Metrics

How should we choose the control group in a situation where we have multiple treatments and time periods? A simple statistical simulation exercise