How the numbers are calculated

Every chart follows the same few rules. They are fixed, so a number means the same thing on every chart and every dashboard.

Calendar days

All durations are counted in calendar days, weekends and holidays included. The people waiting for the work count calendar days, and a calendar day is the same for every team, while working days vary with holidays and part-time schedules. Cycle time is therefore elapsed time, not effort: a work item that took 10 days may have been worked on for only some of them.

Charts show durations in days and hours, rounded up to the next full hour, for example "17d 8h". Less than an hour shows as "<1h". Because they are rounded up, a percentile line never shows less time than was measured.

Start and finish

A work item starts when it first enters a status in Jira's In Progress category. Moving back to To Do later does not reset the start. A work item is finished when it has a resolution date; a status in the Done category without a resolution date does not count as finished.

Whether a status counts as In Progress follows its current category in Jira. If an admin moves a status to another category, past work is counted with the new category. A status that no longer exists counts as not in progress.

Work items, not story points

Every chart counts work items. Estimates such as story points are not used, so the numbers do not shift when a team changes how it estimates.

Percentiles

Where a chart shows what is typical, it uses percentiles by default: the 50th, 70th, 85th and 95th. The 85th percentile is the value that 85% of the work items stay at or below; values between two work items are interpolated. The mean is available as an option; it is not the default because a few very long work items pull it up. The two Monte Carlo forecasts read their probabilities from 10,000 simulated outcomes; their pages explain how.

Rework

When a work item goes from an In Progress status back to To Do, the Cycle Time Chart marks it as rework. The time is never subtracted: time spent going back is part of how long the work took.

Time window

"Days previously" counts back from now, in your time zone. For the Cycle Time, Time in Status and Work Item Age charts, the window includes both ends. It can be at most 720 days. Two exceptions: the Work in Progress Chart allows at most 180 days, and charts grouped by hour allow at most 14 days.

Periods

The Throughput, Created vs. Resolved and Scope Change charts group work items by period: hour, day, week (starting Monday), month or year, in your time zone. Their window starts at the beginning of the period that contains the day N days ago and ends now. So the first period is always complete, and the chart can cover up to one period more than N days: with monthly grouping, up to a month more. The header shows where the window starts, for example "Since 2026-09-28". The last period is the current one and still running, so it only holds what has happened so far; its tooltip says "Still running: counts up to now."

Days in Work in Progress

The Work in Progress Chart counts per calendar day: the last N days, today included. A work item's status on a day is the last status it had that day. Today's value counts up to now, and its tooltip says "Today: as of now."

Monte Carlo forecasts

Both Monte Carlo charts take their history from complete periods only (days, weeks or months). The window starts at the beginning of the period that contains the day N days ago, and the current, unfinished period is left out. Each complete period gives one number: how many work items were resolved in it.

Each forecast runs 10,000 simulations. Every simulated period draws one of those past numbers at random, each with the same chance, and the same number can be drawn again. Because the draws are random, results can differ slightly from one load to the next.

The two charts read their tables in opposite directions, so both can be read as "with P% probability". The "When" Chart's 85% row shows the date by which 85% of the simulations had finished, so a higher probability means a later date. The "How Many" Chart's 85% row shows the number of work items that 85% of the simulations reached at least, so a higher probability means fewer work items. Both take the value directly from the sorted simulation results, without interpolating.

Whose data

Every chart reads Jira data as the person viewing the dashboard, so each viewer sees the numbers for the work items they are allowed to see. No chart breaks numbers down by person.