Behavioral Mathematics for Game AI

by Dave Mark

Human behavior is never an exact science, making the design and programming of artificial intelligence that seeks to replicate human behavior difficult. Usually, the answers cannot be found in sterile algorithms that are often the focus of artificial intelligence programming. However, by analyzing why people behave the way we do, we can break down the process into increasingly smaller components. We can model many of those individual components in the language of logic and mathematics and then reassemble them into larger, more involved decision-making processes. Drawing from classical game theory, "Behavioral Mathematics for Game AI" covers both the psychological foundations of human decisions and the mathematical modeling techniques that AI designers and programmers can use to replicate them. With examples from both real life and game situations, you'll explore topics such as utility, the fallacy of rational behavior, and the inconsistencies and contradictions that human behavior often exhibits. You'll examine various ways of using statistics, formulas, and algorithms to create believable simulations and to model these dynamic, realistic, and interesting behaviors in video games. Finally, you'll be introduced to a number of tools you can use in conjunction with standard AI algorithms to make it easier to utilize the mathematical models.

Genres: Programming, Mathematics, Artificial Intelligence, Education, Game Design, Games

459 pages, Paperback

First published January 1, 2009

Reviews

[thirtytwobirds · Rating 4 out of 5]

A good, basic, accessible introduction to the subject of game AI. Does a wonderful job of explaining the mathematical concepts without assuming too much preexisting knowledge.

It does move a bit slowly at times, but I suppose an introductory book *should* err on the side of being a bit too simple, instead of risking losing the reader.

[Ashkan Saeedi Mazdeh · Rating 3 out of 5]

Overall the approach is interesting and practical and good with the potential to open possibilities up however it did not discuss things like how to get info from the environment and other agents well and fast and how to make knowledge data bases efficiently and instead sometimes stuff were described too much. it feels like it could be 200 pages or at most 250 but is much bigger.
The philosophical parts sometimes iterated and iterated the same thing again and again. the technical part was a bit more dense and better on that regard.

[Emre Ergin · Rating 3 out of 5]

left at page 180, since it is basically a microeconomics text book, while pretending to be a book about AI.

[David Hunter · Rating 4 out of 5]

Great explanations and examples.