Monte Carlo: Delivery Date
Open this chart when you need to calculate the delivery date for a number of tasks instead of the number of tasks that can be completed within a timeframe.
Monte Carlo: Number of Tasks
What it answers
What are we likely to deliver?
Forecast how much work your team can complete within a set timeframe and the probability of each outcome.
Example
How to read it
-
Possible task count
Each bar represents one possible task count and how often it occurred in the simulation.
-
Percentile line
Each vertical dotted line shows the probability of completing a certain number of tasks within the selected timeframe.
-
Hover details
The hover tooltip shows the task count, its frequency, and its probability. The probability expresses the associated risk as a percentage.
-
Frequency axis
The Frequency (Number of trials) axis shows the number of simulation trials associated with the possible scenarios.
See the Monte Carlo: Number of Tasks in action
Understand the probability distribution chart
Monte Carlo simulations use thousands of random trials based on your past throughput data to forecast throughput for a future timeframe. Enter a start date and end date to see a range of possible task counts and the probability of each outcome.
The simulation runs tens of thousands of times and presents the results as a probability distribution. Percentiles increase from right to left, and outcomes farther to the left have greater certainty. The 50%, 70%, 85%, 95%, and 98% percentile lines show the risk associated with each outcome. Compare probability levels, such as 70% and 85%, to balance speed and confidence.
Use percentiles to set a reliable task count
The vertical dotted lines are percentiles. They show the probability of completing a certain number of tasks within your timeframe, including 30%, 50%, 70%, 85%, 95%, and beyond. Use them to compare confidence levels and decide how much risk to accept when committing to a task count.
For easy work, a stable delivery workflow, and no expected obstacles, use a lower percentile such as 70%. For complex or unfamiliar work with many unknowns, commit to a higher percentile such as 95% or 98% to increase the probability of meeting your goal.
Define your simulation scope
Set the start date and release date for the Monte Carlo forecast. The simulation uses your past throughput data to generate possible scenarios and provide a range of task counts with their probabilities.
The simulation predicts how many tasks will be finished by a given date, including tasks already in progress. It does not predict which specific tasks will be completed.
Scale your throughput to account for performance changes
The scale factor adjusts the forecast for expected changes in team performance. Scale throughput down for an expected slowdown caused by holidays, resource constraints, or other external factors. Scale it up if you expect delivery speed to increase, such as after onboarding new team members. The default value is 0, so the simulation uses past performance without adjustment.
Use the scale factor only when you expect throughput to change and have no historical data that reflects the change. If past data already represents similar conditions, rely on that data instead.
You can export the filtered data shown in this chart as a CSV.
What to look for
Confidence at different percentiles
Compare task counts at different probability levels, such as 70% and 85%, to understand the trade-off between speed and confidence. Outcomes farther to the left have greater certainty.
Complexity and unknowns
Check whether the work is easy and stable or complex, unfamiliar, and affected by unknowns. These conditions determine whether a lower or higher percentile is appropriate.
Expected throughput changes
Look for an expected slowdown or increase in delivery speed, then check whether similar conditions are already represented in your historical data.
What decisions it helps you make
- Decide how much risk to accept when committing to a number of tasks by comparing percentile confidence levels.
- Decide whether to scale throughput down for an expected slowdown or up for an expected increase in delivery speed.
Related charts
Throughput Histogram
Open this chart to review the past performance data used as the basis for Monte Carlo analysis.
Cycle Time Scatterplot
Open this chart as another tool for producing delivery forecasts.
See these analytics with your own delivery data.
Connect your existing delivery data and start exploring your flow.