Some ideas about how the peer review process in academic publishing can be used as a device for preregistering parts of an empirical analysis.