A New Copy!

One note to start with: we are on Chapter 14 today out of 17. So, after today, we have approximately four more weeks. As a result, we will have to choose another book in the next couple of weeks.  I have received some suggestions, and I have also asked the interviewees that appeared in the Software Process and Measurement Podcast in 2017 which was the most impactful book they have read. I would also like your input. What do you suggest that we read next?  

Chapter 14 is titled Introduction to Forecasting in Daniel S. Vacanti’s Actionable Agile Metrics for Predictability: An Introduction (buy a copy today)

One of the definitions of predictability is the ability to make a quantitative forecast about the future state of a process. In this light, a forecast is just a calculation about the probability of the occurrence of some future event; an estimate might be a forecast. At one point in my career, the group I was part of had to collect data on a nightly file maintenance process so we could determine whether the process could finish within the required time window.  

Vacanti suggests that all forecasts must be communicated as a range and a probability. Forecasts are not certainties.  The importance of how a forecast is communicated cannot be understated.  When we forecast file maintenance completion times, senior executives boiled the result of the study down to an average on a PowerPoint slide.  Typically we could make the window, however typically wasn’t all the time and there was quite a bit of  (negative) excitement when we occasionally missed it.  Remember: range and probability. I would also suggest adding a scatterplot to ensure proper communication.

The second major topic in the Chapter is a reminder that Little’s law cannot be used to make a prediction about the future. One the issues with using Little’s Law to forecast is that we can not know which of the assumptions that underpin the law will be adhered to in the future. A team or organization that violate the assumptions of Little’s Law invalidates the exactness of the law.  Exactness (another word for variability) is required for predictability.  A second issue is that nagging concept of distribution.  In most cases, we really do not have a great handle on how the data are distributed (and the future distribution). Normal distributions are not as common as we would wish.  The one caveat to this discussion is that Little’s Law can be used as a rationality check on predictions.

When we do have an understanding of the distribution (let’s say we have plotted the data in a scatter plot). Making a prediction is fairly easy.  The 85% line drawn in previous chapters is a tool for making a prediction.  For any item, we can say there is an 85% chance that the team will complete the story in the same or less time than indicated by where the line is drawn. The amount of variability in the distribution provides the range and the line indicates the probability.

The final major topic in this chapter is a discussion of the foibles of straight line predictions.  Vacanti uses the example of a team that averages completing ten stories per sprint and has 100 stories in the backlog. A straight line prediction would be that the work will be completed in 10 sprints. Why is this fraught with potential issues? First, it is built on averages. Secondly, there will be variability in backlog (and therefore throughput). Third, valid predictions require a date range.  Arguably a fourth would be the lack of a probability.


Forecasts are predictions of the future. All predictions of the future are uncertain.  The cycle time metric coupled by a scatter plot is a good mechanism to support making a predictions.

Previous Installments

Introduction and Game Plan

Week 2: Flow, Flow Metrics, and Predictability

Week 3: The Basics of Flow Metrics

Week 4: An Introduction to Little’s Law

Week 5: Introduction to CFDs

Week 6: Workflow Metrics and CFDs

Week 7: Flow Metrics and CFSs

Week 8: Conservation of Flow, Part I

Week 9: Conservation of Flow, Part II

Week 10: Flow Debt

Week 11: Introduction to Cycle Time Scatterplots

Week 12: Cycle Time Histograms

Week13: Interpreting Cycle Time Scatterplots

Week 14: Service Level Agreements

Week 15: Pull Policies

Support the author (and the blog), buy a copy of Actionable Agile Metrics for Predictability: An Introduction by Daniel S. Vacanti



Get your copy and begin reading (or re-reading)!