How come the Velocity and Sprint reports story points don't match what is shown in an active sprint's story points?

Question

The current behavior of JIRA is that you can update story points to a story in an active sprint, but if you do that, the Sprint Report shows the # of when the sprint started.  The "Velocity Report" is also behaving the same way.  It shows the bar graph using the ORIGINAL story points at the start of the sprint.  This is also of no value.  When we finish the sprint we need to know WHAT we got done and it needs to reflect reality so the reports we share with management are accurate?  Any plans to fix these 2 reports?

Antony Marsh · Answer

We did exactly the same thing as David - possibly bad agile practice agreed, but sometimes things happen :-)

I also tried changing sprint scope by moving the story to be split out, reestimating it (smaller), creating the new story for remaining work, moving it back in to active sprint, moving to done...  didn't change the sprint report value even though.

My clumsy workaround (which did work) was to move the original story out of the active sprint, moved it to a "Won't Fix" release (or you could just delete it), create a new story with the 5 point value, move to active sprint, move to done in active sprint.   Hey presto, job done.

David Tombs · Answer

Andre, I understand your point, but this is hurting our team not because we re-sized stories but because we split off part of a story out of a sprint. We had an 8-point story which we split into one 3-point and one 5-point story, and moved the 5-point story out of the sprint. The Velocity chart, however, shows that we completed all 8 points in the sprint, which is not true!

Andre Borzzatto · Answer

This is currently deliberate because the report is trying to show the effort that the team originally committed to (and therefore the estimate that will form their velocity). For velocity to be meaningful all stories need to be estimated with the same level of non precision, i.e. if you have 50 stories in the backlog and estimate them all with no pre-knowledge then from Sprint to Sprint you can measure how many points of equally unknown work can be completed. If you add additional knowledge to some stories but not others (by changing their estimate) you can no longer compare your velocity against these 'more known' stories to items further down in the backlog where you don't have that knowledge.
The Sprint report will also show you the value committed, but it will contain any other issues added after its start.