tgoop.com/agile_coach_notes/1180
Last Update:
Вы спросите, а как тогда статистика, а статистика берётся по средней или средней взвешенной, а корридор оценки по минимальной и средней, максимальное значние по этой причине не трогаем.
Вы можете переоценить стори внутри спринта, когда поняли, что она сложнее или рискованнее, чтобы быстро скорректировать объем спринта и понять риски достижения цели, а вот переоценка по остаточному принципу… спросите меня - я скажу, не стоит так делать. А вот что думаент автор книги “Agile Estimation and Planning”
====
Suppose you’ve reached the end of the sprint and you have an incomplete story. You likely know that you don’t share the story at sprint review. You might also know that you don’t automatically move it forward, but instead put it back in the product backlog for re-prioritization.
But what people often don’t know is whether to re-estimate the story. After all, you’ve done some of the work to complete the story, so maybe it isn’t a 5 any more. Maybe now it’s a 3.
My answer is no, don’t re-estimate. Take an all-or-nothing stance in regards to velocity credit. That story counts for zero points in this sprint, but in the sprint the story is completed, the team gets all 5 points. There are a couple of reasons for this.
First, people are really bad at estimating percent complete. If you ask the team, “How far done are we” with a 3 point story, they’ll likely say about 2 points done. But really, they’re likely only about 1 point done. So they’re going to underestimate the work that’s left to complete.
Second, we don’t re-estimate the stories we finish, even if we discover they were bigger or smaller than we thought. So we shouldn’t re-estimate stories we don’t finish in the name of being “more accurate.”
Taking credit for a story at its original estimate only when it’s truly complete is the only way we can feel accurate in our predictions about when we’ll be done.
And being accurate (but not overly precise) is one of the ways you succeed with agile,
Mike