There must be smaller wins as well, at some optimal frequency; this is called the “reinforcement schedule” in the sort of behavioral conditioning that relies on random reinforcement (as opposed to the “classical conditioning” of Pavlov’s famous dog, where strict correlations are established between events).
