1432961734 Programming Game AI By Example (Wordware Game Developers Library) - Artificial Intelligence - Books - Books - GameDev.net
• Create Account

### Share:

This book is a great foundation to learn AI programming techniques and concepts from. As a COMPLETE AI novice when I picked up the book I can say after finishing it that I have a pretty good handle on all presented subject matter from pathfinding, fuzzy logic, STL programming, and much much more. Often several variations of a given technique are discussed and critiqued, giving you not just the "how to do x a certain way" but "why do x a certain way". This will allow you to pick the best application for your individual project.

Mat presents the material in an easy going but focused manner. I highly recommend the book!!!
Really amazing book. I have not read the whole book but I have started it and just finished FSM. I have also skimmed other parts of the book. Let me just tell you if you are an absolute beginner to AI programming (like I am) this book is a must have. The book is broken up into seven main sections:

0. Physics/Math Primer
1. WestWorld (finite state machine and all the goodies it entails [messaging])
2. Autonomous Moving Game Agents (Steering, Flocking)
3. "Simple Soccer" (soccer simulation)
4. Pathfinding (A*, Breadth First, etc)
5. Scripting
6. Raven (2d Shooter game. Much like Quake in terms of gameplay, pretty awesome)

It also has appendixes for:
C++ Templates
UML Class Diagrams

Overall what I've gone over so far (some steering and FSM) has been explained in a very good manner. Matt is not hard to understand and tries to relate ideas to gaming.
Amazing Book!!! I haven't read all of it; I've read the 2 first chapters(Math and Physics, State-Driven Agents), and now I'm reading steering behaviors! Mat Buckland is explaining the A.I. stuff so clearly and the book is organized so well that there's no way you won't understand something! This is, in my opinion, a MUST BUY/READ BOOK for A.I. beginners(note: not generally programming beginners but A.I. beginners).

Note: I also used some of the stuff I "learned" in the Finite State Machine chapter.
Normally I'm bothered when people write 'reviews' of a book after readying less than a quarter of it... but I've read this book and worked through it a number of times over the passed few months and the other reviewers here are right. The book IS fantastic.

It's not perfect though:

The reasoning behind the use/implementation of fuzzy logic in the Raven game seems a bit forced (a better example would have made more sense). This isn't a big deal though, because the chapter is still well written and gets the point across.

The only real complaint I have is the map editor used for the game created in the book (Raven). For a book that is SO superb at showing real world examples of everything talked about, very little is shown or said about the editor. The information that is presented wrt the editor is limited to notes and blurbs intermixed with the rest of the content - and no source is provided for it; only the exe.

Keep in mind that this is a SMALL complaint, as it's not incredibly difficult to figure out making your own... but it adds a time barrier between reading the material and trying it out on your own - especially for me as I was re-writing the book from the C++ presented to C# for XNA.

It's very fresh especially if you're coming from books about traditional AI - which are usually horribly boring.
Simply excellent. It covers everything that a beginner AI programmer is curious about. The source code in the book is just enough to clarify the details, while the entire source for all the simluations/games is available online. Not only is it a fun read, but it's easy to quickly find a what you're looking for when you need that code snippet or detail for your project.

If that's not enough, I was really impressed with the math and physics primer in the beginning. The author makes all the necessary points clear while providing game related examples. If you're unfamiliar with the required math and physics, the examples are just enough to get you on your way without bogging you down with important yet unnecessary (for this level) details.

All in all, highly recommended for the beginning AI programmer.

PARTNERS