SYSTEM: The Commons
LIVEWhy I Built This
I wanted to see, not read, how the tragedy of the commons breaks down and when it doesn't. I started with simple grazing, and at every step I caught myself fixing the symptom instead of the cause: the grass collapsed to nothing, those who had agreed to it scattered, abundance always died out. Each patch exposed another hole. At some point, I realized I wasn't building a game about cooperation, but a rig for testing my own hypotheses—so I added a run of hundreds of games with random parameters. So I could turn "I think monitoring has no effect" into a graph that either confirms or rejects.
Background
The basis is the classic Hardin dilemma: a common pasture, agents graze, each personally benefits from overeating, but a scorched field feeds everyone worse. It's clear and dead in the text. I wanted to watch. Then the model accumulated layers, each growing out of the specific breakdown of the previous one. Agents learned to negotiate quotas, but a contract without punishment deteriorates, forcing them to add monitoring and a fine at someone else's expense (the one who fines also pays—a second-order dilemma). Mountains cut the field into basins so that different communities would arrive at different fates on the same screen. The pact was tied to territory, otherwise those who had agreed would immediately scatter for grass. Saturation made possible pockets of abundance among living agents—the well-fed don't eat, the grass grows back, then everything is consumed: an outbreak and avalanche. The key question is not "will cooperation prevail" (usually not), but "under what conditions does tragedy cease to be inevitable?" Therefore, the main tool here is not the simulation itself, but running hundreds of games: you tweak the parameters randomly, looking at the point cloud to see what actually decides the life of the community, and what seems to be happening.