Step 4 of 7
Design the intervention
Applied to Nova+. Control and treatment arms use the identical agent population and random seeds.
Intervention
Model assumption
Experimental design
Control
Baseline strategy
400 agents
Treatment
10% launch price cut
— agents exposed
- • Matched populations (same agents in both arms)
- • Common random numbers per replication
- • 20 replications × 12 rounds
Timeline
R1R12
Model assumption