Run the same city across three policy presets and multiple deterministic seeds. Compare outcomes, test one policy change, and inspect repeatability and consistency checks. These are synthetic experiments, not economic forecasts. In new 0.3.6 games random crises start OFF and citywide crime begins ON.
Same start. Multiple futures.
Average across selected seeds. Colors represent policy presets, not performance ratings.
The vertical axis may be zoomed to reveal differences; compare actual values in the table.
Means and ranges reflect random-shock variability, not statistical confidence intervals.
| Preset | GDP | Poverty | Unemployment | Healthcare | Debt/GDP |
|---|
→ Open Economic Calibration Lab
What this checks: repeatability, identical starting conditions, policy-label neutrality, basic numeric bounds, and response to one policy intervention. These tests cannot establish whether the formulas predict real economies.
What remains unvalidated: multi-year historical backtesting, behavioral parameters, uncertainty distributions, structural assumptions, real-world inflation/interest-rate dynamics, and independent expert review. No winner is declared.