What else is wrong with Story Points as a time unit?

There is one another point against considering a story point as a certain period of time (e.g. half a day). It is difficult to use special planning cards for a planning session because their values are like 1, 2, 3, 5, 8, 13, 20, etc. but the time in other hand is linear. So the problem is what to do if we estimate that something should take 17 days?

The original post with comments

Why Story Points should be abstract units?

Imagine that you assumed that 1 story point is about 4 ideal hours of work and you made all the estimations according to that assumption. When the sprint finished you divided a sum of story points of done stories by a number of spent hours and the result was that 1 story point is 6 hours. So how to proceed during the next sprint planning? If we assume that 1 story point is now 6 hours, we estimate stories in a different way than before. In other hand, keeping in mind that 1 story point is 4 hours, although we know that last sprint it was 6 hours is also confusing. In the described case 1 story point corresponds to some certain time, so it looses its real meaning.

Story points should be some abstract units expressing relations between stories in a way a team asses them. When they are abstract there is no other way to estimate a story than by comparing it to already estimated stories. Thanks to that, estimating is always the same, there is no confusion.

WBS and Scrum

WBS is a great tool to give a thought, especially that recently several people asked me about it. There are several aspects which might be considered...

Some people things that WBS is something totally detached from Scrum. I don't agree with this. First of all if we would show a product backlog and a sprint backlog as a structure, we could consider it to be WBS. The main difference would be related to the fact that WBS must cover 100% of the work effort what means, that we at least have some placeholders for the work which is unknown at the moment. Regarding similarities, PMI Book explains that WBS decomposition for a deliverable or a subproject that will be accomplished far into the future may not be possible. In that case the team waits until they have more knowledge (Rolling Wave Planning) - sounds like Scrum to me:).

Recently my colleague asked an interesting question. He said that he already has WBS for his project but he would like to start following Scrum, he asked what he can do with it? It depends on the WBS which he has, but most probably with small modifications, WBS items could be used as themes, stories and tasks. In addition Scrum doesn't say that you can't have a detailed plan in the beginning, it just says that it might not make sense to have one and that it might be misleading. WBS items could be helpful during Sprint Planning meetings or calculating an initial team velocity and an initial release schedule.

It is also important to remember that WBS is not meant only for PM, but it is also a communication tool, shows a customer why a project is estimated for a certain budget, verifies how well a customer defined the requirements, helps forecasting a risk, etc. In a project where a customer is against Scrum it could be used at some point, in the middle of the project to show how much effort was wasted creating it, since most probably, later on, the modified plan would be far from the original one.

Summarizing, I'm quite sure that in some cases I'll use WBS even in the Scrum projects. It is important to remember that we decide what WBS really contains and how detail it is.

How to get familiar with the Scrum

Below is a list of very good books about Scrum which I fortunately has happened to read. I always suggest them as a good start for someone who wants to get familiar with the Scrum. Every book shows something different, so I added short descriptions to the titles:

1. "The Scrum Primer", Pete Deemer, Gabrielle Benefield, Craig Larman, Bas Vodde
   This is a very short (around 20 pages), very well written, to the point document which gives an overview of the Scrum. It is available for free under the following link: Scrum Primer
   
2. "Agile Project Management with Scrum", Ken Schwaber
   This was the first book which I read about the Scrum. For some reason it took me the most time to read it, but I would suggest you to start with it too. There are explained basics concepts and it mainly contains real life examples.
3. "Agile Estimating and Planning", Mike Cohn
   This is another great book. I read it as the last. It presents solutions for lots of issues which might occur during a planning and estimating phase of a project.
4. "Agile Software Development with Scrum", Ken Schwaber, Mike Beedle
   This book is different then others. When I read that book I had already few years of experience in Scrum projects, I was after a Scrum training and after reading other books related to the subject. My impression is that if someone is not familiar with the Scrum, he might understand the Scrum in a bit wrong way after reading this book. But...I think that everyone should read it. It really gave me another point of view on Scrum, showed that usually you might think that the Scrum is not for you when it actually is, shows that Scrum is also about flexibility and adaptation.