Scrum Issues

Question

We have been using Scrum for about 9 months, and it has been pretty much successful. However, our burnout graphs rarely look like “model” charts, instead resembling more terrible roller coaster rides with some vomiting, causing ups and downs.

To try and deal with this, we spend more time in front of the prototype and sprint design, but we still seem to discover much more work during the sprint than originally intended. Note. By this I mean that the work required to satisfy the lag is more complicated than the first thought, and we did not identify new elements for the lag.

This is a common Scrum issue, and does anyone have any tips to help smooth out the ride?

I should note that most of our developments are not "green", so we support functionality in the existing large and complex application. Is scrum less suitable for this type of development simply because you don't know what problems existing code will cause?

How much time should we spend before the sprint begins to develop development details?

UPDATE: We have more success and a smoother ride. This is largely because we appreciated a more pessimistic point of view, appreciating that gives us more space for breathing to do things when they don’t go to plan. You could say that this allows us to be more “mobile”. We are also trying to change the idea that the combustion diagram is a kind of schedule, not an indicator of the amount of resources v.

Answer 1

Some smoothing tips.

1) As others said - try to break the tasks into smaller pieces. A more obvious way to do this is to try to break down the technical tasks in more detail. If possible, I would advise you to speak with the owner of the product and see if you can reduce the volume or the thin history. I find the latter more effective. Juggling priorities and evaluations are easier if both the team owner and product owner understand what is being discussed.

My general rule is that any estimate that exceeds half an ideal day is probably incorrect :-)

2) Try making shorter sprints. If you do one month of sprints, try two weeks. If you do two weeks, try one.

• It acts as a plot size limiter - encourages the product owner and team to work on smaller stories that are easier to evaluate accurately.
• You often get feedback from your grades - and it’s easier for you to see the relationship between the decisions you made at the beginning of the sprint and what really happened.
• Everything is getting better with practice :-)

3) Use racks and retrospectives to look a little more at the reasons for the ups and downs. This is when you spend time with specific areas of the code base? Is this caused by people misunderstanding the product owner? Accidental emergencies that require development time from the team? Once you have a better understanding of where the highs and lows come from, you can often solve these problems. Again - shorter sprints can help make this more obvious.

4) Believe your story. You probably know this ... but I’ll say it anyway :-) If you mess with this terrible legacy, the Foo package took 3 times as long as you thought it would last a sprint - then it will also take 3 x as much you think the next sprint. No matter how efficient you think you are this time ;-) Trust stories and use things like Yesterday’s Weather to guide your estimates next spring.

Hope this helps!

Answer 2

I am glad to hear that the fight was largely successful for you - this is more important than an ideal sprint burning schedule. Sprint burning is just a tool for a team to help him find out if he is on his way to sprint goals. if the team meets the sprint goals, I wouldn’t worry too much that the chart looks like a roller coaster. Some suggestions

• During a sprint retrospective, ask the team where the extra work comes from
• Additional work may arise due to the lack of good acceptance tests at the start of the sprint.
• Additional work may arise due to the lack of a well-groomed lag. A good rule of thumb is to spend at least 5% of the team’s time thinking about the next sprint stories ahead of time.
• Work with the monitor continues - does the team work too much in parallel?
• During sprint planning, how does the team think about breaking down the tasks that make up the stories?

If you have not met the sprint goal, use the set team speed to accept less during the next sprint. You must train well before you can run.

Answer 3

In my experience, Scrum is definitely focused on new development rather than maintenance. The new development is much more predictable than supporting the old, large code base.

With that said, one of the possible problems is that you do not break tasks into sufficiently small pieces. The general problem of people with program planning as a whole is that they think, "Oh, this task will take me 2 days," without thinking about what is happening with this task. Often you will find that if you sit down and think that this task is to complete A, B, C and D and ends with more than 2 days of work.

Answer 4

As others said, I would expect it to burn up and down. Things happen! You should use the up and down bits as feed for your retrospectives.

Make sure everyone understands what is being done, and use this shared understanding to help you manage your scheduling sessions. Often, having a list of what will be done will (a) help you remember things that you might forget, and (b) will most likely cause more ideas for tasks that might otherwise arise later.

Another thing to think about - if you work for a month with an unpredictable code base, I still expect your speed to normalize to a fairly stable rate. Just track your success from the planned work and use only ready-made elements at maximum when planning. Then focus on your unplanned tasks and see if there are any templates that suggest that there are things you can do differently to incorporate these things into your planned work.

Answer 5

I had similar problems. My previous team (more than a year on it) was large, and we supported a very large, rapidly changing code base for a series of initial product launches. Our burning looked shameful, but that was the best we could do.

One thing that can help (to make your chart look better) is to keep the number of hours / points fixed to a constant. If you underestimated the task and must double the clock, pull something out of the sprint. If you are performing a new task, this obviously has a higher priority than what your team took on to pull out this other thing.

We tried to break the task down into many tasks before the planning, and this did not seem to help. In fact, it just gave us more damn tickets to keep track of during the sprint. Requirements began to migrate to tickets and (not surprisingly) were lost in all the shuffles.

In my new team, we took a rather radical approach and started creating large tickets (several weeks) that say things like “implement v1.2 in ProjectX”. Requirements / functions lists for ProjectX (including version 1.2) are stored on the wiki, so the ticket is very clean and only tracks the work done. This helped us a lot - we had fewer tickets left, and we were able to complete all our sprints, although we continue to shoot our sprint tasks to help other teams or put out fires.

We continue to push elements out of the sprint if (and only if) we are forced (by a person) to introduce new objects.

Another simple piece of advice that helped us: add the “total number of hours in the sprint” to your burnout. This should be the sum of all ratings. Maintaining this line can help and increase the visibility of the problems your team may face (provided that you do not get demotion ...)

-ab

Answer 6

I had similar problems in my overtaking. I “fixed” it, specifying what is included in the burnout.

SiKeep commented:

His progress against lag has been chosen for this sprint, which may or may not end as a release.

Since you have chosen certain things to sprint and what is on burning, I don’t know that all the “new work” should appear on burning. I would see this happening behind (not affecting the burning), unless it is important enough to go into your current sprint (which will then be displayed as an uptrend in burning).

However, slight up and down are normal if the trend line basically matches your expected speed. I would be concerned about the trend on skates that you are talking about. However, the idea of ​​isolating burnout by adding only high priority elements to the current sprint can help weaken these up and down your burnouts.

As others have said, planning before the sprint should be short ... ( no more than 4 hours ).

Answer 7

We use the "time-boxed" task for unplanned tasks. Whenever a high-priority operation occurs or pop-up tooltips appear, we can use the time-box time (but we can never go to zero). This method has the added benefit that we can easily track which tasks were unforeseen and take these factors into account during our next sprint planning.

Answer 8

You can integrate new work into the sprint start date to have a great Burndown schedule.

You can mark additional work with a certain marker and evaluate at the end of the sprint why you could not identify these tasks before.

Answer 9

Now we use the burn up chart. Instead of just setting the scope of work, we will draw two things: the amount of work done and the total amount of work (i.e. Completed + outstanding).

This gives you two lines on the graph that should meet when all the work is done. It also has a big advantage in that it clearly shows when progress is slow because more work has been added.

If you like, PO "owns" one line (shared work), and developers / testers "belong" to another line (work done).

The PO line will rise and fall when they add / remove work.

The dev / tester line will only work as it completes.

Answer 10

Article Is this your burnout schedule? explains what this status means when burning a chart. He also offers tips on what to do about it.

Some examples described in the article:

Answer 11

It is as it should be. If your burnout schedule is similar to the model schedule, you have problems. The chart will help you find out if you can do everything possible and complete all the stories.

Detection of stories during the sprint will always occur. Ideally, you could design and find tasks ahead, but if they worked, why would a large project not work? To answer the last question, sprint planning should take no more than four hours .