The 20-minute post-market review
Lesson 27 · about 8 min
The review is where the playbook improves. Twenty minutes, every trading day, in the same format, is enough; longer reviews get skipped, and a review skipped on the day of a loss is the one that mattered. This lesson gives the format, the log fields that make it possible, and the weekly and monthly roll-ups that turn daily notes into decisions.
The log
Every trade gets one row. Fill it in during the session where possible (entry, stop, size, grade) and complete it in the review. Minimum fields:
| Field | Example | Why |
|---|---|---|
| Date, time of entry | 2026-03-12 09:44 | Time-of-day analysis |
| Instrument | MES | Per-product sample |
| Setup | VWAP reclaim | Per-setup expectancy |
| Day type called (10:30) | Range | Read accuracy |
| Day type actual (close) | Range | Read accuracy |
| Direction | Long | |
| Planned entry / actual entry | 5,018.25 / 5,018.50 | Slippage, chasing |
| Stop (price) and 1R ($) | 5,012.25 / $30 per contract | Sizing check |
| Size | 3 | Sizing check |
| Exit(s) and R | half 5,027 (+1.5R), half 5,018.5 (0R) | Result |
| Net R after cost | +0.7R | The number that matters |
| Grade (before result) | A | Process |
| Note (one line) | Held runner to BE; trail would have got 2.5R | The lesson |
A spreadsheet is enough. The two fields most traders omit are the grade and the planned-versus-actual entry; both are the ones the review depends on.
The 20-minute format
Minutes 0 to 3: complete the log. Fill any blank fields. Net R after cost for each trade and for the day.
Minutes 3 to 8: grade the plan. Open the morning checklist. Three questions, answered in one line each:
- Was the day-type lean at 9:30 right? Was the 10:30 call right?
- Did the eligible setups trigger, and were they the right ones for the day that actually happened?
- Was anything on the sheet wrong (a level, a scheduled event, a filter)?
Minutes 8 to 15: grade the trades. For each trade, confirm the A or F grade and write the one-line note. For F trades, name the specific rule broken, using the checklist's language: "entered above trigger ceiling", "setup not eligible", "fourth trade". For A trades that lost, write "A, stopped, no change" and move on. Do not study A-grade losses; they are the cost of the edge.
Minutes 15 to 18: the one thing. Write a single sentence: the one behaviour to repeat or remove tomorrow. Not three things. One. It goes on top of tomorrow's checklist.
Minutes 18 to 20: mark the chart. Screenshot the session with entries and exits marked, filed by date. Annotate nothing except the trades; the screenshot is for the weekly review and for replay.
Then close the platform. The review is done when the timer says so.
What to look at and what to ignore
Look at:
- The grade distribution. If F trades exceed 10% of trades in a week, the problem is execution, not the playbook, and the fix is smaller size and the checklist.
- The day-type accuracy. Below 70% correct at 10:30 means the Module 5 read needs more replay sessions before it controls size.
- Planned versus actual entry, in R. Above 0.1R average is chasing or bad order types.
- Time of entry versus net R. The bucket analysis from Module 2.
Ignore:
- The daily P&L in dollars. It is one draw from a distribution and the review's job is the distribution.
- Whether an A-grade loss "could have been avoided". It could not, without a rule change, and rule changes come from the sample, not from one trade.
- Setups that worked but were not in the plan. Noting "would have made 3R on the midday breakout" is the first step toward taking it next time.
Weekly roll-up: 30 minutes on Friday
| Metric | This week | Last 4 weeks |
|---|---|---|
| Trades | 14 | 58 |
| A-grade % | 93% | 88% |
| Net R, all trades | +2.1R | +6.4R |
| Net R, A trades only | +2.9R | +9.1R |
| Win rate, A trades | 50% | 53% |
| Avg win / avg loss (R) | 1.7 / 0.8 | 1.6 / 0.85 |
| Day-type accuracy at 10:30 | 80% | 74% |
| Avg slippage, entry (R) | 0.04 | 0.05 |
| Rule breaches | 1 (fourth trade Tue) | 5 |
The gap between "net R, all trades" and "net R, A trades only" is the cost of your F-grade trades, in R. In the example above that is 2.7R over four weeks, which for many traders is the difference between a profitable month and a losing one, and it comes entirely from trades the checklist already forbade.
Monthly: the sample check
Once a month, count the A-grade trades per setup. Any setup with 30 or more A-grade trades gets an expectancy calculation. Anything under 30 stays in the "insufficient sample" column and continues at normal size. The next lesson sets the threshold at which a setup is judged; the monthly check only tells you how close you are.
Key idea: Twenty minutes a day in a fixed format, grading the plan and the process rather than the P&L, is what converts a log into a better playbook. The one-sentence "one thing" is the whole output.
Try it: Do the 20-minute review on your last session, or on a replay session, with a timer running. If it takes 40 minutes, you are studying A-grade losses or reliving the day. Cut those and run it again.
Recap
- Log every trade with setup, day type called and actual, planned and actual entry, R after cost, grade before result, and a one-line note.
- The review is 20 minutes: complete the log, grade the plan, grade the trades, write one sentence for tomorrow, file the chart.
- Study F-grade trades and plan errors; do not study A-grade losses.
- Weekly, compare net R of all trades to net R of A trades: the gap is what breaking your own rules costs.
- Monthly, count A-grade trades per setup toward the sample threshold.
See it drawn
Original diagrams for the ideas on this page. Illustrative, not real market data.