There is no shortage of books about magic. Just search the keyword through your favorite online bookstore. Hundreds of guides, manuals, and how-to books, with or without pictures, large and small, are listed. Some are quite popular, but won’t go beyond a stepwise explanation of the mechanics of the magical acts they choose to represent. A few of them, such as Martin Gardner’s 1956 classic, Mathematics, Magic, and Mystery, are far more exciting, written with a wider perspective over the math behind the magic acts. This is the same legendary Martin Gardner whose 1956–1981 column Mathematical Games in Scientific American inspired a host of young and curious minds to pursue mathematics not as a hobby, but as a career. And this is the same Martin Gardner who, in his letter of recommendation for the young Persi Diaconis to Fred Mosteller, writes “of the ten best card tricks invented in the last 10 years, this kid invented two of them and maybe you should give him a break.”

In a sense, the recent book by Persi Diaconis and Ron Graham is the sum of all literature on mathematics of magic. While revealing the machinery of the magical acts, it covers a host of stimulating materials such as de Bruijn sequences, the Gilbreath principle, the Mandelbrot set, DNA sequencing, the Jordan Curve Theorem, and Markov chains, along with a chapter devoted to the lives of iconic figures in the world of magic, including the great Martin Gardner.

Also in this issue, as documented by Norma Hubele and Mark Arndt, there is so much at stake in the clashes ensuing from the statistical modeling of vehicular roof safety between governmental agencies, insurance companies, and carmakers that it could potentially inspire an edgy political film project by Robert Redford. Here, the actors include the National Highway Transportation Safety Administration, the Insurance Institute for Highway Safety, and private consulting firms.

Prakash Gorroochurn revisits the many insights of Gerolamo Cardano to some of the most fundamental components of probability theory, from the definition of probabilities as relative frequencies of favorable events in a sample space to the multiplication rule for independent events, the law of large numbers, and the Saint Petersburg paradox. Cardano often gets it right, but in some crucial junctures, he commits serious mistakes, partly because of his lack of familiarity with a then nonexistent syntactic language of probabilities. What is astonishing, though, is Cardano’s visionary insight in foreseeing probabilistic concepts that would be expanded by figures such as Bernoulli, Pascal, and Fermat at a much later period.

In their article, Mitchell Watnik and Richard Levine carefully build a counterargument to “we need a new stadium to be competitive,” a catchphrase in the world of team sports. Using a data set from Major League Baseball (MLB), spanning 1998–2010, the authors construct a measure of “competitiveness” comprised of the MLB teams’ performances, payrolls, and payroll rank. Watnik and Levine conclude that there is no evidence to believe new stadiums have been making teams statistically more competitive. Perhaps baseball franchises should look for stronger arguments before selling the idea of building a new stadium to the public.

I am writing this letter the day after Barack Obama was re-elected as the president of the United States. Election-related opinion polls were in full swing, and some correctly predicted the outcome. In his article titled “Prediction Markets: How Accurate Are They?” Jonathan Wilson gives an introduction to prediction markets for eliciting a probability of a presidential candidate being elected. According to Wilson, prediction markets can presumably benefit from taking into account the volatility of polls and the closeness of the poll to the actual event. To examine this, Wilson lays out an algorithmic approach for market-based predictions using 55 days of full trading data from Dow Jones. Devising a windowing technique to calibrate the opinions in favor of the binary outcome of Dow up or down at various time points prior to the closing bell, Wilson’s bottom line is the performance of market-predictions is hardly more impressive than the one obtained from Intrade, an opinion-based European platform.

Finally, in this issue of CHANCE, Julia Lane discusses the issues with privacy and confidentiality in the era of Big Data; Nicole Lazar covers the role of imaging modalities in better understanding fMRI data; the editors of Taking a Chance in the Classroom devise an impressive coursework for using the data around lead poisoning in children; Howard Wainer argues that a careful examination of a national-level 4th-grade mathematics test would demonstrate considerable growth across racial and economic strata in the last 20 years; in his ethics column, Andrew Gelman stresses the responsibility of statisticians to clearly state the assumptions of their models; and, in collaboration with Sophie Donnet, Christian Robert reviews six books, including a hugely popular graphic novel, Logicomix: An Epic Search for Truth.

Back to Top