Our development cycle is as follows (roughly):
- Sprint planning for sprint X
- Sprint X start; approx 2.5 weeks dev. time
- Sprint X ends. All issues should be resolved and closed in accordance with DoD (cut-off)
- Regression testing for about 1 week. Identified issues, if important enough, is added to Sprint X. If not, added to backlog.
Now, at the same time as the regression testing start, the team take on sprint planning for sprint Y. The QA team is divided - half the team run the regression test, the rest join in on the sprint planning to ensure they have a good understanding of what is going to happen in sprint Y.
This means that we have 2 sprints in parallel for about 1 week. Now, my challenge is this: For now, the sprint end date is set for the last day of the regression test. This to include the logging of work for the QA team doing the regression.
The burndown takes this end-date into consideration for the graph. Which means, in practice, that the burndown is wrong. The actual burndown is steeper because all issues must be closed one week before sprint end date.
I guess the only thing I can do to get the burndown correct is to use the actual cut-off date. My question is then - is there any other way for me to illustrate the actual release date on the dashboard?
"Sprint end date: 01. nov. 2014. Release date: 08.nov. 2014" (example only, haven't even checked the days) 
