with Uncertainty
ability management section, the author offers a “solution” for
the flaw of averages.
Explaining the Problem
Throughout the book, Savage presents a multitude of examples, anecdotes, illustrations and explanations of who, how,
why, when and where using the average can be disastrous as
an estimation tool. These examples range from the relatively
mundane, such as why people are frequently late, to the potentially monumental impact of misestimating the consequences
of global warming or the likelihood of a terrorist attack.
Misestimations are often humorously depicted by a Jeff
Danziger cartoon or illustrated with an appropriate graph or
chart. All of these examples show how using the average can
often lead to poor decisions, which drives home the idea that
thinking in terms of “distributions” rather than averages is a
much more effective thought process for risk management or
any task where estimating uncertainty is necessary.
Savage provides numerous tie-ins to his Web site
( www.flawofaverages.com), where simple animations or simulations help crystallize the point he is making, and also uses
examples of random generators (such as a coin toss, dice or
spinners) to illustrate probability simply to readers who do not
have a background in statistics.
The Flaw of Averages not only exposes the problems in dealing with uncertainty but also provides an introduction to a potential solution in the field of probability management. Today,
Savage explains, rather than using an average of an uncertain
number as an input to a model, a complete distribution can be
used to provide a distribution of potential outcomes.
If you are looking for a book with a myriad of elegant formulae that get you to “the number,” this isn’t for you. The Flaw
of Averages illuminates the thought process for dealing with uncertainty without bogging down in technical jargon or complex
mathematical equations. In fact, the author uses a system for
dealing with technical jargon, noting them as “red words” in
special font, while providing a “green words” equivalent, summarized with a handy little table at the end of each chapter. For
example, for the red words “random variable,” he provides the
green words “uncertain number.”
The book is written in a very accessible, self-deprecating style
with a healthy dose of humor and a collection of anecdotes
about financial academic titans. As the son of Leonard Jimmie
Savage, a prominent mathematical statistician, Sam Savage
had the good fortune of crossing paths and collaborating with
such notable financial theoreticians as Milton Friedman, Harry
Markowitz and William Sharpe. I highly recommend this book
as an entertaining and enlightening read for anyone who deals
with estimating uncertainty.
Cory Robinson (CFA, FRM) is a vice president, portfolio manager and trader
with Oklahoma City, OK-based Tom Johnson Investment Management. He can be
reached at crobinson@tjim.com.
