Example 1: Two losses, two different verdicts
Trade A follows the plan exactly: a valid setup, correct size of 1%, stop at invalidation. It loses. Trade B ignores the setup, doubles size to 2% on a hunch, and also loses. Both show red, but the verdicts are opposite: A was a good-process loss (variance, repeat it), B was a bad-process loss (fix it). Judged on outcome alone they look identical; judged on process they are a trade to keep and a trade to eliminate.