In controlled experiments when the reward follows a cue (Pavlov's bell for example) the dopamine spike follows the cue, which in turn drives a goal directed behaviour to get the expected reward.
Now look at the graph above - if the subject receives the reward 100% of the time after the behaviour - there is a moderate spike in dopamine, but if the subject is rewarded only 25% or 75% of the time, the spike which drives the goal directed behaviour is a little bit more pronounced. The biggest spike is when the subject is rewarded exactly half of the time...
Casino's know this, "bad boyfriends" know this, and now you know it. Interesting né?
