Once you’ve added your historical data and found the epics or versions to forecast against - you can run simulations. The forecaster has 2 modes - actual or simulation.
Actual mode
Actual mode will use the current number of epics or versions currently in a status of In Progress.
For Epics this is determined by the Epic’s workflow status. If it’s in status that has a status category of In Progress - then it’s deemed as in progress. For Versions this is determined by a a version configured start date. If it’s before the current date and the version hasn’t been released yet, then the version is viewed as being in progress.
...
When you are in actual mode, the Forecast parameters are limited to the certainty parameter. The concurrent throughput used is set to be the number of epics or versions “In Progress”.
The epics or versions that are “In Progress” are grouped at the top of the table and cannot be reordered. The forecast assumes that you’re working on all of them starting today and once they’re finished - you will continue to move onto the Epics or Versions not yet started. The Epics of Versions not yet started are rearrangeable (for more information - see below).
Simulation mode
Once you enter simulation mode you can configure the throughput yourself and all of the epics and versions that are being forecasted are viewed the same and can be rearranged.
...
Once you change the Epics in Progress or Versions in Progress you’ll need to click “Save” in order for the changes to take effect. This will cause the simulation to be rerun.
Rearranging the epics or versions
To rearrange the epics or versions order to be forecasted - simple drag the row to the new location. Once you have rearranged all of the epics or versions in order of importance - click on recalculate in the bar that is shown.
...
, now you need to choose a forecast certainty in the Forecast Parameters box.
...
A forecast certainty is a reflection of your ability to absorb risk in your forecast. The Monte Carlo Simulations that we use to forecast run thousands of simulated trials using the data you configured in the Portfolio Forecaster.
This allows us to see all possible completion dates and how often each completion date happened. This allows us to make a probabilistic forecast containing a probability (85%) and a range (by the forecasted date or earlier).
A probabilistic forecast: There is an 85% chance that Item X will finish on or before Y date.
This means there’s a 15% chance it won’t. You adjust the forecast certainty according to the risk you’re willing to take.
Please note two things:
We can never be 100% certain. The only things we can be 100% certain about are the things that already happen.
The higher you make your certainty, the more padding your forecasts will have for many items because it is accounting for more less likely possibilities. So, if you go too high you might as well just look at how long it took to do the longest-running item as use that as a forecast for any given epic.