By The 888 VaultPublished Updated 9 min read
Almost every trader keeps a journal for a few weeks and abandons it. The reason is not discipline. It is that the journal they kept could not answer any question worth asking, so keeping it stopped feeling like it was for anything.
The fields that matter are written before the outcome is known. A journal filled in afterwards records the market; a journal filled in beforehand records your decisions, and only one of those is something you can improve.
Why most journals are useless
The standard journal records instrument, direction, entry, exit and profit or loss. Every one of those is knowable after the fact, which means the journal can be filled in at the end of the week from a broker statement — and if it can be, it will be.
A journal built from after-the-fact fields can only answer one question: what happened. That is the least useful question available, because what happened is mostly the market and only partly you.
The question worth answering is whether you did what you said you would do. That requires having said it, in writing, before the outcome was known — which is the entire design problem.
The fields that go in before the entry
Four things, written before the position exists. They take under a minute once the habit is there, and they are the whole of the value.
**The reason.** One sentence on why this idea, in language somebody else would understand. Not "looked good" — the structural thing you are responding to. If it cannot be written in a sentence, that is information.
**The invalidation.** The price or condition at which the idea is wrong, decided before you have any interest in the answer. This is the field people resist, because writing it down makes moving it later obviously a decision rather than an adjustment.
**The intended risk.** The percentage of the account you are putting at risk, and the position size that follows from it. Written as a number, not as a feeling.
**The plan for being right.** Where you intend to exit if it works, or what would make you hold longer. Deciding this while you have no position is a completely different exercise from deciding it while you are up.
- The reason, in one sentence
- The invalidation, as a price or a condition
- The intended risk, as a percentage and a size
- The plan for being right
The fields that go in after
Two, and only two are interesting. The outcome, which is bookkeeping. And the gap: what you actually risked and where you actually exited, against what you wrote beforehand.
That gap is the most honest number in trading. It is entirely within your control, it is invisible in a profit-and-loss column, and it improves faster than any other metric once you can see it.
Record it as a plain difference rather than a judgement. "Intended 1%, risked 1.8%" is useful. "Overtraded, felt bad" is a diary entry and changes nothing.
Separating errors from losses
The most valuable thing a properly-built journal does is let you tell an error apart from an ordinary losing trade, which look identical in a profit-and-loss column and are completely different events.
A losing trade that followed the plan is the cost of doing business. There is nothing to fix, and treating it as a failure is how people abandon a working method after four losses.
An error is a trade where the gap field is non-zero: you sized differently, moved the invalidation, entered without writing a reason, or took something outside what the plan covers. Errors are the only category worth working on, and a month with three errors and eleven losses is a good month.
Tag every trade as one or the other at review time. The ratio, tracked over months, is the closest thing to a progress measure that beginners have available.
Reviewing a month without flattering yourself
Review on a fixed date rather than after a bad week, because reviewing after a loss produces conclusions about the loss rather than about the process.
Read the reason field for every trade before looking at any outcome. If the reasons are vague, the problem is upstream of everything else and no amount of outcome analysis will find it.
Then count the errors, not the losses. Then look at whether the errors cluster — the same hour, the same instrument, the day after a large loss. Clusters are where the actual fix is, and they are invisible trade by trade.
Finally, resist the urge to change the method. Almost every review that ends in a new strategy was a review that failed to distinguish errors from losses.
Format matters less than you think
A spreadsheet is fine. A notebook is fine. A dedicated tool is fine and is not the reason anybody improves. The only requirement is that filling it in before the entry is fast enough that you actually do it under pressure.
If a field is regularly left blank, either remove it or find out what makes it hard to answer. A blank invalidation field is not a formatting problem; it is somebody entering positions without knowing where they are wrong, which is the single most useful thing a journal can surface about you.
Members bring a month of this to the community and go through it in public. It is uncomfortable, and it is where nearly all of the improvement in this room comes from.