Inter-laboratory and inter-analyst experiments on confined fission-track length measurements in apatite have reported considerable variation, which can be attributed primarily to differences in analyst decision-making. Here we explore how different etching sequences, track selection approaches, and analytical methods influence track length and density measurements. The first experiment focuses on analyst judgement across different track selection criteria and light sources, and variation in effective etching time, using moderately annealed induced tracks in Durango apatite. Tracks are selected either with the usual approach of only choosing well-etched tracks, which we term the sufficiently-etched criterion, or a more relaxed criterion of simply being measurable without judging how etched they are. For the sufficiently-etched criterion, we separately analyze tracks found using exclusively reflected or transmitted light. We then follow the evolution of all selected tracks through further etching steps, and compare results based on length, angle, and dip, and placed in the context of a variable along-track etching rate model. All step etches show evidence of along-track etching rate falling from core to tip, and tracks measured using the sufficiently-etched criterion are slightly under-etched compared to the full latent track zone with enhanced etching rate. Minimizing effective etch time variation, by only using tracks selected after 10 s of etching, produced tracks averaging ~0.3 μm longer after 20 s than tracks selected using the sufficiently-etched criterion. Ensuring a minimum track etching time by first selecting measurable tracks after 20 s and etching for 10 s more produced tracks ~0.4 μm longer, but correctable to the latent length. Utilizing exclusively transmitted light diminishes mean track dip, and likely reflects lower efficiency in finding dipping tracks. In the second experiment, we repeatedly measure dipping confined tracks at various stages of annealing with and without using a cross-sectional view of the track, finding that the latter avoids an increase in scatter with increasing dip. We propose a two-step etching procedure that may provide more plentiful and reproducible data by reducing the impact of human decision-making, while still being relatable to published annealing data sets.