How do they do it in practice? How do the neurons know when they succeed / fail?
The paper uses this as the basis of learning and it looks like the neurons are given a random (electrical) stimulus when they miss the ball and a predictable stimulus when they hit it.