I found it difficult to write up my thoughts on An Elegant Puzzle. On the one hand, the book is gorgeous: it’s a beautiful hardback volume, printed in colour on thick paper. On the other, it’s not a very enjoyable read: it’s dense with prescriptive advice, interspersed by mostly superfluous diagrams.
I guess one might think of it as a monograph on software engineering management. It feels like Larson has included everything from his experience that he could think of that might be useful. Hell, it even includes a page on media training.
Larson does do a difficult and rare thing: he tells you exactly what he recommends to do. “On-call teams should have eight engineers,” “these are the stages you should have in your hiring funnel.” Texts on this topic are usually punctuated with “your mileage may vary" or “it depends.” Personally, I really missed the war stories, the background of where this clearly hard-won experience came from.
The author clearly likes systems thinking. It is the book’s explicit lens into engineering management and, as such, the source of both its strengths and weaknesses. It’s the foundation for the advice on how to structure your teams and processes. But it’s also like trying to enjoy a game of baseball by reading its rules.
It’s like the title: clever but discordant with the subject of humans and managing groups of them. The pieces of this puzzle don’t fit.
Gergely Orosz got much more out the book. Graciously, he’s done the legwork of linking to Will Larson’s original blog posts that precede the book.
In the end, An Elegant Puzzle feels a bit thin. It covers too much without sufficient substance.
The illustrative system diagrams that pepper the book are hit and miss. A few of them successfully do the job of a thousand words, the rest add very little.
Postscript: I read Elegant Puzzle in early 2020. I’m curious if I’d feel differently after re-reading it now.