Little’s Law is a powerful tool that relates the amount the work a team is doing and the average lead time of each work item. Basically there are two main applications involving either 1) the input rate of work entering the team, or 2) the throughput of work completed.
In previous posts Applying Little’s Law in agile games, Why Little’s law works…always) I already explained that Little’s Law is exact and hardly has any assumptions, other than work entering the team (or system).
This post demonstrates this by calculating Little Law at every project day while playing GetKanban.
The video below clearly shows that Little’s Law holds exactly at every project day. For both the input rate and throughput versions. Throughput is based on the subclass of ‘completed’ items.
E.g. on the yellow post-it the product of lambda and W equals N on every project day.
The set-up is that we run the GetKanban game from day 9 through day 24. The video will show on the right hand side the board and charts whereas the left hand side shows the so-called ‘sample path’ and Little’s Law calculation for both input rate (yellow post-it) and throughput (green post-it).
Sample Path. The horizontal axis shows the project day running from 9 till 24. The vertical axis shows the work item: each row represents a item on the board.
The black boxes mark the days that the work in on the board. For example, item 8 was in the system on project day 9 and completed at the end of project day 12 where it was deployed.
The collection of all black boxes is called a ‘Sample Path’.
Little’s Law. The average number of items in the system (N) is show on top. This is an average over the project days. Here W denotes the average lead-time of the items. This is an average taken over all work items.
Input rate: on the yellow post-it the Greek lambda indicates the average number of work per day entering the system.
Throughput: the green post-it indicates the average work per day completed. This is indicated by the Greek mu.
Note: the numbers on the green post-it are obtained by considering only the subclass of work that is completed (the red boxes).
References
[Rij2014a] Why littles law works always
[Rij2014b] Applying littles law in agile games