‘Coping With Extreme Numbers’
I think I could read Brian Kernighan’s writing about anything. One of the reasons C did so well as a language was, I think, because of the straightforward way Kernighan & Ritchie’s book ‘The C Programming Language’ explained it all, and continued to serve as a useful reference once read. (When I started programming professionally, we always had a copy of K&R within easy reach.)
This book isn’t about programming though. It’s about trying to make sense of numbers.
Some of the fun bits of the book are headlines and stories from newspapers, where the numbers... just don’t add up. (Sometimes literally.) There’s a tendency to take numbers at face value without really thinking about whether the numbers could be wrong or what the implications would be if they were right. That’s the author’s hobby-horse here - that we should think about the numbers more.
My favourite part of the book covers Fermi estimation - making reasonable guesses at knowable parameters to come up with rough answers to tricky questions. For example, the classic Fermi estimation problem is (I think) how many piano tuners work in Chicago? On the face of it, most people wouldn’t know and would struggle to come up with any answer, but if you break it down into smaller questions that can reasonably be estimated (such as: How many people live in Chicago? What percentage of them own a piano? How long does it take to tune a piano?) you’ll generally get close enough to a decent answer.
That’s only one part of the book though - there are plenty of other fun chapters covering many of the fallacies we humans have when numbers are involved.