felt that the fewer people I involved at this stage, the less likely I was
to upset the delicate balance of what was going on. Something was taking its course, and I was now a willing part of the process.
I wasn’t sure how much baggage it could handle.
I called the agency; it sounded reassuring. I made an appointment. It looked reassuring. I suppose I’d half-expected men who
never took their hats off, lounged back with their feet on the desk, and tipped shots of whisky into their coffee from bottles
kept in a drawer. Instead I was greeted by a matronly receptionist with a pleasant smile who invited me to wait for a few
minutes in a comfortable room with fresh flowers and a stack of current magazines and newspapers.
The associate I saw looked like a junior partner in a midsize law firm. I explained I was tracing a family tree, but he didn’t
seem concerned about my motives. It was routine, he said. He asked one question: If these people were found, did I want them
approached? I replied no, just tell me where they were, and I’d take it from there. He asked me to give what information I
could to his assistant in another office, a Miss Shelley. Nadia Shelley, I remember. I saw the name on her desk and thought
how striking it was. Striking girl too, as I recall, but businesslike and efficient. She assured me that everything would
be passed on to their associates in London and the search would begin at once. A routine search of this kind, she said, was
really very simple. I paid a surprisingly modest retainer, which she told me they would not exceed without my written agreement.
I should expect to hear from them within a couple of weeks.
I spent much of the next few days thinking about the law of large numbers. This is the paradox at the heart of probability
theory as well as the foundation of all statistics. It is also, of course, central to the notion of coincidence. Yet, like
so many of these things, the closer you examine it the harder it is to see and the more difficult to grasp.
The paradox is this. If I toss a coin in the air, there is a 50 percent chance it will come down heads and 50 percent it will
come down tails. No matter how many times I toss it, the odds are fifty-fifty each time. But if I toss it a thousand times,
it will come down more or less exactly five hundred times heads and five hundred times tails. Why this should be so is a mystery.
To many people, I confess, it doesn’t seem like a mystery. It seems somehow—if obscurely—obvious. So try this.
Radioactive substances decay at an absolutely fixed rate, which is known as their “half-life.” This is so precise that archeologists
routinely measure the age of fossils using radiocarbon tests. Yet the decay of each individual atom composing that radioactive
substance is totally unpredictable and spontaneous. So what is the mechanism that causes large numbers of these unpredictable
and spontaneous events to average out with a smoothness that allows the overall decay of the substance they constitute to
be used as the most accurate historical clock we have yet discovered? The fact is that nobody knows.
The mathematician Warren Weaver once came up with another famous demonstration of the improbability of probability. He noted
from the New York Department of Health records that between 1955 and 1959 the average number of people reportedly bitten by
dogs each day in the city was 75.3 (1955), 73.6 (1956), 74.5 (1957), 74.5 (1958), and 72.4 (1959).
Even assuming (and it is only an assumption) that the human and dog populations of the city remained relatively stable throughout
that four-year period, how did each dog know when it was his turn, or that when he’d had one bite, or maybe two, he was not
to have a second or third?
There are many such intriguing examples. Most interesting to me was the idea that the law of large numbers seemed to be the
only thing running seamlessly from the microscopic
to upset the delicate balance of what was going on. Something was taking its course, and I was now a willing part of the process.
I wasn’t sure how much baggage it could handle.
I called the agency; it sounded reassuring. I made an appointment. It looked reassuring. I suppose I’d half-expected men who
never took their hats off, lounged back with their feet on the desk, and tipped shots of whisky into their coffee from bottles
kept in a drawer. Instead I was greeted by a matronly receptionist with a pleasant smile who invited me to wait for a few
minutes in a comfortable room with fresh flowers and a stack of current magazines and newspapers.
The associate I saw looked like a junior partner in a midsize law firm. I explained I was tracing a family tree, but he didn’t
seem concerned about my motives. It was routine, he said. He asked one question: If these people were found, did I want them
approached? I replied no, just tell me where they were, and I’d take it from there. He asked me to give what information I
could to his assistant in another office, a Miss Shelley. Nadia Shelley, I remember. I saw the name on her desk and thought
how striking it was. Striking girl too, as I recall, but businesslike and efficient. She assured me that everything would
be passed on to their associates in London and the search would begin at once. A routine search of this kind, she said, was
really very simple. I paid a surprisingly modest retainer, which she told me they would not exceed without my written agreement.
I should expect to hear from them within a couple of weeks.
I spent much of the next few days thinking about the law of large numbers. This is the paradox at the heart of probability
theory as well as the foundation of all statistics. It is also, of course, central to the notion of coincidence. Yet, like
so many of these things, the closer you examine it the harder it is to see and the more difficult to grasp.
The paradox is this. If I toss a coin in the air, there is a 50 percent chance it will come down heads and 50 percent it will
come down tails. No matter how many times I toss it, the odds are fifty-fifty each time. But if I toss it a thousand times,
it will come down more or less exactly five hundred times heads and five hundred times tails. Why this should be so is a mystery.
To many people, I confess, it doesn’t seem like a mystery. It seems somehow—if obscurely—obvious. So try this.
Radioactive substances decay at an absolutely fixed rate, which is known as their “half-life.” This is so precise that archeologists
routinely measure the age of fossils using radiocarbon tests. Yet the decay of each individual atom composing that radioactive
substance is totally unpredictable and spontaneous. So what is the mechanism that causes large numbers of these unpredictable
and spontaneous events to average out with a smoothness that allows the overall decay of the substance they constitute to
be used as the most accurate historical clock we have yet discovered? The fact is that nobody knows.
The mathematician Warren Weaver once came up with another famous demonstration of the improbability of probability. He noted
from the New York Department of Health records that between 1955 and 1959 the average number of people reportedly bitten by
dogs each day in the city was 75.3 (1955), 73.6 (1956), 74.5 (1957), 74.5 (1958), and 72.4 (1959).
Even assuming (and it is only an assumption) that the human and dog populations of the city remained relatively stable throughout
that four-year period, how did each dog know when it was his turn, or that when he’d had one bite, or maybe two, he was not
to have a second or third?
There are many such intriguing examples. Most interesting to me was the idea that the law of large numbers seemed to be the
only thing running seamlessly from the microscopic
Similar Books
