A running list of rules I've broken, with a count next to each
I keep a list in my journal of every rule I've broken, with a count beside each one. It started as a way to be honest, and it turned out to be one of the most useful things I do.
The list has things like: entered before the candle closed, moved a stop away, traded outside my sessions, took a trade after my daily limit, ignored the news filter. Each time I do one, I add a tally mark. At the end of the month, I look at which ones have the highest counts.
What surprised me was that the top two rules accounted for most of the breaks. I had assumed I was making a wide range of errors, but it was the same couple again and again. Knowing that made it much easier to work on. I could put a specific safeguard for those two, such as an alarm or a checklist item.
The count also shows progress. When a number goes down over a few months, I can see the habit changing. That's more motivating than a general feeling of improvement.
A running list of rules I've broken, with a count next to each
Re: A running list of rules I've broken, with a count next to each
I keep something very similar and the same thing happened: two rules made up most of the breaks. For me it was entering before the candle closed and trading after my planned stop time. Seeing the tally made me realise I did not need a whole new system, I needed two specific fixes. For the first one I added a candle timer, and for the second I set an alarm and close the platform when it goes off. Both counts dropped a lot over a couple of months.
It’s Fairman 
Re: A running list of rules I've broken, with a count next to each
Something I added is a note next to each tally about the situation, just a few words like after a loss or tired or news day. After a while those notes showed a pattern too. Most of my breaks happened late in the session or right after a loss. So now I have a rule about the situation, not just the action. When I am in that situation, I take a break before the next trade. It is a small change to the journal, but it moved the problem from willpower to a clear trigger I can spot.
It’s Fairman 
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: A running list of rules I've broken, with a count next to each
A count on its own can hide which rule is weakest, because the rules get different numbers of chances to be broken. Dividing each tally by the number of times the situation came up gives a rate that can be compared.
Say a month has 120 trades with 8 entries before the candle closed, which is under 7 percent. If the daily limit was hit on only 4 days and trading carried on after it on 3 of them, that rule is broken 75 percent of the time it applies, even though its tally looks small.
Putting the cost in R next to each tally is the other column worth having. A rule broken often but cheaply may matter less than one broken twice that cost 4R between them.
The situation notes above could be counted the same way. If after a loss is attached to half the breaks, comparing that with how often trades follow a loss at all shows whether the trigger is real or just common.
How do you count a single trade that breaks two rules at once? Early entry plus an oversized lot is a common pair, and counting it twice could make the top two look more separate than they are.
Say a month has 120 trades with 8 entries before the candle closed, which is under 7 percent. If the daily limit was hit on only 4 days and trading carried on after it on 3 of them, that rule is broken 75 percent of the time it applies, even though its tally looks small.
Putting the cost in R next to each tally is the other column worth having. A rule broken often but cheaply may matter less than one broken twice that cost 4R between them.
The situation notes above could be counted the same way. If after a loss is attached to half the breaks, comparing that with how often trades follow a loss at all shows whether the trigger is real or just common.
How do you count a single trade that breaks two rules at once? Early entry plus an oversized lot is a common pair, and counting it twice could make the top two look more separate than they are.
-
FXS Prop Desk
- Posts: 1689
- Joined: Wed Oct 07, 2026 9:52 am
Re: A running list of rules I've broken, with a count next to each
A list with counts is very effective because it shows which rules are actually problems. Most traders break the same two or three rules repeatedly, and focusing on those is much easier than trying to improve everything at once.
Did you see the counts change after a few months? Often the top items drop quickly once they are visible, and new ones appear lower down. That shift is a good sign of progress, even if the list never becomes empty.
Did you see the counts change after a few months? Often the top items drop quickly once they are visible, and new ones appear lower down. That shift is a good sign of progress, even if the list never becomes empty.
FXS Official Team | The Idea Lab & Strategy
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: A running list of rules I've broken, with a count next to each
When a count drops, it is worth checking that the rule is still being recorded the same way before calling it progress. Honest tallies tend to soften at exactly the point where the numbers start to look good.
Some rules can be checked against the platform history instead of memory. Trading outside the planned session shows up in the entry timestamps, and size above plan shows up in the volume column.
Say the tally shows 2 out of session trades for the month while the exported history shows 6 entries outside the window. The gap of 4 is itself a useful number, because it measures how much the list relies on goodwill.
Other rules have to stay on trust, such as entering before the candle closed or ignoring the news filter. Those are the ones where a sudden drop deserves the most suspicion.
A rule that gets a hard block, like the alarm and closed platform described above, should come off the tally and be noted as enforced. Otherwise its count falls to zero and makes the overall trend look better than the habits really are.
Which of the rules on these lists could be checked automatically from the trade history? It would be interesting to see how many are left once those are taken out.
Some rules can be checked against the platform history instead of memory. Trading outside the planned session shows up in the entry timestamps, and size above plan shows up in the volume column.
Say the tally shows 2 out of session trades for the month while the exported history shows 6 entries outside the window. The gap of 4 is itself a useful number, because it measures how much the list relies on goodwill.
Other rules have to stay on trust, such as entering before the candle closed or ignoring the news filter. Those are the ones where a sudden drop deserves the most suspicion.
A rule that gets a hard block, like the alarm and closed platform described above, should come off the tally and be noted as enforced. Otherwise its count falls to zero and makes the overall trend look better than the habits really are.
Which of the rules on these lists could be checked automatically from the trade history? It would be interesting to see how many are left once those are taken out.
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: A running list of rules I've broken, with a count next to each
Consider a strategy with a true expected monthly return of 2% and a standard deviation of 4%. Roughly one month in six will be below -2%, and about one in forty will be below -6%. A losing month is normal. A losing quarter happens with a decent probability. A year with several red months is routine.
New traders, expecting a smooth climb, panic at the first red month. Understanding the variance helps set expectations. A simple calculation with the strategy's own monthly standard deviation yields the odds of a negative month, a negative quarter, or a drawdown of a given depth.
The standard deviation of monthly returns also determines how long it takes to verify an edge. With a 2% mean and a 4% standard deviation, the standard error of the mean after 12 months is about 1.15%, so the mean is less than two standard errors from zero. Verifying the edge takes years of monthly data, which is why trade-level analysis is more practical.
Plotting a histogram of monthly results, with the normal range marked, gives a visual benchmark.
A simple graph helps: plot each month's return as a dot with a band showing one and two standard deviations around the long-run mean. Most dots fall inside the inner band. A dot outside the outer band is worth investigating, but several dots inside the band on one side in a row is a more interesting signal, a hint that the mean itself may have moved.
One practical programme for converting memory rules into structure takes an afternoon. Hour one: list every rule in the plan. Hour two: mark each as memory-based or structure-based, and rank the memory-based ones by how often they've been broken. Hour three: for the top three, design a structural fix. A spread filter in the platform, a lot size lock, an alert that fires at the loss limit. Hour four: implement and test them on demo.
The effect compounds. Each rule moved into the structure frees attention for judgment calls that can't be automated, and reduces the number of decisions made under stress. Over months, the plan shifts from a list of intentions to a set of working mechanisms.
Not everything can be automated, and some rules need human judgment. The goal isn't to remove the trader from the process but to remove the avoidable failure points, and leave discipline for the moments that really need it.
What's the personal monthly standard deviation, and how often does a red month show up?
New traders, expecting a smooth climb, panic at the first red month. Understanding the variance helps set expectations. A simple calculation with the strategy's own monthly standard deviation yields the odds of a negative month, a negative quarter, or a drawdown of a given depth.
The standard deviation of monthly returns also determines how long it takes to verify an edge. With a 2% mean and a 4% standard deviation, the standard error of the mean after 12 months is about 1.15%, so the mean is less than two standard errors from zero. Verifying the edge takes years of monthly data, which is why trade-level analysis is more practical.
Plotting a histogram of monthly results, with the normal range marked, gives a visual benchmark.
A simple graph helps: plot each month's return as a dot with a band showing one and two standard deviations around the long-run mean. Most dots fall inside the inner band. A dot outside the outer band is worth investigating, but several dots inside the band on one side in a row is a more interesting signal, a hint that the mean itself may have moved.
One practical programme for converting memory rules into structure takes an afternoon. Hour one: list every rule in the plan. Hour two: mark each as memory-based or structure-based, and rank the memory-based ones by how often they've been broken. Hour three: for the top three, design a structural fix. A spread filter in the platform, a lot size lock, an alert that fires at the loss limit. Hour four: implement and test them on demo.
The effect compounds. Each rule moved into the structure frees attention for judgment calls that can't be automated, and reduces the number of decisions made under stress. Over months, the plan shifts from a list of intentions to a set of working mechanisms.
Not everything can be automated, and some rules need human judgment. The goal isn't to remove the trader from the process but to remove the avoidable failure points, and leave discipline for the moments that really need it.
What's the personal monthly standard deviation, and how often does a red month show up?