One interesting feature of this book on game programming is how graphics does NOT dominate the discussion. Most books on the subject are chock-a-block with every graphics tip you could imagine. But Dickheiser does not write for the novice. He has gathered a set of papers (chapters) that try to delineate the research boundary. It is a measure of the rising maturity of the field, and the complexity of many games, that graphics is relatively deprecated.



Though I hasten to add that there are still the pretty colour plates of renderings, and several chapters on these methods.



Why is the book symptomatic of the changing trends? In part because graphical methods are starting to saturate. The techniques are now good enough for near-photorealistic impressions. If you look carefully at the non-graphics chapters, you see that they tackle non-trivial problems arising from evermore intricate games. One instance is the use of AI-style decision making methods for how a game proceeds. When well coded, it can give behaviour of characters that is more intelligent, hence yielding a more interesting game. Another chapter looks at the learning process in AI, where a player might change strategies. Here, the research use of Support Vector Machines is suggested as apropos for commercial game development.



Professionally, you should pay attention to the book for the trend it exemplifies. You probably got into game programming because you were attracted by the graphics. But as graphics methods peak, the value added aspect of games is starting to shift towards the story line and how elaborate is the game logic.