At the beginning of the week I finished another player behavior. I was able to polish the looping behavior a bit by limiting the number of “past selves” to five.

Most of this week was spent doing research on artificial intelligence for games. I am reading AI for Games by Ian Millington. He goes over all sorts of algorithms that have to do with movement, collision detection, and decision-making. I am investing time in this research so that I can be sure that I choose the right system for my needs: maybe a simpler system would suffice for the behaviors I want.

Some new ideas arose that can be extensions of whatever AI system I choose. Because most of these systems depend on data about internal states of the character, I can choose that data to determine the moods of other characters. For example, maybe I keep a toll of how many objects the character has interacted with (this could also be focused on a specific interaction, like how many times did they write on the paper). This number could determine whether the convenience store worker is in a good mood or a bad mood when they reach them.

Another area of research that arose for me is the relationship between programming a character and writing a character. I find it fascinating that screenplay considerations like a character’s desires versus needs come into play when designing them as goal-oriented systems. Interactions between the main player’s desires/needs and the AI they will encounter can be thought of as the crux of any space where they meet, determining the potential outcomes of player-NPC interactions.