- focus on score-based systems: score-to-llr transformation, llr interval - train system on all data (no train/test splits, or cross-validation) - define the relevant score range(s) - CCF-score (1D): based on min and max observed in train set - CMC-score (2D): transform to CMC/N, or CMC for modus of N, CMC for min & max of N, ... - apply the system to a set of regularly space scores within the range - make plots based on these results