Two popular cases in the post-Gettier wave of anti-luck epistemology can be used to construct a case directly relevant to the epistemology of testimony. Roughly, the case involves a hearer who bases her belief on testimony that is appropriately truth-connected and, in spite of this fact, her belief could have very easily become false. What is the right way to respond to this sort of case? There are at least two ways to respond. We may either argue that there is nothing epistemically undesirable about the case or pinpoint the undesirable features and propose an antidote. Here, I will explore several ways of responding to the case, gradually eliminating each one by one until the appropriate response becomes apparent. My own diagnosis is simple: A modal status fails to transfer from the beliefŐs basis to the belief. The antidote is equally simple: For testimonial knowledge, it is necessary that the modal status transfer.