likely + Pessimistic) / 6. This weights the most likely estimate four times more than the extremes, creating a beta distribution approximation that reflects reality better than a simple average.
The 4× weight is based on the beta distribution assumption that the most probable outcome is more informative than extreme cases. In practice, this produces estimates that are slightly pulled toward the pessimistic side since pessimistic estimates tend to be further from the most likely than optimistic ones.
PERT is better than single-point estimates but still relies on the quality of the three input estimates. Research shows that developers typically underestimate pessimistic scenarios. Adding a "black swan" factor or using 99.7% confidence intervals helps compensate.
Use any consistent time unit—hours, days, weeks, story points, or sprints. The formulas work the same regardless of unit. Just be consistent across all three estimates and all tasks in a project.
For tasks on the critical path (sequential), add all expected values for total duration. The combined standard deviation is √(σ₁² + σ₂² + ... + σₙ²) — NOT the sum of individual standard deviations.
Triangular uses equal weights: (O + M + P) / 3. PERT gives 4× weight to M, producing a tighter estimate. PERT is preferred when the most likely estimate is well-understood; triangular when all three estimates are equally uncertain.