Most trading journals have three columns: date, symbol, P&L. That tells you whether you made money this month — not why, and it gives you no way to separate the setup that’s actually working from the one being propped up by a couple of lucky trades. If you want to move your win rate on purpose instead of guessing, your journal needs five more specific fields.
1. Setup, described precisely
Don’t log “breakout” or “trend follow” as a category — log the actual conditions: breakout from what kind of consolidation, volume confirmation or not, which timeframe. The problem is every time you lump two trades under the same label while the entry conditions actually differed, you’re mixing data and quietly corrupting every stat you pull later. One trader logged 40 “breakout” trades over three months; broken down properly, there were four distinct variants, and only one had a win rate above 50%. The other three were bleeding.
2. R-multiple, not raw dollar P&L
Dollar P&L doesn’t let you compare a small-size trade against a large-size trade on the same scale. R-multiple — profit or loss divided by initial risk — normalizes every trade to the same unit. A trade that made $50 risking $25 is +2R, meaningfully better than one that made $200 risking $180, which is only +1.1R. Looking at raw P&L, the second trade looks stronger. Looking at R, the truth is the opposite.
3. Time since your last trade
Log the gap between the current entry and your previous one, and specifically whether the previous trade was a loss. This is the single most important behavioral field most journals skip. If most of your losing trades cluster in the “entered within 10 minutes of a prior loss” bucket, that’s almost certainly revenge trading — and it only becomes visible once you’re logging timestamps, not just staring at end-of-day P&L.
4. Position size as a percentage of account
Log size as % of account equity, not lots or coin count. It’s the only way to catch two dangerous patterns: size creeping up while you’re losing (trying to win it back) and size creeping up on a winning streak (overconfidence). Both mean emotion is steering, not the plan. One trader who reviewed three months of data found his average size jumped 40% on days with two or more consecutive losses — exactly when he should have been cutting size, not raising it.
5. Execution discipline score
After every trade, answer one simple question: did this trade follow the plan you wrote before entering — right setup, right size, right stop — exactly? A yes/no or a 1-5 scale works fine. This metric is entirely separate from outcome. A trade that followed the plan and still lost was a good process, regardless of the loss; a trade that ignored the plan and won anyway was a bad one, regardless of the win. Plot discipline score against R-multiple over time and the pattern jumps out: stretches of low discipline scores line up almost exactly with the sharpest drawdowns, independent of whether the strategy itself was sound.
None of this needs fancy tooling — a spreadsheet with five extra columns is enough to start. What matters is logging consistently, especially on the days you’d rather not look. After 4-6 weeks you’ll have enough data to answer what P&L alone never can: which setups are worth keeping, which to drop, and when the problem is you, not the market.
More on this in the Trading journal section.