Amount of "done" issues differs from JQL results obtained from Jira

[Anna Henke]

It would be good to gain a little more insight on how the issues are identified that count into the Kanban Velocity Gadget as the data presented in the gadget differs from what we found obvious when querying Jira directly via JQL. Example: We configured the gadget to display data based on a filter. Done has the meaning "Issue status has category "Done""

When we now query Jira with the same filter criteria and limiting the results to one of the displayed intervals, i.e. by adding the filter criteria "and (status changed to "Done" during (2018-08-01,2018-08-14))" we get a different result than what is displayed in the gadget.

We are still in the evaulation phase for the plugin and the fact that it is not clear to us how the gadget calculates the issues that count into the statistic is currently a blocker.

[Danut M [StonikByte]]

Hi Anna,

The add-on applies some extra filtering on top of the filter that you provided: 1. sub-taks are excluded 2. issues created after the end date are excluded (because they were created outside of the interval) 3. issues that were last updated before start date are excluded (because they were not resolved in the time interval)

So the filter is something like this: filter=YOUR_FILTER_NAME AND issuetype IN standardIssueTypes() AND updatedDate > START_DATE AND createdDate < END_DATE

In addition, the option "Issue status has category Done" does not refer to a status named "Done"; it refers all the statuses from Jira that have category "Done" (see picture below).

[Anna Henke]

Hi Danut, thanks for clarifying - calculating velocity is such a pain when not doing it on paper... ;-) I understand the difference between Category "Done" and STatus "Done" - the reason why I only filter for "Done" issues is, that we also have a Status "Accepted", which is only reachable after setting a Story to "Done" (=the team has finished its work). If a story is accepted in the next interval based on my calculation I would count a story twice. I also understand that you are kind of "working around" exactly on the same issue by considering the updatedDate - this will anyhow falsify the velocity retrospectivly, if an issue is updated later. This is likely to happen, e.g. when it is commented on, an issue is linked to it or similar. Any thoughts on that, Danut? As said, calculating Velocity is so simple if done on paper, but querying it retrspectivly is a pain...

[Danut M [StonikByte]]

We recently added a Data tab to the Team Velocity gadget (can be enabled via gadget settings) that indicates which and how issues “contributed” to the velocity calculation. Closing ticket for now.