Now here's a subject you don't hear about very often:
risk in human life, how we measure it, how we control it,
and how we react to it. The first 100 pages of Against
the Gods are devoted to the first faint stirrings of
the mathematics of probability (take note that although
not much math is required to understand this book, it
definitely helps if you're not intimidated by it), after
which the pace picks up and it is mostly devoted to human
reaction to risk, rather than the mathematical
measurement of it. An interesting paradox in risk assessment comes
courtesy of Daniel Bernoulli, who posed the following
problem in 1738. Known as the Petersburg Paradox, it
proposes a game between Peter and Paul in which Peter
tosses a coin continuously until it comes up heads. If it
comes up heads on the first toss, Paul wins one ducat and
the game is over. If it takes two tosses, Paul wins two
ducats. If it takes three tosses, he wins four ducats,
and so forth. The question is, how much should Paul be
willing to pay to play this game with Peter? The paradox is this: the standard calculation of
expected value indicates that the value of playing the
game is infinite. The calculation looks like this: (1/2 * 1) + (1/4 * 2) + (1/8 * 4)
. = 1/2 + 1/2
+ 1/2
. = infinity Nevertheless, no ordinary person would be willing to
pay more than a few ducats to play this game. Bernoulli's
explanation is that our assessment of the value of
the winnings decreases as the probability of the winnings
gets smaller. Mathematically, the probability of the game
going to, say, twenty tosses, is 1/1,048,576, and the
payoff is 524,288 ducats. However, although the expected
value is still 1/2, the probability is so small that we
effectively evaluate it as zero. In our minds, the
expected value of the game is not the infinite series
above, but a series that stops after 10 or 20 terms. (As an experiment, I wrote a program to play this game
a few million times and tell me the average payout. It
came out to around 10 ducats. I imagine you have to play
an infinite number of games to get the infinite payoff,
which certainly justifies our instinctive reluctance to
value the game very highly if we only get to play once.) This eventually leads to a discussion of Prospect
Theory, which examines the way people rank risks. An
interesting result of Prospect Theory is that people
aren't generally risk averse, they are loss
averse. For example, give a group of people the
choice of $1000 or a 50% chance of winning $2000, and
most will choose the sure thing. However, give them a
choice of losing $1000 immediately vs. a 50% chance of
losing $2000, and most will take the risk. Interesting,
no? Apparently, this basic pattern of being unwilling to
accept a sure loss has been verified over and over in a
variety of circumstances (losing money, losing lives,
losing employment, etc.) and certainly says something
basic about human nature. Prospect Theory is still relatively young, but the
results so far (and Bernstein's summary of them) are both
interesting and thought provoking. Against the Gods
is highly recommended if you have an interest in this
kind of thing. Back to DrumNet Home Page |