Process versus outcome

Trading is a domain where feedback is noisy and delayed, which makes it one of the worst possible environments for learning by result. Take a well-planned trade and lose; take a reckless oversized one and win. If you grade yourself by the profit column, you have just been taught the wrong lesson twice, and you will repeat the reckless behaviour because it paid.

The repair is to grade the decision independently of the outcome. Every trade lands in one of four boxes: good process and a win, good process and a loss, bad process and a win, bad process and a loss. The first two are both successes. The third is the genuinely dangerous one, because it rewards exactly what you must eliminate, and it feels like validation while doing so.

This is why a plan-followed flag belongs in the journal and why weekly review reads it first. A week of five losses that were all executed correctly is a good week of trading with a bad result — an ordinary sample from a positive-expectancy system. A week of three wins where two were revenge trades that happened to work is a warning, and reviewing on profit alone would have applauded it.

Bad process that wins is the most dangerous outcome in trading. It rewards exactly the thing you have to eliminate, and it feels like proof.

Outcome bias also distorts how traders adjust systems. Changing rules after a losing streak of ordinary length is the most common way an edge is destroyed: the strategy is replaced just before its normal recovery, and the next strategy is abandoned in its own normal drawdown. Sample size, not last week's colour, decides whether rules change.

Keep a private scorecard that has nothing to do with money: trades taken per the plan, trades skipped correctly, journal entries completed, reviews done. Those are the things you control. Over a year, they and only they are what produce the equity curve you are hoping to look at.