In my company, there are 3 different development teams, newly started scrum. They mainly follow same path in doing scrum, but when it comes to sprint burndown chart, the act differently.
One team, gathered story points (with estimated points in them), broke them into tasks and estimated points for them too, i.e, a 10 point story may have 3 tasks with 4, 1 and 5 points estimated for them respectively. When they completed a task, they updated burndown chart respectively. The result, was a relatively smooth burndown chart, but in case of failed stories it was somehow deceptive; It showed progress that wasn't actually made for the end user.
On the opposite side, other teams didn't estimate points for tasks and just updated the burndown chart when a story fully completed. The result was a stair shape diagram: for some days team did no progress and suddenly made a big step.
The question is, what is the right way to make the burndown chart and which approach was closer to it?