An Interview with Bart Holland
Q: What kind of book is What Are The Chances?
A: It is a reader-friendly introduction to probability and its impact on everyday life. Even though the subject is mathematical, I get the concepts across through examples that are accessible to everyone. Anybody who has thrown dice, played a lottery, or been astonished by a coincidence will be able to understand what I am driving at in this book. There is also some fancy math in the book, in the form of occasional presentation of equations, for those people who want this type of exposition. But most of the book is anecdotes and discussion in words rather than in equations or numbers, and if you don't like the math part you can just skip a bit until the narrative resumes.
Q: Why do you think people should care about probability?
A: Probability is a fundamental part of our reasoning process, whether we realize it or not. In life we constantly make decisions in the face of uncertainty, based on the odds. When we take a prescribed medicine, buy life insurance, or invest in stocks, we are reasoning from information about probabilities when we try to figure out what the result will be. When you are told, "quit smoking, you'll be healthier," the issue is one of decreasing the probability of death from lung cancer or heart disease, for example. How much can you lessen your risks this way? Being able to think clearly about what probabilities mean and how they compare and relate to one another helps us make these decisions in a way that maximizes the chance of having the desired outcome. Also, we can protect ourselves from chicanery by careful reasoning about probabilities. For example, you've no doubt had people tell you that your personality is typical of people with a certain astrological sign. Is it? And, is the statistical association strong? Could they tell you your sign in advance, based on your personality, rather than merely agreeing with the sign once they've been told it? There have been scientific tests of this, which I describe. Interestingly, two thousand years ago the Roman senator Cicero had many of the same thoughts about astrology, and I discuss his ideas in this context.
Q: Are there "big issues" in our society, where probability would be useful in approaching our problems?
A: Megan's law is an example of society's answer to a social question – under what circumstances are people entitled to know about sexual offenders among their neighbors? – and it is largely based upon the probability of another offense being committed. This probability is calculated based upon a formula and there is a discussion of this in the book. The issue of global warming is also discussed. The question is, since world temperatures fluctuate all the time, how do we know when a series of warm years is more than expected by chance? The key to determining this is an understanding of the variability that is typically seen. The number of warm years must exceed the fluctuations that would be likely by chance alone.
Q: Why are "voodoo deaths" and "office gossip" in the subtitle?
A: Medical journal articles are cited in my book to show that there may be an elevated death rate among people subjected to "voodoo hexing." But this is a good example of statistical association not being equivalent to causality, and I explain why. As to office gossip, it's an example of exponential spread – until you run out of people interested in the tidbit. I use this as an easy-to-understand example of a certain type of mathematical model, one applicable to how epidemics can explode, and then die down when they run out of susceptibles. In fact, the book opens with a description of the spread of bubonic plague in medieval epidemics. The same probability model is applied to chain reactions among splitting atoms in atomic bombs. One of the themes in the book is that the same kinds of models apply to many types of phenomena throughout nature and society.
Q: Can you help me with something immediately practical and profitable? Specifically, can you help me get rich?
A: There is a section on fluctuations in stock prices. If you want a sure thing, well, you can be sure of winning a lottery if you buy up one ticket with every possible combination, and the book tells the story of a recent lottery in which a consortium of investors used that strategy and made a multimillion-dollar profit by buying millions of tickets.
Q: That's not practical for me! On a smaller scale, don't your students ever prevail upon you to make a class field trip to a casino?
A: Well, students do ask me to place bets for them. The odds of winning can indeed be improved if you have knowledge and skill in certain games where choice, not just chance, is involved – blackjack or poker for example. But the odds are still greater that you lose than that you win, so I don't place bets for other people. I tell them to read the book and then decide if and how to gamble on your own – hopefully in a way that utilizes thinking about probability!
Bart K. Holland is an associate professor of biostatistics and epidemiology in the Department of Preventive Medicine and Community Health at New Jersey Medical School.
© 2002 The Johns Hopkins University Press
|