In this installment of the Inspector Morimoto series of novels, a man the detectives believe to be innocent seems likely to be convicted of robbing ATMs. A key component of the evidence against him is the testimony of an expert witness on the stunningly low probability that he would have been in the vicinity of each of the crimes by chance.
This provides a perfect opportunity for Officer Suzuki,
the inspector's assistant, to make use of her mathematical training.
The following quote from Inspector Morimoto and the Two Umbrellas, the first book in the series, explains the connection Suzuki sees between mathematics and her work for the police department:
(quoted from Inspector Morimoto and the Sushi Chef: A Detective Story set in Japan)
It was his second trip to Kurashiki that day, and this time he took Suzuki with him. Suzuki had been working with Morimoto since April of the previous year when she had graduates with honors from the prestigious Tokyo University with a degree in mathematics. Not many students join the police force after finishing a mathematics degree, but the disciplines of mathematics and crime detection do in fact have many very basic similarities. The ability to exploit patterns and to identify inconsistencies is a fundamental necessity for any good mathematician, and also for any good police detective, as Suzuki often tried to explain to her friends.

Although her reasoning skills are tested in each of the novels, it is here that it is most explicitly mathematical. Fortunately,
"Timothy Hemion," author of the Inspector Morimoto mystery novels, is the pseudonym of statistics professor Tony Hayter, who proved the TukeyKramer conjecture. So, the mathematics is handled well. (See here for an article about the author on his institution's website and here for an interview with him.)
In the end, although the testimony of the expert witness is shown to be correct from a purely computational point of view, it is recognized that like most results in probability and statistics it should not be accepted naively without considering the validity of the assumptions and the direction of causality.
Thanks to Dan Flath for bringing this series of books to my attention.
