← All posts

The four columns that actually change how you trade

A journal fails in one of two ways. Either it collects nothing, or it collects so much that filling it in becomes the work and reading it never happens. Both end the same way: a folder of records nobody opens.

After a few hundred logged trades, most of what changes behaviour comes from four columns. The rest is useful context, but it is not what you review.

1. R multiple, not dollars

Dollars tell you how the month went. R tells you how the trade went. A +$300 win on a position twice your normal size is not a better trade than a +$150 win at standard risk — it is the same idea with more exposure. Logging the outcome in R makes trades comparable across account sizes and position sizes, which is the only way an average means anything.

2. The setup you called

Not the pattern you saw afterwards. The name you gave it before entry, from a short fixed list. Free text here is fatal: twenty spellings of one idea cannot be grouped, and if it cannot be grouped it cannot be counted. Five to eight named setups is usually enough to cover everything you actually trade.

3. Plan followed — yes or no

One boolean, answered honestly at close. This is the single column that most often explains a bad month, and the one people quietly skip. Split your results by it once and the conversation changes: if your plan-followed trades are profitable and your improvised ones are not, you do not have a strategy problem.

4. Session

Cheap to log, disproportionately informative. Most traders have one session where their edge lives and one where they are donating. You will not see it in a list of trades; you will see it immediately in a breakdown by session across fifty of them.

What you can leave out

Screenshots are pleasant and almost never re-read. Long narrative notes are worth writing when something unusual happened and a waste of a field when nothing did. Indicator values you did not use in the decision are noise. If a field has never changed what you did next, it has earned its way off the form.

The point of the record is not completeness. It is being able to answer, in one screen, whether your process is working — and having enough structure that the answer is not a matter of opinion.