This book is the ur-nerd tome. There is no pretending: you either are the sort of person who is mathy enough, physics and astronomy-obsessed enough and all around nerdy enough to find this fun...or you aren't.

To give an example, most days, because I'm busy being a doctor, I spend a lot of time pretending that I'm not a nerd. But recently, I joined a lab where my boss is about as nerdy as I am. So comparing weekend notes, he says: I spent the weekend solving a Rubik's cross. And I said: I spent the weekend reading the new What If XKCD book. I won that competition.

To be honest, I'm not particularly motivated to write much of a review: if you're that nerdy of a person, you've read the webpage version of what-if xkcd and understand the joy that is Fermi Problems (and probably the annoyance that happens after you do a Fermi problem and you spend the rest of the day unable to stop doing Fermi problems), absurd questions about nuclear physics, random statistics and clever stick-figure illustrations.

The key points are these: I religiously read What If XKCD every week, and have read every single one published on the web. The book still had plenty of new things that I had never seen before. There are some extras in the book: one line answers to particularly weird questions. I was anticipating a major drawback of the book to be the loss of hover text and footnotes that appear in the online version; this is replaced by captions and the old-school form of footnotes (i.e. footnotes). However, this is not a great book to read far apart from the internet: it's impossible to get through the whole thing without having strong compulsions to google side questions.

P.S. The worst part of this book is in the acknowledgements when he says he already has an expert on genetics. Note to self: scheme to take out previous genetics expert and become Randall Munroe's personal brilliant geneticist...