Part 4: The cost and discipline · Chapter 13 of 14 · 10 min

Building discipline

Discipline does not show in the result: a deviation costing 0.1 R per trade is worth about 0.8 R a month against a monthly spread of about 6 R, and about 2,950 trades are needed to see it. You build it on four numbers you can count right away (adherence, deviations by type, cost in R, review), on «if... then» rules written beforehand and on a weekly check. The example plan, with three flaws together, finishes below the band of luck alone (+24.4 R against the +30.4 R threshold): the band says when a difference is too big to be chance.

Why the results are not enough

In chapter 12 we saw that the prudent and the fearful lose a third of the plan's gain and that the result does not show it. The reason is the spread. On the plan the weekly sum has a deviation of 2.91 R (mean 1.49); over a month the spread is about 6 R. A deviation costing 0.1 R per trade, with 8 trades a month, is worth about 0.8 R a month: a seventh of the spread of one month. To see it in the results you need about 2,950 trades, that is about thirty years at this pace.

A big deviation shows sooner. The target cut to 1.5 R costs 0.51 R per trade, with 8 trades a month 4.1 R a month: after a year, about 49 R. Even this one, though, is lost in the noise inside a single month (the spread is 6 R).

Figure 1. Cumulative R of the plan (pink, +58.1 R), of «all together» (blue, +24.4 R) and the 5th-95th percentile band of 5,000 random reorderings of the 79 trades (dashed grey): the band reaches +30.4 R and +85.7 R at the end. «All together» finishes below the band.Data: gold, XAUUSDT perpetual future (real candles) · daily (D1) · from 15 Dec 2025 to 5 Oct 2026.
Read the chart in words

Daily gold chart with, below, the plan's curve, the lower curve of «all together» and two dashed curves forming a wide band around the plan.

The figure shows what can be seen. The plan, with its results reshuffled at random, finishes between +30.4 and +85.7 R in 90% of cases: a band 55 R wide. Only a behaviour that costs a lot, like «all together», leaves it. So discipline is measured on behaviour, not on the account.

The four numbers

They are counted on a journal (chapter 2), trade by trade, and require waiting for nothing.

NumberHow it is calculatedWhat it says
Adherencetrades executed as per plan ÷ trades the plan foresaw (skipped signals included)How far behaviour matches the plan
Deviations by typecount: skipped signal, early exit, moved stop, changed size, entry outside the planWhich deviation is the most frequent
Cost in Rthe method of chapter 12: same trades, one deviation at a timeWhat each one costs
Review doneweekly reviews done ÷ plannedWhether the check really exists

Adherence is a count, not an estimate: you know for certain how many trades you executed and how many you did not. It becomes an estimate only if you use it to say how you behave in general. With 79 trades and a true adherence of 90%, the error on the measured share is ±6.6 points; with 80%, ±8.8. With 30 trades they rise to ±10.7 and ±14.3. A 90% adherence on 30 trades cannot be told from an 80% one: compare two periods only when each has at least 70-80 trades, where the error falls to 7-9 points.

What counts as «executed»

Before counting, the plan must say what executing means. A trade is executed if, all four:

  1. the signal was the one foreseen by the plan;
  2. entry, stop and target were those of the plan, set before entering;
  3. the size was the one calculated by the risk rule;
  4. you did not touch stop and target, except for the exceptions written beforehand.

If a single condition fails, the trade is a deviation. A signal skipped without a written rule is a deviation even though you opened nothing: the plan said «enter» and you did not.

The «if... then» rules

The will of the moment is the weak part of a plan. A rule written beforehand takes the decision away from the wrong moment. It works like this: for each deviation you discovered in the journal you write a line, in the form «if X happens, then I do Y».

Situation«If... then» rule (example)
After two losses in a row (chapter 4)If I have lost twice, then I execute the next signal at normal size. If I skip it, I write it down with the reason.
The price goes half a risk against me (chapter 3)If the price drops half a risk, then I do nothing: stop and target stay where they are.
I want to close early with a small profit (chapter 6)If I want to close before the target, then I wait until I have written the reason; the only exception allowed is the one written in the plan.
I want to move the stop (chapter 7)If I want to widen it, then I do not: the stop moves only to protect, and only if it is written.
I want to raise the size after a loss (chapter 8)If I have just lost, then the size stays the calculated one.
I see a signal the plan does not foresee (chapter 9)If the signal is not in the plan, then I do not enter and I note it in the ideas log.
A negative day or week (chapter 10)If the week closes at a loss, then I look at adherence first: if it is full, I change nothing.
Days without signals (chapter 11)If more than 10 days pass with no entries, then I check that the plan's parameters have not changed and I wait.

They are examples: your rules must come from your journal, that is from the deviations that appear most often. Three or four rules are enough; a list of twenty you will not read again.

The checklist before every trade

Five questions, in order, written on the sheet or in the journal. If an answer is «no», you do not enter.

  1. Is the signal the plan's (and not an idea I just had)?
  2. Are stop and target calculated with the plan's rule?
  3. Is the size the one that risks the decided share of the account?
  4. Does the total open risk stay inside the limit (book 11)?
  5. Am I in a state to judge? That is: have I just taken a loss that pushes me to change something?

The fifth question measures nothing: it is the moment you notice you are under pressure. If the answer is «no», put it off by one candle. If the signal expires, you note it as a skipped signal with the reason: it is a deviation and you count it as such.

The weekly review

Ten minutes, always at the same time, with the market closed. Four lines:

  1. Adherence: how many trades executed out of how many planned.
  2. Deviations: which ones, and which «if... then» rule they point back to.
  3. Cost: how many R they cost, with the method of chapter 12 (even approximate).
  4. One thing to change in the written rule, only if the deviation has repeated at least three times.

The week's result in R is looked at afterwards, and decides nothing: 36% of the weeks of a perfectly executed plan are negative (chapter 10).

What to do

  1. Count the four measures on a real journal, for at least twenty trades.
  2. Write three or four «if... then» rules for the most expensive deviations.
  3. Use the checklist before every entry, for a month, without skipping it.
  4. Do the weekly review at a fixed time, and note whether you did it.
  5. Judge the result only on the window decided beforehand (chapter 10): to tell an edge of 0.4 R apart takes about twenty months (chapter 11), for one of 0.1 R decades.

Procedure in four steps

  1. Count adherence and deviations by type on the journal of the last trades.
  2. Calculate the cost in R of the two most frequent deviations.
  3. Write an «if... then» rule for each and put it in the checklist.
  4. Every week review adherence before the result.

Exercises

Chapter exercises

1. A deviation costs 0.1 R per trade, with about 8 trades a month. How much is it worth a month and why does it not show?

2. A 90% adherence on 30 trades and an 80% one on 30 trades: what can be said?

3. Which «if... then» rule is well written?

4. When do you change a rule of the plan on the basis of the weekly review?

Have a question or want to share your exercise?

Post in the community, or join the free signals room on Telegram.

Educational content, not financial advice. Trading involves risk.