Game Development Reference
In-Depth Information
chapter 3
Finite State Machines
Finite state machines are our first formal structure for game AI. The regularity of
the structure helps us to manage program complexity, a major problem of hard-
coded scripts. Using FSMs, we break down the AI into a fixed—hopefully
small—number of states and the transitions between those states. If you have
studied formal FSMs in classes, our usage may differ. ''It's important to
understand that when game developers speak of state machines or finite state
machines , they are only loosely referring to the traditional Computer Science
definition.'' [Rabin02]
This chapter looks at what FSMs are and considers their design and analysis. The
analysis includes single-transition review and multiple-transition review.
(Multiple-transition review unearths the problems of ambiguous transitions and
race conditions.) We will touch briefly on complexity and examine three failure
modes. At the end of the chapter are projects to implement an FSM for a game-
related AI.
What Are FSMs?
Finite state machines have a finite number of states and a set of transitions
between those states. So what do we mean by states? The easiest way to think of
states is to think of different ways to finish the statement that begins, ''I am....''
We could create a list of verbs: flying, walking, running, or sitting. Another way
to model uses adjectives: tired, happy, or angry. The set of states might represent
Search Nedrilad ::

Custom Search