On Wardley Maps Imperfection

theory

#1

Maps are imperfect!

That statement often haunt even very experienced mappers. There is no such thing as a perfect map, and attempts to create one are often futile. The only thing that matters is whether your map is useful for you, whether it helps you to solve whatever problem you face. That’s the only metric you should care about. Your solution, your action, your decision is important. The map itself - not very much. I would dare to say it is a by-product.

There are five major sources of maps imperfection:

Abstractions & Symbols
Have you ever thought what exactly ‘watching TV’ means? Does it require a colourful screen? A broadcasted signal? A special receiver? If you are watching BBC on your phone that is send to you over the Internet, are you watching TV? Is it really so much different from watching BBC delivered by your Cable Company? Every person has a slightly different view on different things. ‘Watching TV’ are just two words, two symbols that trigger a specific meaning. As long as people mean the same things using the same symbols, communication can be quite efficient. But if you ever witness people with Eastern and Western cultures discussing Democracy, you will find they mean different things. The bigger are the differences, the more difficult is the conversation. While maps help you to be explicit about your assumptions, there is still a lot of subjectivity involved.

Size & Generalisation
Depending on how do you define a component, you may end up with a map with a large numbers of nodes. Which will be accurate, but impossible to use. The noise will hide the important bits. Therefore, you have to modify your map and adjust the granularity having purpose in mind. Your map will become less detailed but more useful. You can easily miss something. It is a tradeoff, and a subjective one. Double imperfection!

Purpose & Perspective
The importance of components changes depending on the purpose of your map. You do not map a situation for the sake of mapping, but for the sake of making a decision. The choice you face determines what is relevant. Since, again, it is all subjective, there is no such thing as an objective map. The best you can achieve is a group consensus “This is how we think the situation looks like.”.

Uncertainty & Complexity
Finally, uncharted components do not have clear borders. And it is difficult to measure their exact level of Evolution. More, you probably have not understood all the dependencies between your components, and some of them have not been identified by anyone yet. Some of them will manifest themselves through second-order effects, and it may be economically unjustified to test the system for every possible, unexpected consequence.

To sum this up - maps represent our understanding of the situation, our situational awareness. Situational awareness is not the purpose per se, what you want and need is to use it to improve your situation. Therefore, it needs to be good enough, not perfect.


#2

Thank you for outlining the four aspects. This is very useful.

Would you be able to expand on second-order effects or provide any pointers?

Secondly, maps are a collaboration tool. Maps of a scenario will evolve over a period of time as new information comes to light and situations change. Do you think it is important to keep a journal of assumptions, generalisation rationales, deliberate omissions, etc? I’m thinking a reference-driven mapping approach will generate documentation, shared learning and possibly historical insights.


#3

I think the story of wolves in yellowstone is a nice example of a second order effects. Naive advice would be to analyse the impact of wolf reintroduction, but I am not sure if it is even possible to figure out how the ecosystem would change. We may not know nor understand all the dependencies (f.e. how wolves affect river flow).

In business, a second order effect example, I’d say, is AirBnB in Warsaw. 1% of owners controls 25% flat for rentals. It is no longer sharing economy and it hits traditional rental services. I doubt it was expected to work that way.

I also totally support running a log of events related to the map - key assumptions needs to be verified, but it is also possible that circumstances change - some assumptions being false today may turn true tomorrow.