Flocks, Herds, and Schools: A Distributed Behavioral Model

Authors: Craig W. Reynolds
Year: 1987
Source: SIGGRAPH Computer Graphics, 21(4)

Contributions

Introduced the “boids” model: three simple local rules (separation, alignment, cohesion) produce emergent flocking behavior indistinguishable from natural flocks
Demonstrated that complex collective behavior requires no central controller — purely local interactions suffice
Provided a computer graphics tool for generating realistic crowd and flock animations used in film and games
Established a foundational paradigm for swarm intelligence, multi-agent systems, and artificial life research

Agent-based simulation: each boid computes its next move based only on the positions and velocities of nearby neighbors
Three behavioral rules applied in priority order: (1) separation — avoid crowding neighbors; (2) alignment — steer toward average heading; (3) cohesion — steer toward average position
Simulated in 3D with obstacle avoidance and predator evasion extensions
Evaluated by visual comparison with filmed footage of real bird flocks and fish schools

Showed that three simple local rules reliably produce stable, realistic-looking flocking across a wide range of parameters
Demonstrated that the model scales from small groups to large flocks without modification
Established that separation (collision avoidance) is the highest-priority rule for stability
The model became standard in computer graphics for crowd simulation and is used in films such as Batman Returns (1992)

No learning mechanism: rules are hand-coded and fixed; the model does not adapt to novel environments
Simplified physics: ignores momentum, inertia, and energy constraints present in real animal locomotion
Real animals use more complex perceptual processing (vision, hearing, pressure) than the model’s idealized neighborhood radius
Does not explain how the rules themselves evolved or are instantiated in biological neural circuits