Predator Protocol
To Become the Ultimate Predator
Huang, Yuankai, Coll of Arts, Media & Design

Predator Protocol is a browser-based evolution game I built with p5.js. You play as a simple geometric predator moving through a living ecosystem. From the start, I wasn’t just trying to “make a game,” but to explore something I’ve been curious about: how simple rules can start to feel like real life when they interact. The main question for me was: how do you make a world feel alive instead of scripted? So instead of designing fixed behaviors, I let every creature follow its own simple logic. The small triangles use a Boids system, so they naturally move like a group without any leader. The rectangular hunters just patrol and react to things nearby, while the larger predators constantly chase the player and create pressure. I also added drifting obstacles to keep the space always changing, so it never feels stable. The growth system is roguelike, but I think of it more as evolution. Each level gives random upgrades like speed, size, or cooldown, and different rarity levels push you toward different playstyles. I didn’t want one “best build”—I wanted each run to feel like a different creature developing over time. One thing that surprised me was how simple systems can feel intelligent. Most of the AI is just basic vector math, but players still read it as intentional behavior. That gap between what’s actually happening and what it feels like became really interesting to me. I also focused a lot on visual feedback instead of UI, using color, flashes, and motion to communicate what’s going on. At this point, I see Predator Protocol less as a finished game and more as a small experiment in simulation and emergent behavior.