Comment by LoganDark

So, an optimal policy is basically one that does not put you into a suboptimal position?