Two different methods for morally evaluating actions can be found in Albert the Great's works. The first, which I call the stages theory, requires that an act be evaluated in three stages: 1) its generic quality, or type; 2) its circumstances; and 3) whether it proceeds from virtue or vice. According to this theory, an act might be generically good, but done badly, or generically bad, but done well. The second method is an all-or-nothing principle. It requires that an act be good in each of a number of ways in order for it to be good overall. I call it the Dionysian principle, after pseudo-Dionysius, who stated that good is from a single complete cause, while bad is from any particular defect. It is not clear how these two methods go together, and at least one formulation of the Dionysian principle appears to be inconsistent with the stages theory. Thomas Aquinas, who studied under Albert, resolves the tension between the two methods by rejecting the stages theory and embracing the Dionysian principle. I argue that this disagreement between Albert and Aquinas leads them to describe certain hard cases very differently. Albert can maintain that agents in resolvable moral dilemmas perform a bad action, although it is the right action in the circumstances; Aquinas cannot, but rather claims that the agent in an apparent dilemma does not in fact perform the act he appears to perform. Finally, I draw out these differences by comparing both the stages theory and the Dionysian principle to certain Stoic antecedents.