Wednesday, July 25, 2007

Book review

Review: Game Design: Principles, practice, and techniques--the ultimate guide for the aspiring game designer. Jim Thompson, Barnaby Berbank-Green, Nic Cusworth. Wiley, ISBN 0-471-96894-3. Full color, 192 pages including brief index and gloassary.

This is a very good book, particularly for teens. It is written primarily by a teacher, and makes strong use of color and illustrations. Each topic is covered in just two facing pages, usually. There is very little long text, again a plus where young people used to reading (skimming) the Web are concerned.

In the end, the book is not about game design generally, but about game design and production of video games that focus on a single character--FPS, action, and the like--the kind of game that particularly appeals to teenage boys.

This is the first book I've read that describes the process of modelling characters and then making them ready to be manipulated by programming.

There is almost no recognition--in common with most other books about digital games--that you can plan everything about a game down to a "T", but you won't really know whether you've got something good until you have a playable prototype. I've just been reading a history of the original Civilization game on Gamasutra that describes Sid Meir's process. He programmed, Bruce Shelley (who later made Age of Empires and earlier was the Avalon Hill "developer" for the American version of Britannia) played the game, they discussed what worked and what didn't, Sid modified, Bruce played, and so forth. The playable prototype was the key to success.

Perhaps genre games such as FPS are so similar to an archetype that you can plan it all beforehand and still get it (mostly) right. This "front-loaded" attitude primarily comes from the necessity for game studios to present detailed plans (the Game Design Document) to potential publishers. If the publisher likes the plan, they put up the money to enable the studio to produce the game. To put it another way, it's now too expensive to produce working prototypes of A-list games, so studios produce written plans. No wonder there's little risk or innovation in these games.

1 comment:

Ian Schreiber said...

I think the general rule is, number of iterations is proportional to design risk.

Large numbers of iterations (high risk, i.e. lots of innovation) means you have to prototype early and playtest often. This was certainly the case for Civ.

Highly derivative products (clone-and-tweak method) have significantly lower risk, which means you can do a lot more planning up front and still be confident that your game is fun.