The Real Cost of a Single Exception
You broke a rule once. The trade worked. Nothing bad happened, the account is slightly larger than it would otherwise have been, and by every measure available at the end of that session the decision was correct. This is the most expensive outcome that could have occurred, and understanding why is worth more than most things a trader can learn in a single sitting.
The Trade Is Not What Was Damaged
Evaluating an override by its own result is a category error. One trade is a sample of one drawn from a distribution, and a good outcome from a bad decision is entirely ordinary. What actually changed was not measured by the profit and loss at all. It was the status of the rule.
Before the exception, the rule was a boundary. It described something you did not do. Afterwards it describes something you usually do not do, and there is now a precedent for the circumstances under which you do. That is a different object, and it will behave differently the next time it binds.
The Second Exception Is Cheaper Than the First
The first override requires overcoming the fact that you have never done it. That is a substantial barrier and it is spent permanently the moment it is crossed. The second requires only a justification at least as good as the first one, which is a much lower bar, and it can point at a precedent that worked.
What follows is not usually dramatic. It is gradual. The conditions that qualify for an exception broaden slightly each time, always for reasons that seem sound in isolation, until the rule is being applied on the ordinary days and set aside on the interesting ones. Since the interesting days are the ones with the most at stake, the rule has been inverted: it is now in force precisely where it matters least.
Why the Profitable Break Is the Dangerous One
An override that loses money is self correcting. It hurts, the lesson is immediate, and the connection between breaking the rule and the outcome is easy to draw even though a single result proves nothing either way.
An override that wins provides the opposite feedback. It teaches that your judgement in the moment beat the written rule, which is a specific and confident conclusion drawn from a single observation. Nobody would accept that standard of evidence for a change to the strategy, and yet it is accepted routinely for a change to whether the strategy is followed. The break that pays is the one that recruits you.
It Also Corrupts the Record
There is a quieter cost that shows up later. A log containing a mixture of rule following trades and overrides, undistinguished, cannot answer the question of whether the method works. The results describe a hybrid of the strategy and your improvisation, and the two cannot be separated after the fact unless you marked them at the time.
This matters most when performance disappoints and you sit down to work out why. With a clean record you can tell whether the strategy underperformed or whether execution drifted, and those call for completely different responses. With a contaminated record you will most likely change the strategy, which is the wrong repair, and the actual problem continues untouched.
Handling One That Has Already Happened
The exception has occurred and cannot be undone, so the useful question is what treatment limits the damage. Recording it as a rule break, flagged as such, regardless of the outcome, is the first step and the one most often skipped when the trade made money.
The second is to describe the circumstances plainly. What time it was, what the session had done, what state you were in, what the justification sounded like. Overrides are rarely random. They cluster around identifiable conditions, and a short honest description of two or three of them usually reveals the pattern without any analysis at all.
The third is to decide, deliberately, that the rule stands. Not as a resolution or a promise, which have poor records, but as a stated conclusion that the exception was an exception and did not amend anything. That sounds like a formality. It is the difference between a rule with a break in its history and a rule that has quietly become optional.