Massively multiplayer online games (MMOGs) maintain archival databases of all player actions and attributes including activity by accounts engaged in illicit behavior. If individuals in online worlds operate under similar social and psychological motivations and constraints as the offline world, online behavioral data could inform theories about offline behavior. We examine high risk trading relationships in a MMOG to illuminate the structures online clandestine organizations employ to balance security with efficiency and compare this to an offline drug trafficking network. This data offers the possibility of performing social research on a scale that would be unethical or impracticable to do in the offline world. However, analyzing and generalizing from clandestine behavior in online settings raises complex epistemological and methodological questions about the validity of such mappings and what methods and metrics are appropriate in these contexts. We conclude by discussing how computational social science can be applied to online and offline criminological concerns and highlight the "dual use" implications of these technologies.