Writing a book that can teach others how to create art is a difficult undertaking. Creating art is part technique, part instinct. Teaching technique and how to apply it is straightforward: hold the brush so, apply and strike the chisel thusly, learn the rules of English grammar. Teaching instinct, or at least how to tap into and use what instinct you possess, is harder. We are left to fall back on vague rules and analyses of art that works, all the while hoping that the budding artist will, through time, exposure and practice, hone what instinct he or she has.
When the art in question is computer game design, well, the task becomes harder. The concept of a "game designer" is a relatively new one, and lacks the corpus of writings other forms of art have. Computer game design is sometimes viewed as an anti-intellectual pursuit, even by its practitioners. Games are supposed to be fun, and anyone who plays enough games knows what's fun, right? Besides, who wants to read some dry academic tome about game design when there are so many other things to be done? There's C++ or Maya or Photoshop or RenderWare to be learned, or meetings to attend, or a design document to finish. Despite these pressures, though, game designers would do well to think about their discipline, and seek out writings that explore the theory of what they do. If nothing else, doing so will actually save time by providing an opportunity to learn by a method other than trial and error.
Writings on computer game design are few and far between. A number of books will teach you how to program games, but few will tell you how to design that game you're planning on programming. Chris Crawford's The Art of Computer Game Design, published in 1984, was the first book to cover computer game design in great depth; to all intents and purposes it was the last for the next fifteen years. Other books and essays have discussed bits and pieces of game design, but none have the scope and range of Chris Crawford's work.
Game Design: Theory & Practice, by Richard Rouse III, follows in the footsteps of The Art of Computer Game Design, as the book itself acknowledges. Within its hefty 584 pages lie a wealth of information, from high-falutin' theory and interviews with other designers to nuts-and-bolts explanations of how to work with playtesters. The book's scope is large, and it suffers from the yin-yang problems of such ambitious endeavors: it is at once a little unfocused and not comprehensive enough. And as with any book purporting to teach how to create art, you can take issue with a number of its statements. But Game Design: Theory & Practice is a worthy addition to a slim canon, one which every beginning game designer (and many experienced ones) would do well to read.
The book is intended for anyone who is interested in game design. Rouse assumes some slight familiarity with game design, but not much beyond what most people will have learned simply by playing computer games. The book's glossary explains terms which a reader might not know, and clarifies Rouse's personal take on some definitions. While the book often refers to the "team" designing a game, its principles are applicable to small one- or two-person design teams as well.
At the beginning, Game Design focuses more on theory: what gamers want, coming up with an idea for a game, the elements of gameplay. As the book progresses, it addresses more practical concerns such as how to write a design document, how to design levels for a game which utilizes them, and how to deal with beta-testers. Sprinkled throughout are interviews with game designers and analyses of notable games. Included with the book is a CD containing the text of the book as a PDF file, as well as a number of programs which might be of interest to nascent designers, including the text adventure languages Hugo and TADS, Chris Crawford's Erasmatron, DarkBasic, a free sprite library, and the Visual SlickEdit source code editor.
Most of the book is dedicated to theory. Rouse covers many aspects of game design in a comprehensive manner, introducing topics and then discussing various ways in which games deal with these topics. Examples are plentiful, though they often reflect Rouse's early involvement with Macintosh games. While I'm a big fan of Marathon, which Rouse often cites, how many of his audience will be familiar with it?
I include in these theory sections the chapters on game analysis and the interviews with famous game designers. Rouse takes a close look at Centipede, Loom, Tetris, Myth, and The Sims, deconstructing them in an attempt to explain what makes them good games. On the whole he is successful, and his chosen games represent a varied, albeit limited, cross-section of groundbreaking games. The interviews, however, are where the book truly shines. They are in-depth, running thirty to forty pages long, and offer a fascinating glimpse of the thought processes of some amazing designers. I found Rouse's interview with Ed Logg, creator of Asteroids, Gauntlet, and Centipede, to be extremely interesting. Logg's take on games involves more commercial concerns than most designers are comfortable confessing to, a view undoubtedly shaped by his work in the coin-op industry.