World 0: Money & Safety · Lesson 2 of 3

Luck is loud, process is quiet

9 min read

Judge decisions by the reasoning available beforehand, and recognise why winning badly is the most dangerous outcome.

What this lesson covers

  • Explain why a good outcome does not prove a good decision (and vice versa)

In markets you can make a bad decision and get a good outcome, or a good decision and get a bad outcome. Over short periods, results are dominated by : the ordinary scatter of random events.

This is not a minor technicality. It is the single biggest reason people learn the wrong lessons from their own experience, and it is why a beginner's first year can teach them habits that take a decade to unlearn.

Judging a choice purely by how it turned out is called . Everyone does it, because outcomes are loud and obvious while reasoning is quiet and invisible. Nobody feels the quality of a decision. They feel the profit and loss.

The fix is not to ignore results, results matter enormously over large samples. The fix is to evaluate the decision and the outcome as two separate things, because only one of them was under your control.

Two traders, two lessons, one of them false

Nia bets her rent money on one volatile stock and it doubles. Good outcome, terrible decision: she risked money she could not afford to lose on something close to a coin flip.

Here is the part that matters. If she concludes that worked, do it again, the win has just taught her a habit that will eventually be very expensive. The market paid her to learn something false.

Leo risks 1% of his savings on a researched plan with a defined exit, and loses. Bad outcome, sound decision. If he concludes my process is broken, one ordinary loss has talked him out of something that was working, and he will probably replace it with something worse.

Check your understanding

Which pair of judgements is correct for Nia and Leo?

Because a decision and an outcome are separate, there are four combinations rather than two, and reviewing both dimensions is what makes improvement possible at all.

FigureThe four combinations, and which one is dangerous
Bad decision · Good outcome. 'It worked, do it again'Risk of learning the wrong lesson10

The most dangerous quadrant

Good decision · Bad outcome, 'my process is broken'Risk of learning the wrong lesson7

Talks people out of working methods

Bad decision · Bad outcome, painful but honestRisk of learning the wrong lesson2

The lesson and the result agree

Good decision · Good outcome, the goalRisk of learning the wrong lesson1

Nothing to unlearn

Illustrative weighting of how much each quadrant teaches. The bottom-left is the trap: winning badly feels like proof, and reinforces a process that will fail later.

Notice that the two quadrants where the decision and the outcome disagree are the ones that mislead. When they agree, experience teaches accurately and cheaply.

The dangerous one is bad-decision-good-outcome, and it is dangerous precisely because it is pleasant. Nobody reviews a win. The reckless choice that paid off gets filed as skill, repeated with more size, and eventually meets the outcome it always deserved. Usually at the worst possible moment, because size grew while the process stayed broken.

Check your understanding

You review your month and find one trade that broke every rule you set, and made your largest profit. What is the correct response?

Common misconception

"I'll risk more when I'm confident, and less when I'm unsure."

Confidence is a feeling, and it peaks right after wins. Exactly when overconfidence is most likely and when a run of good luck is most likely to be mistaken for skill. Sizing off confidence means betting most when you're most likely to be fooled, and least when you're being appropriately careful. It is the sizing rule that maximises damage. Size off what you can afford to lose and what your rules say, not how certain you happen to feel today.

Go deeperWhy small losses and large losses are not symmetric

There's an arithmetic reason experienced traders talk about avoiding large losses far more than they talk about catching large gains.

Losing 10% requires an 11% gain to recover. Losing 25% requires 33%. Losing 50% requires 100%. Losing 90% requires 900%.

The recovery needed grows much faster than the loss that caused it, so damage is not linear. A controlled 1% loss is an ordinary cost of doing business. Twenty of them in a row still leaves you with about 82% of your capital and a functioning account. A single 50% loss requires you to double what remains just to get back to level, which most people never manage.

This is also why the previous lesson's warning about repetition matters here. Good decisions with bad outcomes are survivable when each one is small. The combination that ends accounts is a bad decision, a large size, and a bad outcome arriving together, and the first two are entirely within your control.

Key takeaways

  • Short-term results are mostly variance. They are weak evidence about decision quality.
  • Judge the decision by what you knew beforehand: reasoning, risk, and whether the money was yours to risk.
  • Winning badly is the most dangerous outcome, because success reinforces a habit that will fail later.
  • Confidence peaks after wins, which makes it the worst possible input to position sizing.
  • Losses are asymmetric: recovery grows far faster than the loss, so avoiding large ones matters more than catching large gains.

Think about it

Recall a decision of yours that worked out well. Would you make it again knowing only what you knew at the time, or did it simply happen to land?

Practice this lessonSort Quest: Good Call or Good Luck?

Terms introduced here: Expected value (EV), Variance, Outcome bias, Decision quality.

Sources and how to check them

  1. Barber, B. M., Lee, Y.-T., Liu, Y.-J., & Odean, T. (2014). The Cross-Section of Speculator Skill: Evidence from Day Trading. Journal of Financial Markets, 18, 1–24.

    What it found: In the complete record of Taiwanese day trading from 1992 to 2006, fewer than 1% of day traders predictably earned profits net of fees. The top 500 by past performance went on to earn 37.9 basis points a day after fees; the bottom group earned −28.9.

    Used here for: The separation of luck from skill. The study can distinguish them only because it has years of records per trader, which is the point the lesson makes about sample size.

    Read it carefully: It also shows skill exists and persists in a small minority, so it is evidence about the base rate rather than proof that nobody can do it.

Everything else in this lesson

  • Definitional: Outcome bias and the decision/outcome matrix are standard decision-analysis concepts, applied here to trading decisions.
Sort Quest: Three Ways to Risk It

Marking a lesson complete tracks your reading only; mastery still comes from the chapter quiz.