Ah, useful to know. So I guess what my post was about, was preffering to scale the rewards between positive and negative ends of the spectrum, instead of punishment.