Dancing with Chance

blackswanThe book Dance with Chance is written by three scientists each with contributions in the field of judgment and decision making.  They are Spyros Makridakis, Robin Hogarth, and Anil Gaba. I have particular respect for Robin Hogarth and his work on intuition. It is aimed at a popular audience so it makes a stab at that Malcolm Gladwell glibness. They are not stringing anecdotes like Gladwell or making them up like Jonah Lehrer, but they are trying to give you the strongest case and not provide you with too many confusing facts.

Their examination of amniocentesis provides an example of this.  They assert that the the new advice in 2007 from the American College of Obstetricians and Gynecologists was that all pregnant women, regardless of age or blood test results should have the more expensive test.  The expensive test is amniocentesis.  Actually, the recommendations would be more accurately interpreted as saying that:

  • All women regardless of age should have the option of diagnostic testing, such as amniocentesis.

Maybe they are asserting if it is an option, then everyone will insist on having it.  In this case, I doubt it–being stuck in the midsection with a really big needle–is not going to be done lightly based on nothing.  In the book, they then provide the case for amniocentesis for a 25 year old woman with no indicated family history.  This is a straw man.  They do not even mention the much greater risk of Down’s syndrome and certain other issues with increasing age.

I happened upon this five minute video that includes Nassim Taleb, author of the Black Swan and Fooled by Randomness along with the three authors of the book.


It provides a quick summary of a bunch of books. Taleb, as you might expect, takes control and points out the Dance with Chance idea that regular medical checkups have not been proven to increase life expectancy. He calls financial experts psuedo experts.  He seems to call medical doctors psuedo experts and mentions doctors killing patients.  The authors are not so strong and mostly say that we should understand that we only have an illusion of control and that we underestimate the role of chance in our lives.  Taleb mentions the generalized expert problem. The authors say that we can reduce this by delimiting the range of expertise; knowing what we can forecast and what we cannot forecast; requiring experts to put their past predictions with their current predictions to increase transparency; and tempering our unwarranted respect for huge models that require us to forget the many assumptions made to create them. Taleb keeps referring to type 2 error as the problem.  I think he means type 1 error.  Type 1 error is the one where doctors think that they can help you with a new procedure that does not really work.  Type 2 errors are false negatives.  Maybe Taleb did this on purpose to show you that his expertise needs to be more narrowly defined.

The video provokes the following comments from me:

Financial experts are not necessarily psuedo experts. They can still help you, even though they cannot predict the future accurately or pick the stock that will go up.  They can make certain that you know that you cannot do it either.  Doctors are medical experts and they can help you in the same way.  Your team needs a medical expert.  You do not have to do what he tells you.  Mass screenings where you are screened for one thing at the grocery store are much more prone to abuse.  Your checkup doctor works for you and can help you sort things out.  Maybe a regular checkup should be done every two years instead of one to be cost effective.  Hygiene and sanitation did not increase life expectancy until it reached a certain level.  Then it made a big difference.  Not having your own doctor and then getting sick makes it very difficult to sort out information. Once you do get sick,  every specialist has something for you. One has radiation, one has surgery, one has chemotherapy, one has pain management. You need someone to put it together.  You need someone to give you quick referrals.  You need an advocate.  These four guys have danced with chance successfully and may not realize how lucky they are.  Moreover, they probably each have their own personal physicians.

Once they get past trying to convince you, I believe that their arguments are strong. They say that there are two types of basic decisions and four ways of making them. There are repetitive decisions and one of a kind decisions.  The four ways to decide are blinking–going with your gut, thinking–going with the most analytic deliberative process, sminking–using simple models or decisions rules, and going with true experts.

Blinking is great for repetitive decisions that must be made quickly–like hitting a tennis ball.   Blinking is good for Grandmaster chess players, but even then a little thinking after blinking is helpful.  Sminking is great for repetitive decisions. Find a good rule and you will make better decisions.  However, rules can become obsolete in changing environments so you need to reevaluate your rules from time to time.  Experts can help you come up with your own decisions, but you need to be careful as experts are often overconfident and wrong especially when they venture slightly outside their fields.  The authors want you to remember their “sminking” as it not only acknowledges the illusion of control, but it embraces the paradox of control by relinquishing control to a simple model rather than your own thought processes.  You gain more control of the outcome by letting a simple model or rule take over.

1 thought on “Dancing with Chance

  1. Pingback: Superforecasting - Judgment and Decision Making

Comments are closed.