How to review your trades: tags, R, costs and your own base rates
Reviewing scalps by setup tag, in R and after costs, turns your own log into base rates and shows where execution left the plan. The same losing trade can be reviewed in a way that teaches or in a way that misleads.
Start in plain English
A review turns trades into information. For a scalper who takes many small trades, the useful review is not "which trades made money?" but "under which conditions does my plan make money after costs, and where did I stop following it?" This lesson shows a review method built on four columns: a setup tag, R, costs and plan adherence.
The core idea: review the decision against the plan, and the outcome against the base rate. A single trade answers the first question. Only a group of trades with the same tag answers the second.
What to record in a journal in general is covered by FennEdge's article Trading journal review. This lesson is about reading the record once you have it.
Two paths from the same start
The trade below is Path B from VWAP reclaims and rejections. All figures are Illustrative; Brindle Optics is fictional.
The trade. VWAP reclaim at 09:56 ET per the written definition. Plan: entry $2.89, stop $2.81, target $3.05, time stop 10:06, 625 shares, $50 planned risk. At 09:58 the stop triggered and filled at $2.79. Result: ($2.79 − $2.89 − $0.005) ÷ $0.08 = −1.31R (−$65.63).
Here is the same trade reviewed three ways.
Path AReviewed as a "bad trade"
Note: "Bad trade. Should have known the reclaim was weak; the 09:57 bar went right back under VWAP." The trader tightens the definition to need two closes above VWAP, effective tomorrow.
Path BReviewed against the plan
Note: "Plan followed: reclaim per definition v1, entry at plan, stop at plan, 2 cents of stop slippage. Unfavourable outcome. No change; v1 count now 7 of 20."
Path CNot reviewed
No note. The loss is in the broker statement, the tag is missing, and the slippage is not recorded.
The money lost was identical. What the trader learned was not. Path A used information from after the decision ("the 09:57 bar went back under") as a reason, and changed the plan after one draw. Path B judged the decision by what was knowable at 09:57 and left the outcome to the count. The rest of this lesson covers the columns that make Path B possible, how to read them in groups, and how to keep the outcome out of the verdict.
The four review columns
| Column | What it holds | Why it matters |
|---|---|---|
| Setup tag and version | "vwap-reclaim v1" | Groups trades so each plan builds its own base rate |
| R after costs | Gross R, fees, slippage, net R | Compares trades of different sizes and shows what costs take |
| Plan adherence | Yes / no, with what differed | Separates plan results from deviation results |
| Knowable at the decision | One line, written from the decision-time screen | Stops hindsight from entering the note |
The FennEdge practice log uses the same columns for practice trades: setup_tag, definition_version, gross_r, net_r, fees_usd, slippage_penalty_usd, plan_followed, deviation_note and knowable_at_decision.
R and cost columns
Record the planned risk once (1R in dollars), then express every result in R:
- Gross R = (exit − entry) × shares ÷ planned risk.
- Net R = gross R minus (fees + slippage) ÷ planned risk.
- Stop slippage = stop price minus stop fill, in cents and in R. In the Brindle trade: 2 cents, or 0.25R.
- Cost drag = the average of (fees + slippage) ÷ planned risk (1R), in R per trade across a group; say whether the spread is included.
Cost drag is often the largest number in a scalper's review. A plan with +0.15R gross expectancy and 0.12R cost drag keeps +0.03R. Expectancy and Slippage cover the parts.
Execution deviations
A deviation is any difference between the written plan and what happened:
- Entry above the plan's ceiling, or before the trigger.
- Size larger or smaller than the size rule.
- Stop moved or cancelled.
- Exit before the target or time stop without a plan reason.
- A trade the plan did not allow (outside its conditions, or after its daily cap).
Record deviations whether the trade won or lost. A deviation that made money is still a deviation; the trading plan lesson shows a cancelled stop that turned −1.25R into −2.55R, and the same habit can show a profit on another day.
Grouping by condition
One trade cannot show a base rate. Groups can. Useful groupings:
- By setup tag and version. Path A, B and C shares, expectancy after costs, with the count beside each.
- By plan adherence. Followed vs. not followed. A gap between them is the cost of deviations.
- By condition. Time window (Best time to day trade), spread at entry, catalyst type.
- By entry type. Marketable vs. passive, which also checks your fills against Paper trading assumptions.
Example (Illustrative): 30 VWAP reclaims; 22 with the plan followed, net +4.4R (+0.20R per trade); 8 not followed, net −6.0R (−0.75R per trade). The plan as written is not established as positive, since 22 trades leave a wide interval, but the deviation trades averaged −0.75R against +0.20R, a gap worth tracking as both counts grow.
Keeping the outcome out of the verdict
Outcome bias is judging a decision by its result (Baron and Hershey, 1988); hindsight bias is believing afterwards that the result was predictable (Fischhoff, 1975). Three habits help:
- Write the decision note from the decision-time screen. Name what was knowable, as in Path B.
- Grade adherence before looking at P&L. "Followed / not followed" first, the number second.
- Use a 2 × 2 view. Plan followed and won; followed and lost; deviated and won; deviated and lost. The third box is where lucky habits hide.
Observe vs. infer
From the Brindle Optics review (fictional; Illustrative):
| What the record shows (observed) | What a reviewer might infer | Why it is not established |
|---|---|---|
| Stop filled at $2.79, 2 cents below the stop | "My stops keep slipping" (from one fill) | One fill is one observation; average it over a group |
| The 09:57 bar closed back below VWAP | "The reclaim was obviously weak" | The 09:57 bar was not visible at the decision |
| Plan followed | "Plan followed, so nothing to learn" | The slippage and the count are still information |
| Third loss on this tag this week | "The setup stopped working" | Three losses are common even for a positive plan; see the sample-size section |
Odds box
That is what a personal base rate looks like: a point estimate, a wide interval and an after-cost expectancy close to zero. It is still more useful than any single trade, and it grows with every tagged trade.
What shifts the odds
In a review, "what shifts the odds" is what your own groups show. Until a group holds enough trades, treat any difference as a hypothesis to record, not a finding. Questions to ask of each group: does the Path A share differ by time window, spread bucket or catalyst type by more than the intervals allow? Does cost drag differ by entry type? Does adherence differ after a loss?
Cost of being wrong
| Review error | What it costs (Illustrative) |
|---|---|
| Changing the plan after one loss | The v1 sample stops at one trade; v2 starts with no evidence |
| Leaving out costs | A +0.03R plan recorded as +0.15R |
| Not recording deviations | Deviation losses (−0.75R per trade in the example) counted as plan results |
| Only reviewing losers | The sample loses its winners; the plan looks worse than it is |
| Only reviewing winners | Lucky deviations look like skill |
How many trades before this means anything
Every grouping divides the sample. Thirty trades split by time window into four groups leave about seven or eight per group, too few to compare. Review weekly for adherence and costs, which are visible quickly; compare base rates only when each group holds at least 20 and ideally many more trades. How many trades before your numbers mean anything explains the method.
What this lesson does not claim
- It does not claim any plan in this course has positive expectancy. The review figures are Illustrative.
- A thorough review does not improve the market's odds; it improves what you know about yours.
- Brindle Optics is fictional.
Go deeper
- FennEdge first.
- The Trading Journal syncs real fills from supported brokers and holds broker-synced trades only, not practice entries. Its Review page (Pro and Max, as of 2026-10-05) covers daily, weekly and monthly periods with net P&L after the fees your broker reports, a Setup tag, personal tags, notes and a "Within plan", "Outside plan" or "Not recorded" mark per trade, statistics by setup, and observations such as an entry within five minutes of a completed loss. Those observations are counts, not judgments of emotion or mistakes. An optional AI review cites only your own statistics and trades and uses your plan's AI allowance.
- Level Up (Plus and above) is a private challenge scored on your Journal net P&L, with an optional daily max loss per level. It does not track tags or entry timing, so keep process goals such as "every trade tagged" in your own log.
- FennEdge's article Trading journal review.
- MAE and MFE. How far a trade moved against and for you before the exit; the Advanced course covers them in MAE and MFE.
- R-multiples and profit factor. Comparing trades of different sizes; R-multiples in the Advanced course.
- Deliberate practice. Ericsson and colleagues (1993) tie improvement to feedback on specific skills; a review is that feedback.
- Related lessons: A trading plan as a hypothesis, Thinking in probabilities, Capstone: a measured 50-trade sim block.
Exercises
Six exercises follow this lesson: three concept checks, two calculations (cost drag across a group; the average result of deviation trades) and one practice task. Exercises, answers and progress are available after signing in.
Practice task: review your last 20 practice trades. Take the last 20 entries in your FennEdge practice log (download the practice log (CSV), column guide). Fill any missing tag, net R, cost and adherence columns; rewrite any review note that uses information from after the decision time; build the 2 × 2 view; and compute expectancy, cost drag and plan adherence by setup tag, with counts. Keep the log separate from the Trading Journal, which holds only real broker-synced trades; review real trades on the Journal's Review page (Pro or Max) instead.
Primary sources
- Baron, J. and Hershey, J. C. (1988). Outcome bias in decision evaluation. Journal of Personality and Social Psychology 54(4).
- Fischhoff, B. (1975). Hindsight is not equal to foresight. Journal of Experimental Psychology: Human Perception and Performance 1(3).
- Ericsson, K. A., Krampe, R. T. and Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review 100(3).
Sources checked 2026-10-05. Brindle Optics and all trades, counts and costs in this lesson are fictional and Illustrative. This lesson is education, not a recommendation to buy or sell any security. Past patterns do not indicate future results.