We've been using Jira for a couple of years now. We always create a new Sprint before the current one is ended (to plan out work for the next Sprint ahead of time). When we End a Sprint we roll over incomplete issue into the next Sprint, then start the next Sprint right away.
Issues that rollover are counted as part of the "Sprint Started" list in the Burnup report, as expected. But in 3 of the past 8 Sprints, those issues get treated as 'Added to the Sprint', with the same time stamp as the start of that Sprint.
So on the surface of it, those issues are counted the same way as issues that are created from scratch and manually added to the Sprint.
I'm not doing anything with manually changing the Sprint Start date/time, whatever the system time is, I'm letting it be used.
Any idea on what could be driving this inconsistency? Any ideas on how to prevent it in the future?