Scrum: Theory

Scrum: Theory
By Alexander Boden

Scrum is based on empirical process control theory. This immediately makes it sound complex: however, if we break this down we’ll see why this is actually easy to use, understand and why this is an incredibly powerful methodology.

Empirical process control is achieved through the frequent inspection of the outcome or results of a process (set of activities) and adapting one’s actions based on the outcome of those initial set of activities.

To make this easier to understand let use an example. Say I have to dig 20 holes in a row. I may start with the planning and the digging of the first hole. Having completed the work I may have noticed that it was easier to dig in a certain way or that it took longer than Anticipated. This learning may change the anticipated time I thought it would take to dig all the holes or my tactics of how they would be dug, and thus the outputs of the first process have led to me changing the inputs to the following processes.

Now this may seem simplistic, but these types of feedback processes are often lacking in a lot of activities organizations undertake as their processes are too lengthy and the status of the project undertaken has already considerably changed once the often lengthy process of measuring and reporting the results of the activities has taken place.

Coming back to Empiricism – we now understand this asserts that knowledge is based on the experience we gain and making decisions on what we have learned. Scrum uses small feedback cycles to improve the predictability of the outcome of actions and to control risk better as the planning horizons are a lot shorter.

In order to implement an Empirical process you need to ensure the following:

• Transparency: Tasks/parts of the process need to be visible to those who are involved or responsible for their completion. What is visible needs to be defined by a common standard so that everyone involved has the same understanding of that which is displayed i.e. a common language related to the process and a common understanding of when a task has been completed.

• Inspection: The variance of tasks needs to be evaluated against desired outcomes: however, this should not happen too often so as to avoid it getting in the way of completing the tasks. This is best done by a skilled inspector with team where the work is done.

• Adaption: Should the inspector find that the tasks are deviating from the desired outcome to such a point where it will no longer be acceptable, then the process needs to be adjusted. Scrum provides four formal meetings that can be used initiate the change. These will be covered in detail later, but here is a short list of the meetings:

  1. Sprint Planning
  2. Daily Scrum
  3. Sprint Review
  4. Sprint Retrospective

In the next post I will be looking at the Scrum Team: looking at the different roles and responsibilities the team members need to fill.

You can read more from Alexander Boden on his blog.