Build-Run-Improve-Repeat: that's the name, but what's the game?

In a previous post I explained how the DevOps game Build-Run-Improve-Repeat came to existence, where I got the inspiration from. In this post I will dig deeper in the game itself: what is the aim of the game, the learning objectives and how does it go?

Main learning objective

A common misunderstanding about DevOps is that the aim is to release new features faster, more frequently. This will be a consequence but it will not work if you have poor agile development practices - automated builds, continuous integration, test driven development and other test automation and so on. You will just release crappy quality faster. So the main learning objective of this game is to experience the impact of doing the right (or wrong) investments in improving your way of working.

A common misunderstanding about DevOps: if you have poor agile development practices, you will just release crappy quality faster

Playing the game

This is a board game so the entire game evolves around the board. The board has several spaces of which one is marked as “Queue here” and one as “Cash here”. All the other spaces represent an activity or an aspect of the different stages of the DevOps cycle:

• Plan
• Code
• Build
• Test
• Release
• Deploy
• Operate
• Monitor

The spaces on the board are actually placeholders for cards corresponding to these activities. In the previous post I wrote that I got to read the book Accelerate, which explains the research and the findings behind the annual Devops survey of DORA – DevOps Research and Assessment. I also wrote that the biggest inspiration for this game was a webinar Gene Kim did about technical must do’s for DevOps. After all, this game focuses more on the technical aspects of DevOps. Now, some of the things of the DORA-report did find its way to the game. This report talks about different performance levels:

• Low performers
• Medium performers
• High performers
• Elite performers, which are the top of the high performers

Something of these performance levels found its way into the game: the cards to be placed on the board correspond to a certain performance level, ranging from 0 – similar to low performer – to 3 – similar to elite performer.

Invest wisely. Don't spend all your revenue. You don't want to go bankrupt with the next incident.

Default you simply start at the lowest level and work your way up by implementing features and investing some of the revenue in improving your way of working. Or if you play this game with all participants of the same organization, you can start at the performance level corresponding to your organization – for each of the activities. For instance, the lowest level of Plan Approach is that you still work according to a waterfall approach, with big projects and big specifications up-front. If your organization still works in a project mode, but using iterations, you can start with level 1 for that aspect.

Implement features to create revenue

I already mentioned it: you need to implement features. How do you do this? You need to move the blue pawns from the "Queue here" space to the "Cash here" space. Rolling the die determines how many moves you can do, taking into account the possibilities/limitations of your performance level. If for instance you apply automated testing, you can immediately move to the next activity, because automated testing does not require extra effort during the test execution. You also need to respect the queue size and flow principles. In a waterfall approach for instance, your batch size is a lot bigger than in case of a Kanban approach (represented by the number of pawns you need to move together across the board to implement features), which means that it takes longer before your effort is converted into revenue. Not only does this impact the batch size, also, in case of waterfall projects, you only start implementing new features when the current batch is completely delivered (in other words: start a new project when the current project is done).

What can possibly go wrong?

And then things can go wrong… Various incidents can can occur, based on the roll of the die:

• A reported vulnerability needs to be fixed as quickly as possible
• A bug needs to be fixed as quickly as possible but also causes financial impact
• A security breach needs to be fixed as quickly as possible but also causes financial impact

To know the financial impact of a security breach or a bug you need to flip all the cards on the board:

The symbol next to “Incident impact” indicates what type of incident has an impact on this activity. For instance, you have a bug, then you need to sum up all the incident costs of cards that have a bug as incident impact. That is your total cost of this incident. And then you still need to fix it. That’s where the red pawn comes in. For each incident you need to fix – of whatever type – you put a red pawn on the first coding activity on the board. You’ll need to move it to “Cash Here” as quickly as possible, again by rolling the die.

Invest to improve

You receive a revenue for each feature you have implemented – each blue pawn that lands on the “Cash Here” space. Part of that revenue can be invested in improving your way of working. Don’t spend it all, because you still need to be able to pay the losses of incidents. You don’t want to go bankrupt with the next incident. As far as these investments are concerned: initially the return on investment (if done correctly) is rather low. You won’t immediately get:

• a much shorter lead time
• a much higher deployment frequency (unless you do the wrong investments)
• a much lower change failure rate
• a lower mean time to recovery from incidents

But this is an important foundation to build upon. And that is exactly what the DORA report explains about the high performers becoming elite performers and the gap between the low and elite performers that keeps growing: they experience the benefits of their initial investments in a good foundation. And that's what you will experience in the game too, when you continue improving your way working - according to the right priorities.

Want to know more about this game? Then keep an eye on my news feed, send me a message or check my web site. I am really looking forward to doing workshops with this game - once the corona crisis is fading away...