Overall, the book seems more to describe than explain, more to report than intrepert. There arises no general, well defined thesis from its 500+ page volume. At best, this book can be said to raise a lot of issues which a designer ought to have in mind when desining a game.
However, the vast majority of the issues raised are either of secondary importance or generally irrelevant. It breaks down the process of game design into topics in a way which is neither natural nor logical, and proceeds to pursue a rather sizyphian discussion of each of these topics in turn. These are: What is Game Design?, Game Concepts, Game Settings and Worlds, Storytelling and Narrative, Character Development, Creating the User Experience, Gameplay, and The Internal Economy of games and Game Balancing.
This division makes very little sense. These topics are all so closely realted, some to the point of overlapping, that attempting to develop a theorem which deals with each of them separately would result in exactly the kind of negligable book we have before us.
Actually, it would be impossible for the authors to develop any meaningful discussion of their subject, because they fail to define a) what we are trying to create and b) how do we measure our success. Nor can such a definition be induced from this overflous and superficial book. Without this definition, there is nothing that binds the book's pieces together (and, actually, had the authors bothered to provide a rigorous definition, they would have relized that no reasonable definition could be found for the garbled mess they've created), and it remains a pile of expressions in the spirit of "some people did this in some games, and some people did that in some other games". In short, the book does an admirable job in showing how NOT to perform a critical analisys of a subject, not to mention attempt to construct a wholesome theory.
While the book can be interesting at times, mainly because it makes one think on how such a book SHOULD be written, it is chuck full of assertions obviously made on the basis of misunderstandings, like the authors' curious misuse of the term Suspension of Disbelief, or their suggestion of the Hero's Journey narrative template as an object of imitation rather than a tool for analisys.
The authors' goal with this book also seems qustionable. At one point, they assert that, even were it possible, we wouldn't like our player to be tormented by remorse after taking an immoral action in the game. Why? isn't moral education one of the most important and unique roles of art? If it were indeed possible, and I'm sure it is, it would've been a glorious achievement for this medium, one which would put all its previous achievements far behind.
Or are the authors only interested in computer games as a source of pure fun? If so, I suggest they invest their impressive talent and enthusiasm in cooking or adult toy design - a medium's greatness lies not in the fun it offers, and these repectable fields are all about fun.
An interesting book for raising a large scale discussion, but one which falls short of grasping the deeper principles of its subject, and is, therefore, unimportant.
Also, while Rollings' other book is most suited for people making strategy games, this book really is general enough to be a worthy read for anybody working on any kind of game.
I only gave it four stars because, for me, the last half of the book--summary chapters of different game genres--was mostly throwaway, rarely going into very much depth or telling me information I didn't know already.
For instance, the back cover of the book On Game Design posits: "How do you turn a great idea into a game design? What makes one design better than another? Why does a good design document matter, and how do you write one? This book answers these questions and stimulates your creativity?"
It is important to note that the book does not limit itself to console video games or computer games. The essence of the rules discussed in this book are those of creating any type of game. Right away that should tell you whether or not you're going to find the book useful. Are you looking for a book that tells you, in general and abstract terms, what concepts are involved with creating a game, or are you looking for a book that actually works examples of concepts?
While this book does a good job of providing many checklists for consideration, advice for certain conditions, and a dictionary of possible ways to view game design, the writers do not follow through. There are few solid examples of checklist scenarios or of worked-through versions of a game scenario which a game designer would find helpful. Without a practical means to an end, there is little purpose in reading these examples except for reassurance that you're facing the same problem that other people have faced. There are many psychology texts available for that situation already.
If you're used to reading programming books, like I am, you're probably aware that they follow a standard format: Propose a problem, choose a method of solution, work through several to many versions of the solution, solve the problem. With only a proposal, it is rather unhelpful to not see why one solution is better than another when it comes to game design. For that matter, as you might have guessed, the level of abstraction to design presented in this book leaves no space for any code examples.
While the advice given in certain situations might be helpful to someone who knows nothing about game design, it is highly likely that whoever reads this book will have little need of it since the advice is so much common sense that a gamer of several years would already be aware of much of this. It's like a senior in college having to take freshman seminar.
But, aside from this little discussion of fault, there is much to be savored in this book. Don't let this review scare you off! Get a copy of the book. Read it. Keep it as a reference for when you might need a more formalized way of presenting a problem you face in game design.
And as I'm sure you know, once you've found a way to state a problem, you're almost ready to find a way to solve it.