Assumptions and method

When to be cautious

Assumes independent triangular inputs. It is not a schedule-network simulator or evidence that the supplied ranges are calibrated. Percentiles are estimates, not guarantees.

Sample each triangular input and sum; repeat 10,000 times

Test a scenario

Use your own inputs

Inputs stay in your browser. Shared scenario links include your inputs.

Use the result in a decision

Describe events, consequences and assumptions before assigning numbers.

Assumes independent triangular inputs. It is not a schedule-network simulator or evidence that the supplied ranges are calibrated. Percentiles are estimates, not guarantees.

Set a baseline and change one assumption to compare outcomes.

A worked example

Follow the fixed teaching example
  1. For each trial, draw one value from each triangular component and add them.
  2. Repeat 10,000 times using seed 42, sort the totals, and read the reported percentiles. These are simulation results conditional on the inputs and independence assumptions.

These illustrative inputs describe a project scenario, not a published benchmark. All monetary inputs use the same currency and price basis.

One cost component per row: minimum, mode, maximum
8000, 10000, 14000 3000, 5000, 9000
Reproducible seed
42
Mean of 10,000 simulations
16,350.4965
Median
16,271.325
80th percentile
17,894.3931
95th percentile
19,463.0992

Interpret the output only within the assumptions above. Changing the inputs changes the result; it does not validate the inputs.

Source & credit

Standard estimating practice; the cited guide documents use rather than claiming invention.

GAO: Cost Estimating and Assessment Guide

This is independently written code and explanation of the underlying method. The linked publication, its diagrams and its trademarks remain its owner’s material; no permission to reuse them is implied.

Learn and apply

Explore risk: guides, calculations and sources