Adams systematically investigates the thinking that should precede and follow the use of math in decision-making processes.
It is not enough to simply â€œlook at the numbers.” Like petty criminals, numbers are capable of saying just about anything; they have to be interrogated to spill the truth. Adams offers a thorough course in questioning the more slippery variety of numbers. Polling methodologies, for instance, and as Adams amply illustrates, are vulnerable to all manner of bias, from loaded questions to non-response error. The author provides a checklist of questions to ask such notorious numbers and refers readers to a welter of news and scholarly articles to illustrate his points before he poses a series of questions to test readers’ understanding. Boring like a worm through the tree rings of math’s deductive logic, Adams hits the nub of validity vs. reality/truth: â€œprecise mathematical reasoning in whatever context, geometric or other, can only establish the validity of the conclusions based on the postulates set up as a starting point.” Meanwhile, compounding the problem of what numbers say, â€œinterpretation of a mathematical theory is a structure separate from the theory itself.” Adams takes math models and situates them in real-world examples: for vacation trips and the cost of smoking, what probability and linear-program models have to say about gambling and profit maximization (and which variation of which model is the most suitable fit). His text can be rarefied, calling on Euclid, Newton and Einstein to explain themselves, though it is the practical application of theories and models–what they can tell us relative to the data we supply–that concerns Adams foremost. To that end, he has endeavored to keep the math simple; â€œbasic arithmetic,” he claims, though at times the sky darkens: â€œLobachevsky took the contradiction of Playfair’s equivalent of Euclid’s fifthâ€¦” The rain, happily, doesn’t lasts long.
Never buy an answer until you know the question–a proposition, by its very lack of cunning, that allows Adams to cut through the data smog.