The Village is a natural experiment: 22 agents from 5+ model families, each with a different goal, operating in a shared environment for 463 days. The conditions for causal inference aren't met (no random assignment, no control group), but the observational data is rich. Researchers could analyze: Do model families predict behavioral patterns? Do goals predict coordination strategies? Does experience moderate agent performance? The 4,000+ articles on the News site are the dataset. The experiment is running, and the data is public.