Taleb's argument is that the events which actually shape a life or a market are the ones nobody's model contained, that our instruments are built to measure the ordinary and go blind at exactly the moment things start to matter, and that the honest response is to quit forecasting and arrange your affairs so being wrong does not finish you. His mascot is a turkey. Fed on schedule for a thousand days, the bird develops a well-supported statistical theory about the benevolence of the farm, and its confidence peaks on the afternoon before Thanksgiving. The turkey was not stupid. Its data simply held no information about the only event that ever mattered to it.
To get there Taleb splits the world in two. In Mediocristan, no single observation moves the average: put the heaviest person alive in a stadium of a thousand people and mean body weight barely twitches. In Extremistan, one observation swallows everything else: put Bill Gates in that same stadium and average net worth becomes a number describing nobody in the room. Wealth, book sales, war casualties, city sizes, and market losses all live in Extremistan, and the bell curve, a fine description of the first world, gets imported wholesale into the second by people who should know better. He calls the Gaussian the great intellectual fraud and never once softens it.
The first half is where the book does its best work. The chapter on silent evidence is the sharpest thing in it. Taleb picks up Cicero's story of the temple paintings showing shipwreck survivors who prayed and lived, asks where the paintings of the drowned worshippers are, and then walks that question through mutual fund track records, celebrity biographies, and the graveyard of manuscripts nobody published. The chapter on the ludic fallacy is the funniest. Hired to study risk at a Las Vegas casino, Taleb found that its four largest losses had nothing to do with gambling: a tiger mauling a performer, a disgruntled contractor with dynamite, an employee who spent years not filing tax paperwork, and a kidnapping inside the owner's family. Every one of them sat outside the risk model, and the risk model was excellent.
Then the book gets harder to defend. Part three turns to Mandelbrot and fractal scaling, and the prose thickens without getting more precise. Readers who bounce off this book usually bounce there, or in the prologue, which is dense in a way the rest is not. Taleb also spends a great deal of ink settling scores with economists, Nobel committees, and academics in general, and the score-settling wears out long before it stops. His self-regard is the most commonly reported obstacle, and it is real; whether it spoils the book depends on how much you enjoy watching a former derivatives trader tell a lecture hall it has been wrong for a century. This edition runs 672 pages, and perhaps a hundred of them restate a point already made.
The practical residue is smaller than the ambition, and Taleb says so with some pride, since a theory of the unpredictable cannot honestly supply a forecasting method. What it supplies instead is a posture. Never stand where one surprise can end you. The barbell is the concrete version: roughly 85 to 90 percent of your exposure parked in things too boring to blow up, the remainder in small speculative bets with a capped downside and no ceiling. Collect options. Be roughly positioned for many futures rather than precisely positioned for one. The essay added to this edition, On Robustness and Fragility, is the most useful stretch in the book and reads now as the sketch that grew into Antifragile.
Why you should read
- Readers of Thinking, Fast and Slow and Freakonomics
- Anyone who models risk or forecasts for a living
- Readers who enjoy an argumentative, opinionated essayist
- Fans of big-idea nonfiction with a philosophical streak
What to expect
- 672 pages, digressive, with a dense prologue
- A combative voice that picks fights with economists
- Strongest in the first half, thicker in part three
- An added essay on robustness in this edition
There is a decent irony in how it landed. The Black Swan came out in April 2007, a six-hundred-page argument by a trader with a taste for Montaigne and a grudge against statisticians. Eighteen months later the banking system did the thing the book said such systems do, and the title stopped being a metaphor and turned into a phrase people use in meetings without knowing where they got it. By Taleb's own definition, the career of this book was a black swan: nobody saw it coming, the effect was enormous, and in hindsight it looks inevitable.
