First published in 2003, Stephen Senn's 'Dicing with Death' received exceptional praise for being a unique addition to 'pop-science' literature. Books on statistics - popular books, that is - are somewhat of a rarity and preceding Senn's book, only a handful of titles managed to permeate outside of academic circles and begin to find their way to the bookshelves of the layman. However, to regard 'Dicing with Death' as purely a 'pop-science' book is a very naïve misclassification.
'Dicing with Death' is a chronicle of statistics through the ages; it is the tale of "Galton meets Pearson meets Fisher", how mathematics lured medical hopefuls away from their plans, how Cambridge and London acted as the bastions of statistics over the last 200 years. The stories of the greats such as the Bernoulli family, Bayes, Laplace, Poisson, Greenwood, Ross, Bradford Hill, Doll, Galton, Kermack, Pearson (both Karl and Egon), Fisher and Cox - to name a few - are retold and connected to paint a wonderful picture of how statistics, mathematics, epidemiology and medicine have interacted and developed over time.
Concurrently, it is also a detailed insight into medical statistics and clinical trials. With help from fundamental theorems, unexpected paradoxes and other insightful results from mathematics and probability, Senn describes the important constructs and tools used today to aid the medical sciences and drug development. Randomised controlled trials and their (mis)interpretation, life tables, epidemiological modelling of diseases, meta-analyses and even Markov chain theory are addressed with care and used effectively to explain how many aspects of public health work rely on statistical science. Perhaps one of the most impressive features of this text is Senn's ability to explain the philosophy behind the Bayesian school of thought with great skill and consideration, ensuring the reader is not left confused or lost with respect to the underlying theory or its scope for application.
For the icing on the cake, Senn's fluent writing style and acerbic wit keep the reader turning the pages. Forays into popular topics such as the MMR-autism controversy, frequent courtroom abuses of statistics by lawyers and journalists' misinterpretations of statistics (the 'post hoc passed hack fallacy', as Senn calls it) ensure that the fusion of history, mathematical theory and medical applications make for an entertaining and enlightening read.
In the very first chapter, Senn writes, "I want to explain how important statistics is". It is fair to say that 'Dicing with Death' achieves this and so much more. It tells us that statistics is exciting, full of vibrant history and one of the key components in modern-day science. It is a book we all should have read nine years ago.