Measure your discipline, not your P&L
Financial results are too noisy to tell you whether you're improving. Four behavioural indicators to track instead.
PUBLISHED JULY 6, 2026
P&L answers a question you’re not asking
Over a day, over a week, often over a month, your result mostly measures variance. Two traders can execute exactly the same plan and finish the week at +3 R and −2 R. If you use that number to decide what to change in your method, you are learning from noise — and you mostly change things after losses, which is to say at the worst possible moment.
P&L is an excellent measure of outcome over a large number of trades. It is a very poor measure of progress over a small number of days. And days are what you live in.
What can be measured without noise
A good progress indicator has three properties: you control it entirely, it can be measured daily, and it doesn’t depend on what the market does. Four candidates meet all three conditions.
1. Preparation rate
The proportion of sessions where preparation was done before the first position: levels marked, calendar checked, size calculated. Binary, one line per day.
It’s the first indicator to track because it’s the first to degrade when things go wrong, and it always precedes the others. A drop in the preparation rate announces the run of impulsive trades — not the other way round.
2. Adherence to the calculated size
The percentage of positions whose size actually matches what your risk calculation produced. The point isn’t whether you can do the maths, but whether you execute what you worked out.
It’s the indicator that reveals a degraded mental state most reliably: the oversized position almost always shows up after a loss, never after a gain. Tracked day by day, it shows you the revenge trade before it costs you.
3. The number of rule breaks
Not their severity, not their cost: their count. Stop widened, position added into red, setup taken outside the written conditions. A broken rule is an event, and events get counted.
Counting rather than weighting avoids the classic trap: a rule break that ends in profit is still a rule break. If you weight by outcome, you learn that “it depends”, which is exactly the lesson you don’t want.
4. Journal upkeep
How many of the last thirty sessions have a closing entry. It’s a meta-indicator: when it falls, the other three become unverifiable, so their warning value disappears.
How to read them
Don’t look at them in isolation, and don’t look at them daily as a score. Look at them over thirty rolling days, compared with the previous thirty, and ask a single question: is the gap explainable?
A month at 70% preparation after a month at 85% is explained by a holiday, a busy period, a change of hours. That’s information, not failure. A month at 70% with no explanation is a signal — and that is exactly why you measure.
What this doesn’t replace
These four numbers say nothing about the quality of your strategy. A bad plan executed perfectly is still a bad plan — and iron discipline applied to a method with no positive expectancy will ruin you with admirable consistency.
They answer a different question, narrower and far more useful day to day: are you doing what you decided to do? For as long as the answer is no, analysing your P&L teaches you nothing, because you haven’t yet executed the thing you want to evaluate.