The VariDrive technique was used to determine the reliability of transfer of initiation between two detonators in an SAU (Safety and Arming Unit). It was deduced that flying particles from the donor are the main cause of initiation, with air shock playing a minor role. The donor explosive mass that results in a 50% chance of initiation was determined in a computer-driven trial run using optimality statistics that delivered reliable estimates of the standard deviation. This result was used to determine the reliability of the donor/acceptor pair. However, the emphasis is on an investigation of pitfalls and the effects of underlying assumptions on the validity of the approach.