Application of social network analysis to reveal doping entourage of professional cyclists

Code: R10M1MA

There is a consensus in the international community that action must be taken against those behind-the-scene individuals who propagate and facilitate doping in sport. The first step must be to identify those responsible. Law enforcement agencies have used Social Network Analysis (SNA) to tease apart and identify key figures in complex investigations concerning drug trafficking, counter terrorism, money laundering, organised crime and people smuggling. It is proposed to apply SNA to study the social affiliations of convicted blood dopers. This will provide a baseline from which to gain a more sophisticated understanding of the blood doping environment.

Social Network Analysis requires three successive phases of activity. First, information about the links and relationships between the entities under investigation are extracted from existing data (e.g., newspaper reports, internet sites, public databases). Next a structural analysis is conducted to identify central members, subgroups and patterns of interaction between the parties. Finally, those relationships and associations are visualised using a software program that spatially distributes individuals according to their social interactions (i.e., individuals with strong ties are plotted closest together).

Main findings

The use of banned blood transfusions by athletes erodes the integrity of elite sport. Athletes are lured by the substantial performance advantage bestowed by blood doping, together with the realisation that its use cannot be detected by doping controls. Rogue doctors and transfusion technicians have set up extensive and sophisticated doping networks to meet this demand and the covert provision of specialist advice and equipment has proven to be a lucrative trade.

Novel strategies are required to identify participants in blood doping networks. One avenue to gather evidence about networks, and indirectly the athletes who are utilising such facilities, is to obtain eyewitness testimony via investigative interviews. For example, with regard to the transfusion networks that have so far been discovered, invariably some peripheral teammates and support staff of the doped athletes were aware of the existence of the network. However for this interview-based approach to be cost effective, there must be some way to flag persons of interest for interviews amongst the pool of several thousand athletes and support staff.

This study evaluated whether network analysis is an effective method for targeting interviewees for investigations into blood doping networks. Specifically, whether it was possible to identify and rank teammates and staffers in terms of their closeness to riders who had been implicated in blood doping practices.

The study determined that network analysis was capable of prioritising individuals based on the premise that individuals with the greatest exposure to doped athletes would have the highest likelihood of possessing relevant information. These findings have implications for antidoping authorities who seek to rationalise their allocation of scarce investigatory resources. This study makes recommendations for how network analysis can be incorporated into investigative operations.