Thinking Through the Chaos

Russell Mosier, Senior Vice President

A well documented, Decision Analysis based study is needed in only a small percentage of the many decisions that confront us.  However, the basic tenants of Decision Analysis can be applied to many of our non trivial decisions.  In other words, a "structured thinking" approach to making a decision.  I will use an example of buying a car to illustrate a structured thinking concept.

Step 1: Define Your Problem

This is generally the first thing one needs to do when making a decisions.  "What exactly am I trying to decided?

Be sure to write this problem statement down, I prefer pen and paper to computer for this.  I believe there are many cognitive advantages to writing something, but at the very least it is easier to remember what you were thinking.

So here is my first cut at a problem statement.  "I need to decide what new car to buy"

Step 2: Think about the CHAOS

This is the "Structured Thinking" part and CHAOS is a handy pneumonic.  CHAOS stands for Context, Holistic, Assumption, Objectives and Swing.  I sometimes think through these areas in a different order and most of the time I iterate.  However, since OCSAH is not a word, I will discuss them in the CHAOS order.

Context:  The context about the decision.  I think about these as the environmental factors around the decision.

Some questions you might want to answer are the W's. Who are the stakeholders?  Who is going to buy/drive the car?  Who is going to ride in it?  What do I need to do with it? Where am I going to store it?  Is it replacing a car or is it an additional car or a first car?  When do I need to buy this car?

You will notice none of these questions deal with the car itself, they all deal with things surrounding the car.

Holistic:  The holistic view is at odds with the analytic view.  That is, the analytic view (covered by the C_AOS) is analyzing the problem from specific angle or details. The Holistic view is taking a step outside the system you are evaluating.

First think about what scenarios are you going to use this car.  Commute, transport kids, recreation, vacation...

Some questions you might want to answer are:  What need is the car fitting within those scenarios?  Transportation of things or people or both?  What am I not able to do now that I would like to be able to do?  Can I imagine my life without a car (or Do I need to buy a car at all)?  What must I still be able to do after I make my decision?  Do I want to move so I don't need a car?

This is the time to stretch the bounds of your decision.  You might grow broad and then narrow back down, which is fine.

Assumption/Constraints:  This is the time to document, and more importantly to challenge, the assumptions you are making and challenge the constraints you are putting on the potential solutions.

I need to decide what new car to buy. 

Looking back at my problem statement, the core of the statement is "Need..Car."  There are other assumptions I put in, probably without thinking.  Do I really need to "buy" a car?  What about lease, or Zip Car or rent, or borrow.  Does it have to be new, would I consider Used?  You could even challenge whether the solution needs to be a car at all.

The point here is not to solve the problem and make the decision but to be aware of the assumption and constraints you are making on the decision.  It is fine to have the assumptions and constraints in your problem statement but they should be intentionally put there. 

Objectives:  What am I trying to achieve?  How do I define "success" for this decision?

These objectives are not the broad overarching ideas you might address in Context or Hollistic.   These objectives are what do I want from the car itself.  Performance, Reliability, Safety, Style, Durability, Cargo Space...  Next step would be to take this further in detail.  What do I mean by Cargo Space?  These should tie back to the Context discussed above.  Don't forget to include things that are not quantifiable.  Just because you can't count it, doesn't mean it doesn't count.

Don't get into defining "How" to achieve those objectives yet, that is the topic of another article

Swing:  This is often overlooked but explicitly but we all take into account implicitly.

What is my trade space?  What is the absolute minimum and maximum of something I will accept.  Would you be willing to trade below the minimum for "X" capability, it might not be the minimum then.

Step 3: Re-define your problem statement based on what you just learned.

Chances are after you thought through your problem in a structured way you will want to expand or shift what you were going after.

In summary, Define the Problem, Define the CHAOS, Re-Define the problem.