Empty trials are visually awkward. They sit as gaps, short lines, or footnotes. Designers of slides often remove them so the remaining points can breathe. The laboratory then loses the only public reminder that a setup already failed.
Keep failed runs in the figure even if they break the scale. A broken scale is more honest than a clean one. If the failure is a missing measurement rather than a bad value, show a mark at the date with no point. Captions should count them.
Marvel-themed stickers on a lab door or a character print on a notebook cover do not change the record. They are decoration. Sourceenginegrid has no connection to those rights holders; the notebooks still need stop reasons.
If a series has more failed runs than completed ones, say so in the first sentence of the briefing note. Directors can work with a high failure count. They cannot work with a figure that implies the path was already smooth.