Black Swan Modeling

BLACK SWAN MODELING

 

Swans are white.  At least that’s what we all thought before the discovery of Australia.  “People in the Old World were convinced that all swans were white, an unassailable belief.”  Why?  Because it was completely confirmed by empirical evidence.  Every single swan that had ever been seen had been white.  Bingo.  So all swans are white.  Then the sighting of the first black swan— a single observation completely invalidated a general statement derived from thousands of years of sightings that had confirmed over and over and over again that swans were white.

 

This is the way Nassim Nicholas Taleb opened his book, The Black Swan (2007) to highlight a several cognitive biases.  We have a bias to draw generalizations and then to confirm them.  When we do this we close off our minds to other possibilities.  We have a bias to predict the future based on our confirmation bias.   We have a bias to think that what we know is all there is to know and to be over-confident in that knowledge.  We have a bias to not question that we may not know something.  And all of these biases add up to the fact that we, as human beings, are consistently being surprised and caught off-guard by outliers, the unsuspected, the Black Swans.

 

9/11 was an outlier Black Swan, so was the fall of the Soviet Union (1991), so was the more recent world-wide financial crisis (2008), so was the election of Donald Trump (2016).  We also have a bias to concoct explanations after the fact that make it seem explainable and predictable so that we do not go forward thinking that it will happen again.  Until it does.  And it always does.  Nassim Taleb writes:

“What is surprising is not the magnitude of our forecast errors, but our absence of awareness of it.” (p. xx)

“Contrary to social-science wisdom, almost no discovery, no technologies of note, came from design and planning– they were just Black Swans.” (p. xxi)

 

Taleb notes that we lack a particular meta-learning.  We do not learn that we are not learning from this.  We live in a world of uncertainty and unknowns.  But we develop knowledge about what we know and what we think is certain.  We hardly have a way to even talk or think about the unknown.  A Catch-22!  If we knew it, it would not be unknown.  This is one weakness of knowledge he points to, and using a notion from G.L.S. Shackle, he calls it unknowledge.

 

In a chapter about “The Problem of Silent Evidence,” Taleb says that “silent evidence pervades everything connected to the notion of history.”  By history he says this refers to “any succession of events seen with the effect of posteriority.”  That is, we create “history” as we look back on things and concoct a historical theory yet we do so while avoiding “looking at the cemetery.”   He says that this is not a problem with history as such, but a problem with the way we construct samples and gather evidence in every domain.  It is a problem of a bias that we have, that is, a systematic error that views what we’re looking for more than what is there.  And this rises from the confirmation bias: we have a bias to confirm, not to disconfirm (p. 102).

 

Along this line he asks about all of the talents and books of geniuses that were not preserved.  What does that missing or silent evidence tell us?  Talking about the thousands of manuscripts that are rejected by publishers, he asks about the hundreds of literary masterpieces that perish by rejection.   How do we take into account all of the great manuscripts that were never published and that you never hear about?

 

The Danger of Model and Ignoring Missing Information

Suppose you were searching for the secret of wealth creation.  Since this was one of my modeling projects in the early 1990s, I took special note of Taleb’s words.  “We look for traits of those who succeeded, but don’t look at the cemetery of those with same traits that failed.” (p. 105).

“Numerous studies of millionaires aimed at figuring out the skills required follow the following methodology.  They take a population of hotshots, those with big titles and big jobs, and study their attributes.  They look at what those big guns have in common: courage, risk taking, optimism, and so on, and infer that these traits, most notably risk taking, help you to become successful.  You would also probably get the same impression if you read CEOs ghostwritten autobiographies or attended their presentations to fawning MBA students.

Now take a look at the cemetery.  It is quite difficult to do so because people who fail do not seem to write memoirs, and, if they did, those business publishers I know would not even consider giving them the courtesy of a returned phone call.  Readers would not pay $26.95 for a story of failure, even if you convinced them that it had more useful tricks than a story of success.  The entire notion of biography is grounded in the arbitrary ascription of a causal relation between specific traits and subsequent events.  Now consider the cemetery.  The graveyard of failed persons will be full of people who shared the following traits: courage, risk taking, optimism, et cetera.  Just like the population of millionaires.  There may be some differences in skills, but what truly separates the two is for the most part a single factor: luck.  Plain luck.”  (pp. 105-106)

 

The mere presence of traits, characteristics, and even actions of those who succeeded at something may be like white swans.  That’s all we have seen—so far.  Yes those factors do seem to contribute to the success.  Yet for all we know, these same factors may also be present in those who did not succeed.  There may also have been some other factor— a Black Swan factor— that we have not taken into consideration.  Besides “luck,” the small number bias, other factors could be involved.  Therefore we need to ask, “What unknown factor/s have I not even considered?”  “What variable could be unknowledge at this point that our modeling is not considering?”

 

Before the theory of germs arose, doctors never even considered that washing their hands between surgeries could have any affect on the mortality rate in the hospital.  That was unheard of?  How ridiculous!  Today we know that germ theory was a key variable, yet it was unknown then.  It was silent information that could not even be considered.  When it was discovered and later confirmed, it was a Black Swan event.  No one expect it and it too the Medical community by surprise.

Comments are closed