I kept a trade journal for two years before I noticed it was useless. Entry, exit, size, P&L. Every row accurate. Every row silent about the only thing that mattered.
I could tell you what I did on any day. I couldn't tell you why.
The column that changed my log
One day I added a single column: what I felt at entry. One or two words, written before I clicked buy. Bored. Anxious. Behind. Sure of myself. Annoyed at the last loss.
It felt silly. It took five seconds. Three months later it was the most useful field in the spreadsheet.
Sorting by that column showed me something the P&L column never could. My trades tagged "bored" or "behind" were a minority of entries, but they made up most of my losses. Trades tagged "calm" looked like the system I thought I was running.
These are my numbers from my own log, a small sample, and I wouldn't build a theory on them. But the direction was impossible to ignore.
The worst cluster was after a loss. Two or three entries in a row, each tagged "angry" or "behind," each slightly larger than the one before. In the P&L column that looked like three unrelated bad trades. With the feeling attached, it was one bad mood that lasted an afternoon.
Why fills alone can't show you this
A log of fills tells you the outcome of a decision. It says nothing about the state of the person who made it. And your state is the variable you can't see from the inside, because in the moment every trade feels reasonable.
Revenge trades don't announce themselves. They feel like "a good setup, finally." Size creep feels like confidence. Only on paper, with the feeling written next to the entry, does the pattern show up.
Think of it like a bug report with no stack trace. You know something broke. You have no idea which conditions produced it.
How to do it without turning it into homework
Keep the vocabulary small. Five or six words you reuse, so you can actually count them later. Mine are calm, bored, anxious, behind, angry and excited. If you need a paragraph to describe the feeling, you're writing a diary, not collecting data.
Write it before the order goes in, not after. Afterward, you'll tell the story that fits the result. A winner becomes "calm" in hindsight. A loser becomes "I knew it was a bad idea." Hindsight is a terrible witness.
Add one more field if you want it: did this trade follow the rules, yes or no. Then look at the two columns together. The overlap between "no" and a particular feeling is your personal trigger list.
One caution. Don't grade yourself on the column. A "bored" entry that wins is still a bad entry, because you got lucky. The goal is to learn which states to distrust, not to punish yourself for having them.
What to do with the pattern
After a month or two, count. How many entries per feeling, and what happened after each. You're not looking for a precise edge. You're looking for the moods you shouldn't be trading in.
Then turn it into a rule. Mine is boring: if I'd tag the entry "bored" or "behind," I don't place it. I wait for the next planned signal. That one rule has cost me almost nothing and saved me from plenty.
This is also, I think, the honest argument for automation. A bot has no column for feelings because it has no feelings to log. If you trade by hand, you do, and ignoring them doesn't make them go away.
If you want to see how we handle the rules side, our methodology and backtest data are published at v33systematic.com. But start with the spreadsheet. Add the column tomorrow, before your first trade.