Review: An Elegant Puzzle (Book)

In Philosophy, Reviews by J Michel MetzLeave a Comment


Rebecca Weekly, Chair of the Open Compute Project, started a book club for managers - particularly engineering types. At least, that's the first group of people who joined to read the book An Elegant Puzzle: Systems of Engineering Management (affiliate link).

When I saw that she had written a short write-up on her blog about the book, it reminded me that I had wanted to do the same. I agree with Rebecca's overall assessment, but I thought I might add in a few additional perspectives.

Review: 6.5/10 (Recommend)


Let me start with a confession: I hate self-help books.

Wait. No. I loathe them. The Men Are From Mars, Women Are From Venus-like mentality of watered-down pseudo-intellectual claptrap garbage are the snake-oil of the literary set.

Remember The Rules? No? Good. Ever read For Men Only? Dear God, don't. Run away. Avoid like the Black Plague. You'd be better off spending your time filing your toenails with a belt sander. And less painful, too.

Nevertheless, I went into An Elegant Puzzle with an open mind. I'm glad that I did, because otherwise I would have missed out on a mostly insightful and quite useful engineering management tome.

This isn't by accident. Will Larson's book is based upon his rather useful blog. If you're not familiar with Mr. Larson (I wasn't until I started reading the book), he is a software engineering executive with stints at companies like Stripe, Uber, and Digg.

This perspective makes it well suited for an engineer's mindset. Go figure - a management book written by, and for, engineers. Who would have thunk?


Book Strengths

Mr. Larson draws from not only his years of experience, but also his experience in developing methods for engineering efficiency to create (and diagram) practical and useful flows.

I found these flow charts and basic diagramming to be helpful, even if rudimentary (I'm a visual person by nature, and even basic graphics can properly orient me to complex and in-depth subjects). His direction is extremely approachable in the first half of the book, coupling his fondness for lists (more on this below) with appropriate anecdotes as examples.

"I believe that management, at it's core, is an ethical profession." - section 4.3.1

I'm not sure why Larson waited until Chapter 4 for such a revelation, because it underscores the philosophy behind the entire book. It seems to me that he would have done the book - and the audience - a better service if he had started this theme (and reinforced it more) at the start of the book.

Why? In my 30+ years of professional career, I've found that this is a minority attitude - so much so that ethical managers truly stand out (in a positive way). It was such a relief to see it explicitly stated, even if it was late in the book.

As it is, in retrospect it helps the reader understand much of the motivation of the earlier chapters; from team size to change management to vision and strategy, ethical and moral principles guide his process - and it shows.

By the way, the section on Vision and Strategy (Section 3.3) is a must-read reference for anyone responsible for putting together those concepts for any organization. It really is that good.

Unfortunately, Larson doesn't carry many of these strengths throughout the entire book, and winds up undermining himself at times.


Book Weaknesses

As I mentioned earlier, the book is derived from a series of Larson's blogs. The last two chapters make this painfully clear, making the read quite a slog.

Books chapters are not blog articles. Audiences consume them differently and have different expectations. As it happens, it appears even Larson was getting a bit exhausted with the writing process, as these chapters tend to devolve into mere list after list (and lists of lists).

While this is useful for reference material, it became an exhausting trek through what was little more than a means to check off items. No understanding or internalization required.

While that's annoying to be certain, perhaps the largest flaw in Larson's book is that every anecdote is a perfect scenario. Conflicts are handled reasonably, managers are wise, employees are rational.

This is, of course, a very software-engineer-oriented approach. Human foibles are errant bugs in the system that can be eradicated or commented out. As noted above, he makes the (astute) comment that management is an ethical profession, but never addresses what happens when a manager is faced with an unethical employee or upper management.

Therein lies the disappointment in the later chapters. By casting the "ethical management" as a throwaway line in Chapter 4, it risks undermining many of his finer points. It's such a simple truth that underscores Larson's entire philosophy to the point where it elevates the lists beyond the "this is what you should do" to "this is why you should do them."

By the time you get to Chapter 6 (the last chapter), for instance, that ethical motivation is completely absent. Larson may have well labeled the entire chapter as an object-oriented subroutine to #include in a management workflow program.

This is exacerbated by one pretty significant flaw in Larson's approach. Human beings are not programs.

Larson spends an extraordinary amount of time discussing creating roles and the interviewing process, but management lifecycle extends to the dissolution of problematic reports as well. This is particularly salient when you look at the minefield that is cross-cultural communication in engineering (particularly SillyConValley) roles.

This means that each one of his lists have a major random() element to that risks derailing anyone who attempts to follow these lists without a guiding principle (moral, ethical, or practical). As a result, the utility of such lists become more limited and restricted, and the unethical manager can easily use them as a means to avoid doing the Right Thing™.


Book Recommendation

All things considered, the book is well-worth a read.

(Note: I made the mistake of getting the accompanying audiobook so that I could continue to listen in the car. I do not recommend getting the audiobook, as the chosen narrator read the material as if she were a talking elevator announcing the next floor. Avoid at all costs.)

As I mentioned, there is a lot of gold in them thar hills. Larson has a gift of talking to engineers who love straightforward I/O methods of programmability, and his lists do just that.

Yes, humans aren't programs and yes, blindly following checklists will get you into trouble if you can't read the room. However, you've got to start somewhere. As a starting point, Larson's approach is likely eye-opening to many engineers who have been thrust into the uncomfortable world of human leadership, and should be commended for that.

Overall, I think the book is well worth a read, as well as a place on the reference shelf.

Rating: 6.5/10


Leave a Comment