Changes in the law on secondary participation in joint criminal enterprises since 1985 have been driven by a judicial policy aiming to counter a perceived threat to social order from criminal groups. This article argues that these changes enable prosecutions to succeed through the use of tenuous forms of evidence and epistemologically unjustified inferences which juries are permitted to draw from such evidence, a combination which effectively lowers the standard of proof. The lowering of the standard of proof has two effects. First, it makes it easier and cheaper for police investigators to construct prosecution cases, and secondly, it is likely to result in injustice. Many claims have been made that it does indeed have the latter effect. The police view the joint enterprise doctrine as a useful tool, and other parts of the criminal justice system are willing to collaborate with the development and refinement of the tool. This article explores the reasons why support is given to a development which requires such a drastic change in accepted legal standards. The Law Commission justifies these changes by its adoption of a theory that an individual who joins a group engaged in criminal activity changes her or his normative position to that of the group, and so may be held responsible for any crimes committed by any member of the group. However, the theory is unsupported by empirical research. Rather it is grounded in the findings of courts trying cases in which the joint enterprise doctrine is applied, and has its origins in theories which assign to criminal courts the responsibility to identify certain groups as enemies of society. Its application produces the populist characterisation of casual groups of young people as 'wolf packs'.