Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We examine the optimal disruption of dark (covert and illegal) networks. Of central importance is that an interventionist will generally have incomplete information about the dark network's architecture. We derive the optimal disruption strategy in a stylized model of dark network intervention with incomplete information and show how it combines features of two types of disruption considered in the literature: random failure and targeted attacks. In particular, the optimal disruption strategy encourages greater risk as less of the architecture is observed. A laboratory experiment finds that subjects tasked with disrupting a dark network qualitatively mimic the theoretical predictions.