A major challenge of any search for gravitational waves is to distinguish true astrophysical signals from those of terrestrial origin. Gravitational-wave experiments therefore make use of multiple detectors, considering only those signals which appear in coincidence in two or more instruments. It is unclear, however, how to interpret loud gravitational-wave candidates observed when only one detector is operational. In this paper, we demonstrate that the observed rate of binary black hole mergers can be leveraged in order to make confident detections of gravitational-wave signals with one detector alone. We quantify detection confidences in terms of the probability $P(S)$ that a signal candidate is of astrophysical origin. We find that, at current levels of instrumental sensitivity, loud signal candidates observed with a single Advanced LIGO detector can be assigned $P(S)gtrsim0.4$. In the future, Advanced LIGO may be able to observe single-detector events with confidences exceeding $P(S)sim90%$.