The results of experiments in quantum mechanics can be predicted correctly either by assigning a forward-evolving state to the system based on the preparationoutcome or by assigning a state that evolves backwards in time based on the measurement outcome. The latter picture admits some retrocausality without allowingmessages to be sent at a faster speed than that of light. This retrocausality allows some standard quantum paradoxes to be examined from a different viewpoint. It also allows closed causal cycles to be examined in the context of laboratory experiments. For a particular experiment, we find agreement with the principle that inconsistent causal loopshave zero probability of occurring, that is, only self-consistent loops can occur.