Three Point Estimating
Three point estimating uses three scenario based values (optimistic, most likely, pessimistic) to produce a more reliable estimate than a single guess, either through a simple average or the PERT weighted formula.
How Three Point Estimating Works
Three point estimating is a technique that improves estimate accuracy by replacing a single guess with three scenario based estimates: optimistic (the best realistic case), most likely (the expected case under normal conditions), and pessimistic (the worst realistic case). Combining these three values produces a more reliable estimate than any one of them alone because it forces the estimator to consider uncertainty rather than anchoring on a single number.
The technique is applied to both duration and cost estimates. For each task or work package, the estimator provides three values, and one of two averaging formulas is applied to produce the final estimate.
The Two Formulas
The simple average (also called the triangular distribution) gives equal weight to all three estimates: E = (O + M + P) / 3. This is appropriate when the estimator has no reason to believe the most likely scenario is more probable than the others, or when the distribution of outcomes is roughly symmetric.
The PERT weighted average gives the most likely estimate 4x the weight: E = (O + 4M + P) / 6. This reflects the beta distribution assumption that the most likely value has the highest probability. It is the standard formula for duration estimating in PMI methodology because task durations tend to cluster around the most likely value with a skew toward the pessimistic end.
When to Use Three Point Estimating
Use three point estimating when tasks have meaningful uncertainty: the optimistic and pessimistic cases are meaningfully different from the most likely case. If a task could take 5 days or 15 days depending on conditions, a single estimate of “10 days” hides critical schedule risk. Three point estimating surfaces that risk and quantifies it.
The technique is especially valuable for tasks the team has not done before, tasks with external dependencies (vendor response time, client approval), tasks involving new technology or integration, and any task where the estimator says “it depends” when asked for a duration.
Three point estimating also provides the inputs needed for PERT calculations and Monte Carlo simulations. It is the foundation for probabilistic schedule analysis.
When Not to Use Three Point Estimating
Routine tasks with well known durations do not benefit from three estimates. If a weekly status report always takes 2 hours to write, the three estimates would be 1.5, 2, and 2.5 hours, and the formula would produce approximately 2 hours. The calculation overhead adds no value.
Teams that produce superficial three point estimates (most likely plus or minus a token amount) gain nothing from the technique. The value comes from genuinely considering what conditions would produce the optimistic and pessimistic outcomes, not from arithmetic on trivially different numbers.