All Experiments start with a hypothesis. Early on at The Commission we were deliberate about creating hypotheses for all the changes we wanted to implement. After a while, we stopped creating hypotheses, except for changes that had a large blast radius or high uncertainty.
