PERT estimate calculator
Enter three estimates for a task or project, the optimistic, the most likely and the pessimistic one, plus your deadline in the same unit (hours, days or weeks). The calculator gives the PERT expected time, its standard deviation, a range that holds about 95 % of outcomes, and the chance of finishing by the deadline.
Runs in your browser: nothing you enter is sent anywhere.
How to use it
- Enter the optimistic time: what it takes if everything goes right.
- Enter the most likely time: what it usually takes.
- Enter the pessimistic time: a realistic bad case, not a catastrophe.
- Enter the deadline in the same unit to see the chance of meeting it.
All four numbers can be hours, days, weeks or story points, as long as they use the same unit.
How it works
- Expected time: E = (O + 4M + P) ÷ 6
- Standard deviation: σ = (P − O) ÷ 6; variance σ²
- Chance to finish by deadline T: z = (T − E) ÷ σ, then the standard normal probability Φ(z)
- About 95 % range: E − 2σ to E + 2σ (the lower end is never shown below 0)
- 90 % confidence estimate: E + 1.2816 σ
If all three estimates are equal, σ is 0 and the chance is 100 % when the deadline is at or after E, otherwise 0 %.
Example
A task estimated at 4, 6 and 14 days with a deadline of 9 days:
- E = (4 + 4 × 6 + 14) ÷ 6 = 42 ÷ 6 = 7 days
- σ = (14 − 4) ÷ 6 = 1.67 days; σ² = 2.78
- z = (9 − 7) ÷ 1.67 = 1.2, and Φ(1.2) = 0.885, so the chance is about 88.5 %
- About 95 % of outcomes fall between 3.67 and 10.33 days
- The 90 % estimate is 7 + 1.2816 × 1.67 = 9.14 days
Common mistakes
- Using a “best case” that never happens as the optimistic estimate: it pulls E down and makes σ look bigger than it is.
- Adding standard deviations of tasks instead of variances, which overstates the uncertainty of the whole project.
- Mixing units, such as hours for one estimate and days for another.
Frequently asked questions
What is the PERT formula?
The expected time is E = (O + 4M + P) ÷ 6, a weighted average that counts the most likely estimate four times. The standard deviation is σ = (P − O) ÷ 6, on the idea that the range from optimistic to pessimistic spans about six standard deviations.
How is the chance to meet the deadline calculated?
The outcome is treated as normally distributed with mean E and standard deviation σ. The deadline T gives z = (T − E) ÷ σ, and the chance is the normal probability of a value at or below z. A deadline equal to E gives 50 %.
How do I estimate a whole project?
Work out E and σ² for each task on the critical path, the chain of tasks that sets the end date. Add the E values to get the project's expected duration, and add the variances (not the standard deviations); the square root of that sum is the project's σ. Then use the project totals as one estimate.
What is the 90 % confidence estimate?
The time you would meet nine times out of ten under the same normal assumption: E + 1.2816 σ. It is a common choice when you need to commit to a date rather than state an average.
How reliable is this?
It is only as good as the three estimates. The normal curve is an approximation; real projects tend to overrun more often than they finish early, so treat the probability as a guide, not a promise.