Wednesday, February 27, 2013

Calculating Risk

Calculating Risk 

I have been doing a ton of reading about Risk Management recently.  I find the concept intriguing.  Risk has been a mathematical study for hundreds of years.  It is intimately tied to statistical analysis and mathematical concepts including Game and Chaos theory.  Risk is not only for gamblers, bankers and insurance dealers, it is a useful tool for assisting Quality Management in the decision process.

I suspect that the recognition of risk is innate in all, or at least most organisms.  Certainly cats seem to go through a process of doing a full sensory scan when confronted with something that resembles food and then make a decision to eat or not eat.  One sees the same behavior with fish, when they look at bait and swim by.  I don’t know for certain what is going on their mind, but it sure looks like some form of risk assessment. 
I suspect that the human reaction to risk has a huge overlay that distorts that innate process, sometimes in the direction of being Risk Seekers, but more often as Risk Reduction, and sometimes as Risk Aversion.  Recently I had a meeting with some folks with Transport Canada who were very concerned about the level of risk associated with putting a small safely packaged box in an airplane that contains tiny amounts of bacteria for proficiency testing.  Imagine is there were an in-air accident and my box crashed to the ground and my  100,000 bacteria were released.  The risk would be severe.  True, but I wonder how that would compare to the effects of the metal of a crashing airplane or the 1000s of litres of jet fuel, or the trillions of bacteria from each of the damaged human bodies. 
After doing a lot of reading and reflecting I began to appreciate that Risk is something that every Qualitologist needs to know about.  Here is the essence:

1.    Risk is the effect of uncertainty on an outcome where uncertainty is the product of incomplete information.  The more information the more one can calculate Risk, and by corollary, the less information that is available, the less certain one can be about Risk assessment. 

2.    Chaos theory says that for every effect there is a cause, but the cause may be so subtle that it is indistinguishable.   There may be many causes that impact on the effect, making Risk predictions increasingly difficult.

3.    What Donald Rumsfeld called “unknown unknowns” Frank Knight called “unmeasurable uncertainty”.  Both make Risk prediction very difficult. 

4.    Some risk is inevitable because the “unknown unknowns” can have an effect at any time.  Setting a policy of risk aversion will always fail.

5.    Game theory says that accepting  a risk level that avoids a big loss is more likely to be more successful than accepting a risk level that might lead to a big win.

6.    As much as we talk about Risk in calculated percentages, Risk is at best semi-quantitative estimate. 

7.    The single most effective tool in the Risk toolbox is the creation of a Severity – Occurrence estimate.  Outcomes that may be catastrophic in effect and occur frequently should be avoided,   Outcomes that are negligible or trivial in effect and occur rarely if at all should be considered as not a problem.  The challenges fall somewhere in between.

8.    But the greatest risk occurs when people who don’t have the same innate ability as a cat or a fish use tools they neither know or understand and  try to make predictions they cannot support or justify  based less on information and more on incompetence.  
Me bitter?  You betcha. 

No comments:

Post a Comment

Comments, thoughts...