Read the current forecast
Compare actual progress with the median and the 50% and 90% simulation bands. The wider band communicates more uncertainty than the single line at the center.
Reporting & Forecasts
A delivery date with room for reality.
Explore delivery ranges, scope changes and risk drivers. Have a more useful conversation than a single promised date.
ACDFMC+1
Illustrative workspace · interactive demonstration
See the range
Compare the median forecast with wider uncertainty intervals.
ENG-284Linked context
Test a scope change
Move the scope slider and see the demonstration recompute.
Review the range
Expose the assumptions
Keep historical throughput and model limitations alongside results.
Ready for reviewHuman approval
A practical workflow
Use the range to have a better scope conversation.
A delivery forecast becomes useful when the team can inspect its history, compare a change and explain the tradeoff.
Change the scope
Move the added-scope slider. The demonstration recomputes 1,000 paths from the same seed so the before-and-after comparison reflects the scope change rather than a different random sample.
Discuss the deadline
Inspect the completion distribution and probability by the target date. Bring dependency changes and capacity assumptions into the conversation before choosing a commitment.
A connected working example
A date needs a capacity conversation.
The Atlas migration example shows how additional scope changes the delivery distribution. A narrow range is not a promise, and a wider range is a useful signal to discuss uncertainty with the team.
Pair the forecast with recorded capacity. A planned task may compete with support work, leave or another project. Review whether the historical throughput still describes the team that will do the next sprint.
ACDFMC+1
Illustrative workspace · interactive demonstration
Inside the approach
How the model works.
This demonstration samples weekly throughput from a lognormal distribution fitted to fictional history. Its 50% and 90% bands summarize simulations, not measured product accuracy. Real estimates require relevant history and calibration.
Methodology and evaluation01Throughput history
021,000 sampled paths
03Date distribution
Before you move forward
Keep the interpretation next to the chart.
| Review | What to look for |
|---|---|
| History | Check how completed work is counted, which period is used and whether the process has changed. |
| Intervals | The 50% band uses the 25th–75th percentiles. The 90% band uses the 5th–95th percentiles of simulated paths. |
| Probability | The displayed probability counts paths that reach the current scope by the deadline. It is conditional on the simulation assumptions. |
| Decision | Use the range to compare scope and timing. Review external dependencies and unrecorded work before relying on a real forecast. |
Questions & answers
A few useful answers.
Are these production accuracy numbers?
No. The charts use seeded fictional throughput and project records. They demonstrate the calculation and interaction; they do not measure production model accuracy.
Why does added scope change the probability?
Each simulated path has more work to complete with the same sampled throughput. Reusing the seed makes the comparison consistent and prevents added scope from improving the simulated completion chance.
How should a real forecast be evaluated?
Compare it with outcomes from later, held-out projects. Review interval coverage, date error and changes in work practices. The methodology page explains the evaluation questions.
Ship what you promised.
Bring the plan, the work and the next decision together.