Comment by lotsofpulp

Then how would the feedback loop be completed to provide information about which strategies are good and bad?