The article discusses a well-known type of counterexample to Causal Decision Theory (CDT), in which CDT recommends an option that probably causes the best outcome while itself being evidence that it causes the worst. Intuition disagrees. Many philosophers accept that this justifies either modifying CDT or dropping it altogether. I argue to the contrary that (a) if intuition is right about this case, then transitivity of preference must be violated in another, but (b) this violation is untenable. I conclude that CDT stands.
The author wishes to thank Rachel Briggs and Huw Price for helpful discussion and two Philosophy of Science referees for helpful comments. The author began this article while on a Leverhulme Trust Research Fellowship, and he gratefully acknowledges the trust's support.