The Power of Three: Part 2: Complex Mechanical/Adaptive Systems

In the Power of Three: Part I: Mental Models, I distinguished between complex systems and complicated systems. Both systems have many interacting parts, but a complex system is what remains after all irrelevant parts have been weeded out of a complicated system.

I now want to explore further the various types of complex systems. First there are two broad categories of complex systems. These are complex mechanical systems and complex adaptive systems.

Complex mechanical systems are systems whose many parts and their many interactions remain fixed or which have a very limited range of change. Most complex machines are complex mechanical systems and hence the name. But not all complex mechanical systems are machines. For example traditional command and control organizational structures that characterized armies and business until recently are complex mechanical systems. Complex mechanical systems can be understood by understanding how the individual parts work and how they interact with each other in producing a limited amount of behaviours. For example, consider an airplane. An airplane has many thousands of individual interacting parts; some of these parts are themselves made up of several hundred subparts. Nevertheless, a few interactions on some of the parts are transmitted to many other parts in a predetermined manner so as to allow a trained pilot to fly the plane. The trained pilot can rely on a dashboard, which reports the state of some critical parts, and ignore the state of all the other thousands of parts, to determine what to do next. The pilot can ignore the other thousands of parts because their behaviour is completely determined by the few critical parts whose state is displayed on the dashboard. If something goes wrong in a complex mechanical system, by which we mean that it does not behave as expected, the root cause can be determined, the defective part can be repaired or replaced and the system is ready to start functioning as expected once again.
To summarize then, if we understand the individual parts and the interactions of the individual then we can understand and therefore control the behaviour of the complex mechanical system as a whole.

Applying the principles of complex mechanical systems over and over again has created the most advance machines and most of the organizations structures that our society daily depends on.

Many business managers, and nearly all management consultants and business academics believe that organizations are complex mechanical systems. Hence their penchants for developing dashboards, scorecards (balanced or not), and root cause analysis and their ten rules for this and their sixty rules for that.

However we are reaching the limits of what we can next create with complex mechanical systems. Sure we can create a billion more variations of the millions of complex machines we have already created but that is what they will be: variations of what we already have. To create whole new ideas, products and services, not simple extrapolations of what we already have, we will have to look for a new mental model. That new mental model is derived form studying complex adaptive system. And organizations, which act as complex adaptive systems, will not need dashboards, scorecards, root cause analysis or the ten rules of this and the sixty-five rules of that to be well managed. Instead they will need the tools and rules that come from complex adaptive system. And these will usually be the Three Things, but more of this will be explored in a future posting.

The Power of Three: Part 1: Mental Models

In his book: the One Thing You Need to Know, Marcus Buckingham (link) presents his conclusion that even the most complex topic can be reduced to a core concept or insight. This One Thing is all that really matters and all else is either details or distractions. Now that is a very powerful idea. Instead of wasting time considering all that I need to know to successfully deal with a particular issue, be it leading, managing, investing, or whatever, I only have to discover the One Thing, and then apply that one thing. In this over complicated, information overloaded world, this simplification is truly a welcomed relief. And yet perhaps, to paraphrase Einstein, this is the case of having made the world not as simple as possible but crossed over into having made it simpler that its. Topics, at least those that are important, are unfortunately complex. Note I said complex not complicated. A complex topic has many necessary interacting parts; a complicated topic is one with many unnecessary parts. Most topics when first encountered appear to be complicated because we lack the knowledge to distinguish what is relevant and want is not. However, once this knowledge is gained, and all the unnecessary parts are removed, we are often left with a complex but uncomplicated topic.

In my experience, to deal with complex topics what is needed is not the One Thing, but the Three Things. Three clear and simple things, that is.

Why the three things? Well because, each one of these three things can then be broken down into three things and so now there are 9 things and if each of these is broken down into 3 things then there are 27 things and so on. In so doing, any topic from the most simple to the very complex can be reduced to a coherently structured lattice of parts and their interrelationships. Some of you may already be familiar with this approach. It is has been applied successfully to decision analysis and goes by the name of decision trees.

You cannot do this with the One Thing. Pursuing the One Thing approach leads to only One Thing at any level no matter how many times it is applied, and therefore cannot possibly capture the complexity of the thing. (See picture to the right). Some will say that I am missing the core point of the One Thing. They will say: you start with the One Thing and then you build out as many branches and sub branches as are required. This allows you to replicate the complexity of any topic, just like the Three Things. This is in-fact what Buckingham does in his book. He first identifies a core One Thing for each of the topics he is investigating and then attaches many branches to the core thing.

There is, however a very important difference between having a core thing with many branches and a lattice of sequential Three Things. The One Thing simply shifts all the complications one level down to the numerous branches and as with any complicated thing, makes it very difficult to understand and remember. The Three Things lattice provides a coherent simple structure of interconnected parts that is easy to remember, and easy to use. As Charlie Munger said, “If the facts don’t hang together on a latticework of theory, you don’t have them in usable form.”(Link)

At this point you may ask why Three? Why not Eight things? Well I know it is more than One and anything more that Five is very difficult to carry in one’s head. Three works for me.

Below I describe the first level of Three Things lattice for various topics:

Investing: to invest successfully you need to:
1. identify valuable companies (good management, wide moat)
2. buy valuable companies selling at a discount
3. sell when companies can no longer sustain their value.

Leadership: to lead others you need to:
1. gather and keep followers,
2. keep rivals in check,
3. adapt to change

Project Management: to deliver a project on budget and on time you need to
1. well define the scope,
2. specify the time line,
3. secure necessary resources

People Management: to successfully manage others you need to
1.. focus strengths
2. leverage strengths
3. catalyze strengths

Risk Management: to avoid losing value you need to
1. determine the risk (amount and odds)
2. take or leave the risk
3. control the risk if you take it

In future postings I will build the lattice for each of these.