Contributing to theory is considered the quintessential act of organizational scholarship. In response scholars propose novel additions of new theory and extensions of existing theories. Now, after several decades of these types of contributions, there exists a need to make this extensive body of novel additions to theory stronger. Sutton and Staw (1995) offer guidelines for distinguishing theory from non-theory, but less attention is given to distinguishing better theory from that which is simply good. Logical consistency of explanations, falsifiability, generalizability, parsimony, and accuracy have been offered as evaluative criteria. Making this theory stronger requires new perspectives on the nature of theoretical contributions. The most central of these is moving away from interpretations of simply whether a theory predicts to how well it predicts outcomes. Arguments for placing greater emphasis on meaningful measures of predictive accuracy as a means of keeping score are proposed and implications for evaluating theoretical contributions and research progress are discussed.