[PREV - A_FAMILIAR_DOOM]    [TOP]

BLACK_SWAN

                                             February 10, 2008

   Nicolas Nassim Taleb
   "The Black Swan:                                  Note: this page isn't
   the Impact of the Highly                          about Rafael Sabatini!
   Improbable" (2007)
                                                              For once I'm
                                                              talking about
Taleb's basic message:                                        something
                                                              contemporary.
We repeatedly encounter
extreme events that we
pretend were obvious in         And that sets us up for
retrospect.                     being blindsided by the
                                next extreme events.

The commonly used tools of
statistical planning rule out
the existence of these extreme
events, making a subtle
assumption that tomorrow will
essentially be like today.

       We over-apply the Gaussian
       distribution, and ignore the
       "Mandelbrotian": the
       Gaussian presumes that
       there's an average, and a
       scatter of values around            This rules out
       that average that gradually         extreme occurrences
       tapers off.                         by fiat.


   He complains of the "ludic fallacy":
   modeling the world as one of the
   standard "games of chance".

      Rolling a pair of dice
      builds up a bell
      curve centered around seven       You know in advance that
      with values lopped off at         no one will slip in a new
      snake eyes and boxcars.           invention like a die with
                                        more sides that counts
      Often, the real world             higher than six: that
      phenomena we care about --        would be cheating.
      let's say, the traffic on
      your website -- only looks              No such rules
      bounded over the short                  apply in the     (Unless the
      term.  At any moment                    real world       industry is
      strange leaps and crashes                                heavily
      may occur -- and lacking                                 unionized.)
      any good way of dealing
      with those possibilities,    But then, I would             (Or protected
      we tend to ignore them.      not say Taleb has             by patents.)
                                   any really good way
                                   of dealing with the
                                   possibility of        His financial advice
                                   extreme events.       is to go with a
                                                         "barbell" strategy:
                                                         combine the extremely
                                                         safe and the wildly
                                                         speculative.

He warns against trusting people
who act like they know everything
and claim they can manage risk
with whizzy mathematical models.

    Myself, I didn't particularly
    need to hear Taleb's expose
    of financial advisers...
                                         WIDE_FIELD
    People with hard-science
    backgrounds don't tend to            SAN_FRANCISCO_DREAMING
    have a huge amount of
    respect for these folks.


        Comments abound such as:

           "Economics?  But
           it doesn't work."

--------
[NEXT - WIDE_FIELD]