On the standard, 3-place model of desert (A deserves X in virtue of Y), the assumption is that noninstrumental value exists in the obtaining itself of proportionality between treatment (X) and desert basis (Y). But do most people, when they make desert claims, really think of themselves as wanting to preserve this kind of value? A good analysis of a concept like desert will move between concrete usage and abstract theory, and so will need to offer a plausible account of the core concern people typically have when making everyday desert claims. Yet, when we consider the traditional model of desert as a 3-place relation, we find that it cannot plausibly account for the concern people have when making certain standard kinds of desert claims. Specifically, there are cases in which: (1) we insist that deserved treatment be given, even while we concede that there is no value itself in the desert basis or in the treatment; (2) we conclude that a person both does/ does not deserve some treatment; (3) we judge that a villain still not has received what he deserves, even while conceding that he has received punishment proportional to－or even exceeding－his crime.