It was always interesting for me how some of the world's most complex board games have a relatively simple set of rules. The rules of games like chess, checkers and go can be printed on one page of paper and yet these games have enormous strategic depth. We play them for many centuries and still find them infinitely enthralling.
Out of small number of simple rules complexity emerges. This is somehow due to the very nature of our universe. Our world is built from simple rules and patterns. If we look around we can see them everywhere. One of the most illustrative examples of this is swarm intelligence of insects such as ants and bees. The complex behavior and incredible precision of their colonies are rooted in a set of relatively simple rules. Ants for example, exchange chemical signals with each other to point out whether nearby ants need to do less or more of their current activity. This leads to the whole colony to be in sync and regulate the work of each ant without a central control system.
There are many more examples of simple rules at the core of complex systems in nature. The foundations of this behavior are studied in mathematics. From the theory of cellular automata to the study of fractals we can see how incredibly sophisticated and unpredictable are the worlds generated by basic rules. Interestingly, we can find many examples of these abstract mathematical concepts in the nature itself:
Simple rules and self-organisation
As we have seen in the case of ant colonies, simple rules can lead to self-organisation in systems with large number of elements. However, it is not always the case, certain simple rules result in unpredictable and chaotic output. Take the famous 3N+1 problem as example, this theory can be explained to a primary school students, however the best scientists are failing to find a pattern that would let them prove or disprove it. To achieve synchronisation and self-organisation we cannot use arbitrary set of rules. Rules must be designed in order to reach our desired goals.
Unlike abstract mathematical settings it is not always easy to simulate the behavior our rules will result in. Especially, when it comes to other people to implement them. Below are some common ideas that can guide us towards better rules and stimulate self-organisation in our day to day work.
Governance over control and command
Good set of rules are the ones that deliver governance rather than aim to control individuals. They set clear borders within which work can be done autonomously. As opposed to the control and command approach, where every step is tightly specified.
For example, take a policy for requesting day-offs in a company. A control and command way of doing it will be having a precise step-by-step instruction that must be followed by an employee taking some time off. It might include rules around how many days in advance the request should be made as well as rules about who should be receiving and approving the request. As we know, this kind of rules rather create a negative outcome and unpleasant friction in a company. This is mostly because the reasons and the length of breaks that employees need vary a lot and one simple set of instructions cannot be good for all scenarios.
A governance approach to the same policy could be a declarative agreement in a company that everyone aims to organise their holidays in a way that it creates minimum disruption for others. Now, it's the responsibility of each employee to think about best way to inform their colleagues about their plans. At the same time the policy has set a border by defining that holidays should create minimum disruption for others. This way we have a self-organised approach with a shared goal to avoid operational interruptions.
Principles over processes
Principles are more important than the actual processes. They define core ideas around which processes should be built. The key to consistency is to rarely change principles while constantly improving and modifying the processes with a goal to implement those principles in a better way.
Let's take software development lifecycle as an example. A rarely changing process would most certainly cause a major slowdown in an actively changing environment, such as a startup. A well tested two weeks long sprints and ceremonies around them may work great for a team in a pre-release stage. Then suddenly after launch, they may find themselves in a different environment with a large amount of incoming change requests that cannot really wait for weeks of planning and coding. They decide to adopt a kanban approach as they see it as a better process for this time given the new mode of working.
While the process has changed for the team, the core principle of agile development "responding to change over following a plan" is still maintained and their work style is consistently agile. However, if at some point they decide to stop everything and plan every detail of their work for the next six months and then follow that plan no matter what, this will be a significant change in their principles and will lead to a major change in the organisation overall.
Keeping principle fixed and processes flexible helps teams to self-organise and constantly shape the process towards the most efficient way of working.
Less rules more probability that they will work
As with the board games, less rules will usually make it much easier to remember and implement them. Keeping the list short will also make it easier to distribute the rules.
We can have a long list of qualities we want to see in our new hires. However, during an interview it will be impossible to consider all of them and keep them all in mind evaluating the answers from candidates. A better way would be to have just three or four key qualities and values shared across the company that are easy to spot during interviews. This will serve a much better signalling system to the interviewers than a long list of questions.
Making rules simple
Simple rules are the rules that are easy to understand. There should not be many different ways of interpreting them. They have to be communicated using the concepts that are shared and understood in the same way within the organisation.
The rules usually have two parts: the condition and the action. When the "condition" is true, do the "action". To make rules simple, we need to ensure both parts are clearly defined. The most straightforward conditions are always and never. If your rule can use these, go for them. However, we might need to define a more specific condition. In that case we need to make sure it is easy to recognise whether the condition is met or not. Same applies to the actions, they should be easy to apprehend and perform.
As we can see simple rules do not only create complexity, but they can also be used in order to manage and govern it. Many complex problems can be solved by using a few well-chosen principles. Finding these silver bullets is, however, something that requires a lot of creativity.