A trade that breaks every rule and wins is a bad trade; one that follows the plan and loses is a good trade. Over a real sample-size, the process determines the outcome, but any single result is mostly noise.
Grading trades on process in the trading-journal keeps recency-bias and hindsight-bias from rewriting the rules after every result.
Example: a trader takes a valid setup, risks 1R, and is stopped out before the stock reverses and hits the target. The journal grade is A for process, -1R for result. Nothing changes.
Related: trading-plan, trading-journal, sample-size, hindsight-bias