They really oughta have a "blind" approach to testing forensic evidence where samples are prepared by other forensic professionals to be sufficiently similar but not matches and real matches, neither of which are from a real trial. The forensic examiners should never know when they have evidence from a real trial.