27 Nov 2011
Conflict in Agent Based Models
According to Robert McKee and his book Story: Substance, Structure, Style and The Principles of Screenwriting, a conflict is the essence of every story: nothing moves forward in a story except through conflict. A conflict propels the story forward, it causes the protagonists to do something, and it is the driving force behind every change.
Is this true for Agent-based models (ABMs) as well? Is every successful ABM based on some kind of conflict? Let us consider three basic models:
- Thomas Schelling’s Segregation Model for ghetto formation:
neighbors have not the right color
- Craig Reynold’s Boids Model for swarm formation:
neighbors have not the right place or position
- Robert Axelrod’s Dissemination Model Model for culture formation:
neighbors have not the right traits
The driving force behind the change is in each case a small conflict. The agents act because their neighbors have not the right color, place, position, trait or attitude. Due to many cumulative interactions, already a small preference can lead to a large effect.
Thomas Schelling showed that a small preference for one’s neighbors to be of the same color could lead to total segregation. Many small conflicts about the right color in the neighborhood leads to large clusters of similar colors in form of ghettos.
Craig Reynolds showed that a small preference for one’s neighbors to be at the same place could lead to fascinating swarms. Many small conflicts about the right position in the neighborhood lead to large clusters of similar positions in form of swarms.
Robert Axelrod showed that a small preference for one’s neighbors traits could lead to separated cultures. Many small conflicts about the right trait in the neighborhood lead to large clusters of similar traits in form of cultures.