Skip to content

Understanding Complexity (and what to do about it)

2012 August 12
3D_Cantor_set (1)

People like to say, “keep it simple, stupid.”  It’s a down-home remedy for our overly complex, technology infused modern life.  Like much good advice, it is often given, but rarely followed.

The problem is that simplicity is not so simple.  We live in a complex universe where much that happens is beyond our control.  Merely wishing things to be simpler does not make it so.  In fact, making facile assumptions often leads to disaster.

How we deal with complexity determines how we innovate, build organizations that can compete effectively and navigate an increasingly technological marketplace.  We need to take it seriously, not gloss over it.  Fortunately, this has been an area of intense study since the beginning of the digital age and there are some basic principles that can guide us.

The Complexity of an Entity

The idea of complexity is fairly new.  It really began in earnest with Claude Shannon’s 1948 paper that spawned the field of information theory.  It established, among other things, a measurement unit for information – the bit– a binary piece of information.

That led to the first serious thinking about the subject.  The Kolmogorov-Chaitin principle, formally defines complexity (and therefore simplicity) by the number of bits that it takes to describe an object without losing information.

For instance, a googol is a very large number but simple to describe (i.e. 10100).  While 479,001,599 is much smaller, it is also a prime number that can’t be reduced to anything simpler, which makes it extremely complex.  That’s why large prime numbers are used in encryption, because we want our transactions to be hard to decode.

There have been some other approaches, but the general theme is that complex things take a lot of information to communicate; simple ones do not.

Complexity of a System

Simplicity, however, doesn’t just apply to individual entities, but systems as well.  Simple systems are fairly easy to understand.  They tend to look like this:
 

 
Statisticians use curves like these so often that whenever data follows this pattern they refer to it as normally distributed.  We know a lot about systems like these and can predict much about them.  For instance, if we have a relatively small sample of people’s heights, we can know a lot about the heights of other people.

Unfortunately, most real world systems don’t work like that, because they include feedback.  While your height won’t affect mine height, your use of e-mail will.  The same goes for your purchase of a house, your enthusiasm for a brand and so on.

That makes most of the things we interact with in the real world considerably more complex than the ones you’ll find in statistics textbooks.  These type of systems are governed by power law distributions that look like this:
 

 
Unlike so-called “normal” systems, power law systems are problematic for two reasons. First of all, a few large entities drive the system, so an average value means little.  It makes no sense to talk about an “average” social network when Facebook has almost a billion members.

Secondly, while normal systems quickly degrade at the margins, power laws have “long tails,” so there are no possibilities that we can entirely dismiss.  The probability of a Justin Bieber or an Instagram may be very low, but unlike the possibility of a ten foot tall man, we have to take them into account.

Emergent Complexity

Probably the most confusing kind of complexity are systems that are emergent.  The most famous and telling example is the Mandelbrot set:
 

 

While the Mandelbrot set is amazingly intricate (click to enlarge and zoom in), it is actually a fairly simple sequence repeating itself:

zn+1 = zn2 + c

By the formal definition of complexity it’s very simple, because we can describe it with very little information.  However, as the sequence repeats itself we get a very complex structure that we only know how to simplify because we know the starting point.

Emergent complexity explains how  the relatively information poor human genome (about 800 MB) can create a brain which supasses the computational capacity of the world’s most powerful supercomputers.

Scientists (and certain types of investors like Elliot Wave theorists) spend their careers dealing with this type of complexity, taking systems that have built up over millennia and trying to factor them down to a few simple factors.  When they are successful, they are called geniuses (i.e. E=mc2), when they are not we call them quacks.

Dealing with Complexity

As Einstein said, we should make things as simple as possible, but no simpler.  Ignoring complexity won’t make it go away.  However there are some strategies we can utilize that can help keep the mess manageable.

Factor Down:  The first place to start is with entities themselves.  Acronyms and buzzwords can be convenient within organizations and communities which reuse the same terms constantly, but they are a disaster when communicating with a larger audience.

A good rule to use is Wittgenstein’s principle: if you can’t communicate in a common language, you probably don’t know what you’re talking about.  If you are confusing others, chances are that you are confusing yourself as well.

Minimize Options:  Although we like to have choices, they also make things more complex.  Therefore, it often helps to minimize your options, especially in negotiations.  As I’ve explained before, someone with a gun to his head is in a very strong negotiating position (with people other than the gunman, of course).

In his book, The Paradox of Choice, author Barry Schwartz shows that this principle applies in much less life threatening situations as well.  When confronted with fewer choices, consumers tend to buy more.  37 signals built a software company with a cult-like following by stripping away features.

Compare an Apple product to any of their competitors and the first thing your notice is it has fewer buttons and doodads.

Stay Robust:  Nicholas Nassim Taleb, the bestselling author of The Black Swan, has thought profoundly about complexity and how it affects our everyday lives (not to mention the recent financial crises).  He suggests that the best way to deal with complexity is to stay robust enough to survive the volatility that comes along with it.

Certainly, Jamie Dimon and JPMorgan Chase benefited from a stronger balance sheet during the financial crises.  While other banks increased risk when times were good, they stayed conservative and were able to buy up assets on the cheap when the bust came. Later, when their London Whale trade went horribly wrong, they absorbed it easily.

So the key to keeping things simple is to tackle complexity on every level.  Keep entities simple, but understand that once they start interacting with each other a new more complex kind of order will emerge.

Keep simple what you can, make allowances for what you can’t.

- Greg

17 Responses leave one →
  1. August 12, 2012

    Our economies are more complex because we keep on creating structures and using structures that defy Adam Smith’s pricing principle.

    Here is my list:

    These are:
    Mortgage Finance,
    Related Regulations
    Taxation of savings and debt and investments
    Index-linked Bonds
    Fixed Interest Bonds

    Mortgage payments jump around, mortgage sizes are not properly regulated to control interest rate risk, Taxation re-distributes wealth in unstable ways as interest rates change, Index-linked Bonds are linked to the wrong index, and Fixed Interest Bonds cannot adjust to rising incomes, which is a primary source of rising demand.

    By structuring everything in this list so that it can rise in response to rising demand at the appropriate rate, including the cost of mortgages, (not jumping around), we can take so much instability and fragility out of our economies that we will ask ourselves why we did not think to do this before.

    That is what Bankers usually say to me: “Why has no one thought of this before?” Indeed. A good question.

    For further reading and a video describing a really simple mortgage try my blog

    http://macro-economic-design.blogspot.com

    Make life simple – help me to spread the word.

    Thanks Greg.

    [Reply]

  2. August 12, 2012

    You’ve made it simple and clear, Greg, and I really appreciate it. I am working on a new approach to planning for marketing, and your three strategies are like lighthouses on a stormy night at sea. Thanks for your deep thinking and hard work, and for sharing the fruits.
    Graeme Roberts´s last blog post ..Neil Gaiman Makes Art

    [Reply]

    Greg Reply:

    Thanks Graeme. That’s very kind of you.

    - Greg

    [Reply]

  3. August 12, 2012

    Greg,

    The fusion of tech and design remains an area where simplicity is critical but seldom used. John Maeda’s work remains profoundly true. Just because you can add doesn’t mean you should. As true for enterprise software as it is for the next mobile phone.

    Nice article mate.

    [Reply]

    Greg Reply:

    Thanks Hilton. Have a great Sunday!

    - Greg

    [Reply]

  4. August 12, 2012

    Dear Greg,

    I am the editor of CEOCIO China magazine and a big fan of Digital Tonto. I have been inspired by the excellent viewpoints here. CEOCIO China focuses on business and management topics in the hyper-connected era. Might we have the honor to reproduce your blog article in Chinese in our magazine?

    Thanks a lot.

    Jerry Zhanren Yue
    Editor
    CEOCIO China magazine

    [Reply]

    Greg Reply:

    Xi XI Jerry. That’s very nice to hear!

    Feel free to reproduce it, but please give attribution.

    - Greg

    [Reply]

  5. August 13, 2012

    Greg,

    Nice read. It seems to me that the problem of complexity is more methodological than practical. The entities we deal with in informatics – bits, systems – were very convenient when we need to calculate but when we need to interact they are useless.

    Internet is the new business ecosystem where all entities should interact within their business activity and their evolution. The Bit is the entity to be computing, and the System is the entity to produce some (computing) functions. Analysis, reporting and monitoring are only the weak efforts to combine the bits with the time, places and circumstances.

    Basically, the entity should have the opportunity to interact. So, it should have the semantic part, should make a choice, choose a direction and make a decision. We know about Semantic Web, Neuro-Networks and Artificial Intellect but all their failure is that they don’t correspond to business activity.

    Information is the product of men’s activity. And it was a big discussion in 1940’ when Claude Shannon and Norbert Wiener regarded the information theory and cybernetics as computing: data calculating and data transmission. The information as a knowledge entity and semantic category was forgotten.

    Today we need new approach to semantic relationships in economy and Internet, we need new business format for information to provide the entrepreneurships. We need the bit with semantic meta-data, we need the systems as autonomic monads that can effectively interact, we need the automatic making decision based on Business Intelligence.

    So, we need the new business ecosystem within the Internet and complexity will vanish. Because to provide real activity is easier than to make sophisticated analytical mathematical calculations : )))

    [Reply]

    Greg Reply:

    Very good points.

    Thanks Sergei.

    - Geg

    [Reply]

  6. Ton Jörg permalink
    August 23, 2012

    A very interesting read. Me, myself, I am busy with the danger of misunderstanding complexity in complex organizations. The basic misunderstanding is that people want to simplify complexity without knowing what complexity is really about. To my mind we need to link complexity with network thinking (Barabasi, 2003). In my work I make a distinction between social complexity, relational complexity, interactional complexity, generative complexity and hypercomplexity. The misunderstanding of complexity is the confusion about these complexities and how they are connected. A theoretical confusion which is dominant is what interaction between entities is really about. We therefore need a different view of interaction which goes way beyond the mechanistic view of it. Our sciences are failing to offer the alternative view of interaction for various reasons, the main one being locked-in in the traditional Newtonian paradigm of action and reaction. Companies like KPMG, IBM, and the RBS have made reports about “Confronting Complexity” which show the disastrous effects of bad thinking. We really need new thinking in complexity to face the complexity of our world. You need complexity to deal with complexity. As simple as that. But a hardy perennial indeed for most scientists and CEO’s of complex organizations.

    [Reply]

    Greg Reply:

    Interesting ideas. Thanks for sharing.

    - Greg

    [Reply]

  7. August 28, 2012

    Or, as they say:

    Grant me the serenity to accept the things I cannot change,
    Courage to change the things I can,
    And the wisdom to know the difference.
    Paul Clarke´s last blog post ..Lightboxes

    [Reply]

    Greg Reply:

    Nice sentiment:-)

    - Greg

    [Reply]

  8. Ton Jörg permalink
    August 28, 2012

    For tackling complexity, one better start from the wisdom, expressed by a poet

    “In order to arrive at what we do not know
    You must go by a way which is the way of ignorance.”
    (T. S. Eliot, in “East Coker”)

    It is my impression that CEO’s of big companies like KPMG, IBM and the RBS doe not realize that. See their reports about “Confronting Complexity”, published last year.

    [Reply]

    Greg Reply:

    Thanks for sharing.

    - Greg

    [Reply]

  9. October 10, 2012

    HI Greg,

    I teach leadership to business people to inspire innovation. I really appreciate your article and have retweeted and summarized it. It would be great in the classroom classroom as well–cited.
    Thanks so much for a simply brilliant article.

    Elizabeth
    http://twitter.com/audaciousinnov8

    [Reply]

    Greg Reply:

    Thanks Elizabeth. That’s very kind of you.

    - Greg

    [Reply]

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS

CommentLuv badge