Our authorities are not authoritative. They often want the protection afforded by their expertise or authority to nudge others toward their own goals and desires. Stanford epidemiologist John Ioannidis contends that wrong numbers arise because scientists “may be prejudiced purely because of their belief in a scientific theory or commitment to their own findings . . . Or prejudice may prevail in a hot scientific field, further undermining the predictive value of its research findings. Highly prejudiced stakeholders may even create a barrier that aborts efforts at obtaining and disseminating opposing results.”1
This is not an investment book, although it touches on some important trading concepts like the Kelly Criterion, odds formulation, and the misuse of government macroeconomic data. The widespread misuse of statistics across science, however, makes a strong case that finance is not immune to wrong numbers. The misuse of p-values, failure to present confidence intervals, drawing conclusions from small samples, overfitting, and poor back-testing methodology are just a few of the problems faced in financial research, and we have not even touched on the issue of data being used to manipulate in the sales and investment persuasion process. All who read this book will say to themselves that some of the wrong numbers are obvious upon reflection, yet in a world of information overload, how can a reader question everything they read? How can a reader make sense of bad numbers?
Brown asks readers to question authority and think for themselves through critical analysis and basic statistical knowledge, as employed in solving Fermi problems. Use logic to ask whether the statistics being presented make sense when extended to a broader setting. Walk through assertions to their logical extremes and question the underlying data used for any analysis. You don’t have to be a statistical expert; just apply the basics, like confidence intervals, sample size, proper application of p-values, and power levels, to the numbers presented. Brown, through his interesting tales of numerical failure, walks readers through his approach to problem-solving and provides a path to better numerical thinking.
This is a powerful book for anyone who wants to be more numerate; however, I have some minor criticisms. While this is a book of short vignettes about wrong numbers across many fields, the stories are at times disjointed and could use stronger thematic introductions. Brown is occasionally too facile with the numbers, so the reader may need pen and paper, as well as a statistics book, to keep up with some of the key arguments. If the objective is to stop bad statistical thinking, working methodically through the correct way to conduct the analysis will better educate readers. Finally, Brown should spend more time writing about how to manage this information-challenged world. In a crowded news world, how do you make sense of the nonsense? Brown clearly notes red flags in his narratives, but getting at the right numbers is often exhausting work, and, while challenging, providing shortcuts for spotting disinformation is necessary for clearer thinking.
From this work, readers should follow the advice of Robert Solow, the Nobel laureate in economics, who was skeptical of stylized facts: “There is no doubt that they are stylized, though it is possible to question whether they are facts.” Always think of Solow sitting on your shoulder when looking at facts in an argument. The Royal Society, the United Kingdom’s national academy of science, has a useful motto, “Nullius in verba,” which is Latin for “take nobody’s word for it.” Wrong Number takes this to an extreme with a significant splash of cold water on all the readers willing to give those in authority the benefit of the doubt, as well as to professionals who play fast and loose with their numbers. You want to read this book as a warning for all that can go wrong with statistical misinformation.
Footnotes
1J. Ioannidis. “Why Most Published Research Findings Are False,” PLoS Med 2, no. 8 (2005): e124. https://doi.org/10.1371/journal.pmed.0020124.