According to Adams’ thesis the probability of a conditional is the conditional probability of the consequent given the antecedent. According to Stalnaker semantics, a conditional is true at a world just in case its consequent is true at all closest antecedent worlds to the original world. The chapter argues that Adams’ thesis and Stalnaker semantics are ways of cashing out the same ‘Ramsey test’ idea. Unfortunately, a well-known class of triviality theorems shows that Adams’ thesis and Stalnaker semantics are incompatible. Stefan Kaufmann has proposed (for reasons largely independent of the triviality theorems) a revised version of Adams’ thesis, which the chapter calls Kaufmann’s thesis. The chapter proves that combining Kaufmann’s thesis with Stalnaker semantics leads to ‘local triviality’ results, which seem just as absurd as the original triviality results for Adams’ thesis.