Boids

I’ve seen some impressive Starling murmurations recently, and they have inspired me to implement the famous Boids simulation.

How It Works

Boids use three simple steering behaviors to create complex flocking patterns:

Alignment

Boids try to match the direction and speed of their neighbors. This creates coordinated movement within the flock. Try setting alignment to maximum to see boids moving in parallel paths.

Cohesion

This behavior makes boids steer toward the average position of their neighbors. It keeps the flock together. With high cohesion values, you’ll see tight, dense clusters forming.

Separation

Prevents boids from crowding too close to their neighbors. This avoids collisions and creates more natural spacing. Increasing separation will cause the flock to spread out more.

Technical Implementation

I used p5.js and I started off by following this excellent tutorial by The Coding Train.

I also added a mouse/tap interaction that the boids will try to steer away from, give it a try!

There’s probably a bunch of mistakes in the simulation, and I’m certainly missing some more advanced features.

I might come back and implement quadtree spatial partitioning for better performance. The current simulation is pretty heavy and can’t support more than a few hundred boids.

Checkout the source code.