The Physics of Wall Street: a brief History of Predicting the Unpredictable

The Prediction Company 
•
139
lem, contradicted every classical expectation regarding how things like 
weather worked. (Lorenz quickly showed that much simpler systems, 
such as pendulums and water wheels, things that you could build in 
your basement, also exhibited a sensitivity to initial conditions.)
the basic idea of chaos is summed up by another accidental contri-
bution of Lorenz’s: the so-called butterfly effect, which takes its name 
from a paper that Lorenz gave at the 1972 meeting of the American As-
sociation for the Advancement of Science called “Predictability: does 
the flap of a Butterfly’s Wings in Brazil Set off a tornado in texas?” 
(Lorenz never took credit for the title. He claimed one of the confer-
ence organizers came up with it when Lorenz forgot to submit one.)
Lorenz never answered the question asked in the title of his talk, 
but the implication was clear: a small change in initial conditions can 
have a huge impact on events down the road. But the real moral is 
that, even though chaotic systems are deterministic — in the sense that 
an infinitely precise description at any given instant can in principle 
lead to an accurate prediction — it is simply impossible to capture the 
state of the world with such precision. You can never account for all 
the flaps of all the butterflies across the globe. And even the tiniest er-
rors will quickly explode into enormous differences. the result is that, 
even though weather is deterministic, it seems random because we can 
never know enough about butterflies.
farmer finished his physics degree at Stanford in 1973, although not 
without a few bumps along the way (after his first year there, he had 
done poorly enough to be put on academic probation — after which he 
entertained the possibility of dropping out to open a smoothie shack 
in San francisco or maybe smuggle motorcycles). By the end of his 
college years, however, farmer had pulled himself together sufficiently 
to be admitted to a handful of graduate schools for astrophysics. A 
trip down the california coast was enough to make up farmer’s mind, 
however, and he decided to attend the new University of california 
campus in Santa cruz. Packard, meanwhile, had gone to reed college, 
in Portland, oregon, a school famous for the independent spirit of its 
undergraduates.

---

 during the summer of 1975, the year after Packard’s junior year 
at reed’s and farmer’s second year of graduate school, Packard and 
farmer decided to try their hands at gambling. they had explored 
the idea independently, farmer through reading A. H. Morehead’s 
Complete Guide to Winning Poker, and Packard by reading ed thorp’s 
Beat the Dealer. With their analytic minds and disdain for authority, 
gambling systems had a certain appeal to both men. they could make 
money without doing work — and at least in the blackjack case, they 
could do it by being smarter than everyone else. It was a romantic idea. 
the trouble was in the execution.
Packard studied thorp’s system carefully and then, along with a 
friend from reed named Jack Biles, he took it to vegas. they kept 
careful track of their winnings and losses — and observed an awful lot 
of wins. day after day, they would record profits. they would switch 
to higher-stakes tables as their accumulated capital increased, and 
the profits would soar even higher. But then something happened. 
no matter how much success they had, there would always be a los-
ing streak that would bring them back to zero. In the end, they barely 
broke even. It was only at the very end of a summer of gambling that 
they realized they were being cheated. In the years since thorp’s card-
counting system had first been introduced, casinos had become very 
good at identifying — and foiling — card counters, often by simple 
methods like crooked dealing.
farmer, meanwhile, had memorized Morehead’s book. But he had 
never played poker before reading it, so even though he knew what 
to do in any given situation, he didn’t know how to shuffle cards or 
handle chips. He dealt like a kindergartener. But the poor mechan-
ics ultimately worked to his advantage: he looked like an easy mark. 
Playing in the card rooms of Missoula, Montana, under the alias “new 
Mexico clem,” farmer and a friend from Idaho — an accomplice from 
the motorcycle-smuggling scheme named dan Browne — cleaned up 
against the Missoula cowboys. Browne, a more seasoned player who 
had paid his way through college by gambling in Spokane, Washing-
ton, marveled at farmer’s unlikely success.
At the end of the summer, farmer and Packard decided to meet 
up to compare notes on their gambling adventures. farmer had good 
140 
•
t h e p h y s i c s o f wa l l s t r e e t

---