The Discipline Score: How to Put a Number on Rule-Following
Trading discipline stays a feeling until you define what counts as a rule. Here is how to turn your rules into checkable conditions, score adherence out of 100, and read the number without fooling yourself.
Table of contents
Every trading post-mortem ends the same way. You know what went wrong, you know it was avoidable, and the fix you write down is "be more disciplined next month."
That is not a plan. It is a mood. And it fails for a boring structural reason: you cannot manage a quantity you never measure. Your win rate gets measured. Your drawdown gets measured. Your adherence to your own rules, which is upstream of both, gets a feeling.
This is how to turn that feeling into a number.
Why Discipline Resists Measurement
The obstacle is not the maths. It is that most trading rules are written in a form that cannot be checked.
Look at a typical plan:
- Do not overtrade
- Only take A+ setups
- Manage risk properly
- Do not revenge trade
- Wait for confirmation
Every one of those is unscoreable. Two honest traders reviewing the same trade would disagree on whether "A+ setup" was met, and you reviewing your own trade will disagree with yourself depending on whether it won. A rule that can be argued about after the fact is not a rule, it is a preference.
The rewrite is mechanical. Take each rule and ask: on a finished trade, can this be answered yes or no without an opinion?
| Unscoreable | Scoreable |
|---|---|
| Do not overtrade | No more than 3 trades in a day |
| Only take A+ setups | Higher-timeframe bias aligned: yes or no |
| Manage risk properly | Risk per trade at or below 1% |
| Wait for confirmation | Entry taken after the displacement candle closed |
| Do not revenge trade | At least 30 minutes between a loss and the next entry |
The right column is boring, and that is the point. Boring conditions produce a number. The left column produces an argument.
Step One: Rules Become Conditions You Tag
In practice, a rule is a field you fill in on every trade plus a required answer for that field. TradingSFX handles this through confluences: a confluence becomes a tracked rule the moment you set a required answer on it, and stays a neutral data field if you do not.
Three shapes cover almost everything:
Yes/no conditions. The field is a checkbox, the required answer is Yes (or No). "Higher-timeframe bias aligned" with a required answer of Yes means a trade tagged No is a violation. This is the workhorse and most of your rules will be this shape.
Threshold conditions. The field is a number and the required answer is a minimum. "Minimum 2R target" with a required value of 2 passes at 2.5R and fails at 1.5R. Note the direction: the check is at-or-above, so write these as minimums. A cap like "risk no more than 1%" is better expressed as a yes/no field ("risk within limit") than as a number.
Allowed-list conditions. The field is text with a set of permitted answers. "Session" with London and New York on the allowed list means a trade tagged Asia is a violation, and one tagged London is not.
The work here is genuinely front-loaded. Defining five rules properly takes twenty minutes and then never needs doing again, and it is the only part of this whole process that requires thought.
Step Two: The Score Itself
Every trade where a rule had an answer produces one check. The score is:
Discipline score = total checks met / total checks made
That is it. Worked on a real-sized sample of 20 trades with three rules tracked:
| Rule | Checks recorded | Met | Adherence |
|---|---|---|---|
| Higher-timeframe bias aligned | 20 | 18 | 90% |
| Minimum 2R target | 18 | 12 | 67% |
| London or New York session | 12 | 6 | 50% |
| Total | 50 | 36 | 72% |
Discipline score: 36 of 50, so 72%.
Notice that 72% is not the average of 90, 67 and 50, which would be 69%. The score pools every check into one bucket rather than averaging the three rule percentages. That is deliberate: a trade where four rules applied says more about your execution than one where a single rule did, so it carries more weight. Averaging the rules instead would let a rule you barely ever record swing the headline number as hard as one you record on every trade.
The Three Things This Number Does Not Do
Any score you are going to trust needs its blind spots stated out loud.
1. Blank fields are ignored, not counted against you
If a rule has no answer on a trade, it produces no check at all. It does not count as met and it does not count as violated. This is correct behaviour, because a trader running a breakout strategy and a mean-reversion strategy should not be penalised on the breakout rules when trading the other one. Absent is not the same as failed.
But it is also the one way the score can be gamed, so here is exactly what that looks like. Take the table above and imagine you stop filling in the session field on the six trades where you traded outside your windows:
| Rule | Checks recorded | Met |
|---|---|---|
| Higher-timeframe bias aligned | 20 | 18 |
| Minimum 2R target | 18 | 12 |
| London or New York session | 6 | 6 |
| Total | 44 | 36 |
Same 36 rules respected, same 20 trades, same actual behaviour. Score: 36 of 44, or 82%.
Ten points of improvement bought entirely by leaving a box empty. Nobody sets out to do this, which is precisely why it happens. The defence is a habit, not a feature: fill the field in first, look at the score afterwards. A journal is a measuring instrument, and there is no version of this where fooling the instrument helps you.
2. It weights trades unevenly by how much you logged
A trade with five rules answered contributes five checks. A trade with one contributes one. If you are thorough on your calm days and sloppy on your tilted days, your calm days dominate the score. Worth knowing before you read too much into a single week.
3. It says nothing about whether your rules are any good
This is the big one. A discipline score measures conformity to your plan. It has no opinion on whether the plan makes money.
Perfect adherence to a bad strategy is perfect adherence to losing. If your score is 95% and your account is down, discipline is not your problem and grinding harder on execution will not help. That is an edge problem, and the number that answers it is expectancy, not this one. The two are diagnostic in sequence: expectancy tells you whether the plan is worth following, discipline tells you whether you are following it.
How to Read the Score
The level. Above 80% is strong adherence. Between 60% and 79% is mixed, and usually means one specific rule is dragging rather than a general looseness. Below 60% and your equity curve is not really a test of your strategy, because you did not run it.
The direction. More useful than the level. Split your recent trades in half and compare the average adherence of the older half against the newer half. A score climbing from 55% to 70% over 40 trades is a trader fixing something. A flat 75% for six months is a trader who has stopped noticing. TradingSFX shows this as a trend delta on the discipline card once there are at least four trades with rule data behind it.
The worst rule. The per-rule breakdown is sorted worst-first on purpose, because that is the only row that generates an action. Chasing a 90% rule to 95% is fiddling. A rule sitting at 50% is the one costing you.
What to Do With Your Worst Rule
There are only two honest responses, and the wrong one is the default.
The default is to try harder. Sometimes right, mostly not. If a rule has been at 50% for three months, willpower has already been tested and it lost.
The better first move is to check whether the rule earns its keep. Filter your journal into two sets, trades where that rule was respected and trades where it was violated, then compare expectancy across the two. Three outcomes:
- Respected clearly outperforms violated. The rule is real and it is costing you money every time you break it. Now the fix is structural: put the check somewhere you cannot skip it, like a pre-entry field you have to fill before the trade saves.
- No meaningful difference. The rule is folklore. You inherited it from a course or a Discord and it is not doing anything for your results. Delete it. A plan with four rules you keep beats a plan with nine you do not, and every dead rule you carry makes the score less informative about the rules that matter.
- Violated outperforms respected. Uncomfortable and worth sitting with. Usually it means the rule is drawn too tightly rather than being wrong outright, so widen it and re-measure.
Do check the sample size before acting. A rule with six recorded checks is a rumour, not evidence. Give it thirty before you delete anything.
One caveat on this comparison: the trades where you violated a rule are not a random sample. They tend to be the trades you took while tilted, tired, or chasing, so some of that performance gap is the state you were in rather than the rule itself. That does not invalidate the comparison, it just means treat a small gap as noise. This is the same self-selection that makes revenge trading so hard to see in your own data until you count it.
Where This Fits Against Other Journals
Rule tracking is not unique to us, and pretending otherwise would be silly. Two competitors do a version of it, checked on 7 August 2026:
Edgewonk has the Tiltmeter, which tracks how often you break your defined rules and maps that against your equity curve, plus Setup Checklists where you define the criteria for each setup and mark which were met per trade. It is the closest thing in the market to what this post describes, and it is genuinely good at the psychology angle.
TradeZella builds checklists and playbooks tied to trade data, where a playbook documents the conditions, entry criteria and risk parameters for a setup and trades link back to it.
The difference in our implementation is what the same tagged conditions get reused for. A confluence with a required answer feeds the discipline score, and the identical tag also feeds cross-filterable performance analytics, so the field answering "did I follow my rule" is the same field answering "does this rule correlate with my winners". You define the condition once and it does both jobs, which is what makes the delete-the-dead-rule test above cheap enough to actually run.
If you are running a funded account, adherence carries extra weight: prop evaluations fail people on rule breaches more reliably than on bad strategies, and what to track in a prop firm journal covers which rules are non-negotiable there.
In TradingSFX
The Discipline Score card appears on the dashboard Analytics tab, on the per-symbol and per-strategy analysis pages so you can see whether your adherence collapses on one instrument, and in the printable PDF performance report. It shows the score, a trend line of cumulative adherence across your trade sequence, and the per-rule breakdown sorted worst-first. If you have not set a required answer on anything, the card does not render at all, since there is nothing to score.
The AI Coach Profile also reads these numbers, and this is the one place worth being precise about: the respected and violated rule lists in your profile are computed arithmetically from your adherence, not written by the model. A rule needs at least three recorded occurrences before it counts as a pattern, then lands as respected at 80% or above and violated below 60%. The model writes the qualitative read on your trading, but it never gets to have an opinion about your adherence percentages.
On plans: the Basic plan is free forever at 10 trades a month and includes one custom confluence, which is enough to track a single rule and see whether any of this changes your behaviour. Pro at $19.99 a month raises that to five and unlocks the per-symbol and per-strategy analysis pages, the Coach Profile and the PDF report. Premium at $29.99 removes the confluence limit.
Start With One Rule
The failure mode when people first read a post like this is defining twelve rules on a Sunday evening, filling them in diligently for nine trades, and abandoning the whole thing in week three.
Pick one. Specifically, pick the rule you already know you break, because you have written it in your journal three times. Make it checkable, tag it on every trade for a month, and let the number tell you whether you are actually as inconsistent as you think.
Most traders are wrong about which rule they break most. That alone is worth the twenty minutes.
Your rules are already written down somewhere. Start scoring them and find out which ones you keep.
Published August 7, 2026. Competitor feature claims (Edgewonk Tiltmeter and Setup Checklists, TradeZella checklists and playbooks) verified 7 August 2026 against edgewonk.com and tradezella.com. All arithmetic in this post is worked openly from the stated inputs, and no statistic here is drawn from a study or survey.
Turn your trades into a real edge
Stop guessing what works. Log your trades, track confluences, and let the AI Coach surface the patterns you keep missing across every prop firm rule and strategy.
No credit card required · Start for free