The Trader's Mind · Lesson 5 of 6
Judge the Process, Not the Coin
A good decision can lose. A terrible one can pay. Confusing the two is how markets train people backwards.
Poker players call it resulting: judging a decision by its outcome. Folded correctly and the bluff would have failed? Genius. Played terribly and lucked out? Also, apparently, genius. Markets run the same trap at scale, because every trade's outcome mixes decision quality with a large dose of randomness — and the randomness pays and punishes without regard for merit.
The damage compounds through the feedback loop. Rewarded rule-breaks teach rule-breaking (the market just paid you to abandon your plan — a lesson your reward circuitry records enthusiastically). Punished good decisions teach abandonment: follow the rules into three statistically-ordinary losses (see the streak arithmetic in The Trader's Arithmetic) and the temptation is to scrap the rules at exactly the moment the maths says nothing is wrong. Trained by outcomes alone, a trader is being trained by a slot machine — randomly, and therefore toward superstition.
The reframe that fixes it: your job is making positive- expectancy decisions; the coin's job is landing. On any single trade, only the decision was yours. So score yourself on the part you control — was the setup per the rules? the size right? the exits placed? the calendar checked? — and let the scoreboard for early trading be the adherence rate, not the P&L. This is not a motivational poster; it is statistically literate self-measurement: over a large sample, P&L converges toward decision quality (that is what expectancy means), but on any week, the coin dominates. Judging a week's process by its P&L is sampling noise and calling it feedback.
Two practices make it real. In the journal, grade every trade A–D on process at entry time, before the outcome — then review whether your A-trades outperform your C-trades over months (that gap, not any single result, is the honest measure of whether your rules mean anything). And build the ritual sentence for good losses: "correct decision, losing outcome, nothing to fix" — said out loud, it is a vaccine against resulting.
Our machine took a full 40-pip stop on its first live day — and we published it as the system working: the stop did its one job. The same week, refused trades that would have paid were logged as correct refusals. If a machine's record can distinguish good decisions from good outcomes, so can a human's journal. It is the same discipline, one column over.