Journal scores that actually change next-day behavior
Posted: Fri Sep 11, 2026 10:48 am
Journal scores that actually change behavior.
I run short scores, not novels:
- Plan adherence 1–5
- Setup quality A/B/C
- Execution 1–5 (slip, chase, hotkey errors)
- Emotion code (FOMO / revenge / calm / bored)
- Process PnL in R, separate from dollar vanity
Rule: if a score never changes a next-day rule after a month, drop it. Vanity writing is for memoirs; this is maintenance. Long abandoned journals teach nothing.
Sunday aggregation looks at process first, equity curve second. Ego somewhere near the parking lot. I also tag mistake codes so the weekly review can say "chase at open" instead of "bad week." Codes beat paragraphs when you're tired.
A worked example of one day's entry and what it changes. Tuesday, three trades.
- 09:38 ET, reclaim long, A setup, filled 1¢ worse than my limit, exited at plan. Adherence 5, execution 4.
- 09:44 ET, chased a move that wasn't on my list, marketable order, 4¢ of slip on entry, stopped. Setup C, adherence 2, execution 2, emotion FOMO, mistake_code "chase at open."
- 10:15 ET, ORB15 long, A setup, clean fill, exited at target. Adherence 5, execution 5.
Dollar PnL for the day was green. The session adherence score, which I set to the lowest of the day, is a 2. So for Wednesday, size stays at base and I add one line to the premarket sheet: no marketable orders before 09:40 ET. If Wednesday's adherence is a 5, the extra line can come off by Friday. That's the whole loop: score, one rule change, check.
Failure modes:
- Averaging the adherence score so one chase gets hidden by two good trades.
- Scoring after looking at PnL, which quietly inflates the numbers on green days.
- Adding new scores every week until the journal takes longer than the session.
- Writing mistake codes that are too vague to act on, like "bad entry."
When to skip scoring for a day: never entirely, but on a no-trade day I only log "no trades, plan was observe" and the reason. That still counts as adherence.
One practical habit: score the session before you look at dollar PnL. If adherence was a 2, a green day is still a process loss for sizing decisions tomorrow. That single rule has saved me more than any indicator tweak.
Diagram: the five scores I keep.
I run short scores, not novels:
- Plan adherence 1–5
- Setup quality A/B/C
- Execution 1–5 (slip, chase, hotkey errors)
- Emotion code (FOMO / revenge / calm / bored)
- Process PnL in R, separate from dollar vanity
Rule: if a score never changes a next-day rule after a month, drop it. Vanity writing is for memoirs; this is maintenance. Long abandoned journals teach nothing.
Sunday aggregation looks at process first, equity curve second. Ego somewhere near the parking lot. I also tag mistake codes so the weekly review can say "chase at open" instead of "bad week." Codes beat paragraphs when you're tired.
A worked example of one day's entry and what it changes. Tuesday, three trades.
- 09:38 ET, reclaim long, A setup, filled 1¢ worse than my limit, exited at plan. Adherence 5, execution 4.
- 09:44 ET, chased a move that wasn't on my list, marketable order, 4¢ of slip on entry, stopped. Setup C, adherence 2, execution 2, emotion FOMO, mistake_code "chase at open."
- 10:15 ET, ORB15 long, A setup, clean fill, exited at target. Adherence 5, execution 5.
Dollar PnL for the day was green. The session adherence score, which I set to the lowest of the day, is a 2. So for Wednesday, size stays at base and I add one line to the premarket sheet: no marketable orders before 09:40 ET. If Wednesday's adherence is a 5, the extra line can come off by Friday. That's the whole loop: score, one rule change, check.
Failure modes:
- Averaging the adherence score so one chase gets hidden by two good trades.
- Scoring after looking at PnL, which quietly inflates the numbers on green days.
- Adding new scores every week until the journal takes longer than the session.
- Writing mistake codes that are too vague to act on, like "bad entry."
When to skip scoring for a day: never entirely, but on a no-trade day I only log "no trades, plan was observe" and the reason. That still counts as adherence.
One practical habit: score the session before you look at dollar PnL. If adherence was a 2, a green day is still a process loss for sizing decisions tomorrow. That single rule has saved me more than any indicator tweak.
Diagram: the five scores I keep.