was possible to see what really works, and what doesn’t.
Small bet to test hypotheses in a complex system. Lean startup and design approach. Trial group assumes representative of larger group, but nonetheless empirical. Catalog conditions of group and continue to test representative groups to invalidate hypothesis
