The weekly trading review in five questions, measured in R

A 30-minute weekly review in five questions: rule adherence, mistake cost by tag, setup expectancy, risk checks and one change, with a filled-in example week.

A trade journal only pays off if you read it. Once a week, set aside thirty minutes with the market closed and ask the same five questions in the same order. The answers come from numbers you already recorded, so the review is about counting, not about how the week felt.

It assumes you log every trade with a planned stop, a result in R and one mistake tag, as in the trade journal template. Without planned risk and tags, most of these questions cannot be answered.

Why a fixed set of questions

An open-ended review drifts toward the most memorable trade and ends with a vague resolution. Fixed questions make weeks comparable. The order matters too: behaviour first, then what it cost, then the setups, then risk, and only at the end a decision. Judging a setup before separating it from your own execution errors is the most common way to drop something that works.

The five questions

1. Did I follow my rules?

Count trades tagged NONE (plan followed) and divide by all trades. That is your rule-adherence rate for the week.

Rule adherence % = trades tagged NONE / total trades x 100

A losing trade that followed the plan is a pass; a winning trade that broke a rule is a fail. This number measures you rather than the market.

2. What did my mistakes cost, in R, by tag?

For each tagged trade, the cost is the actual result minus what the result would have been under the plan. A trade that should not have existed at all (OFF-PLAN, REVENGE) costs its whole result, because the plan result was zero. Add the costs by tag. The tag with the largest total is a candidate for next week's one change.

3. Per setup: trades, net R, expectancy. Keep, fix or drop?

Group trades by setup name and write trades, net R and expectancy (net R divided by trades). Then repeat on a plan-followed basis. That version tells you about the setup; the first mixes in your execution. Decide only when the sample is large enough (see the decision rule below).

4. Was risk sized correctly?

Three checks:

• Did any trade have a planned risk above 1R? That is a sizing error before the trade even started.
• Did any loss exceed about -1.2R? A little beyond -1R is normal from slippage and fees; much more usually means a moved or missing stop. Write your tolerance down once.
• Was the daily loss limit respected every day, including stopping for the day when it was hit?

If sizing is new to you, the position sizing with 1R guide shows how to turn a stop distance into a contract count.

5. What is the one change for next week?

Only one, written as a rule you can check next Friday, not an intention like "be more disciplined". Copy it into your pre-market plan.

A 30-minute agenda

1. 0 to 5 minutes: reconcile the journal with your official statement (trade count, fills, fees).
2. 5 to 10: question 1, adherence rate and the failed trades.
3. 10 to 15: question 2, cost of each tag.
4. 15 to 22: question 3, the setup table.
5. 22 to 26: question 4, risk checks.
6. 26 to 30: question 5, then add the week to the monthly roll-up.

If it regularly takes over an hour, the journal is missing fields or the tags are inconsistent.

A filled-in review for one week

Illustrative example: twelve trades on a micro index future over five days, with 1R set at $100 and results already net of fees. These are not real trades; the numbers are invented to show the arithmetic. Three setups are in the playbook: opening range breakout (ORB), first pullback (PB) and failed breakout (FB). The daily loss limit is -3R.

Illustrative week of trades (not real trades), results in R after fees

#	Day	Setup	Result	Tag	Result under the plan	Mistake cost
1	Mon	ORB	+1.80	NONE	+1.80	0
2	Mon	PB	-1.05	NONE	-1.05	0
3	Tue	ORB	-1.10	NONE	-1.10	0
4	Tue	FB	+2.10	NONE	+2.10	0
5	Tue	none	-1.00	OFF-PLAN	0 (no trade)	-1.00
6	Wed	PB	-1.60	MOVED-STOP	-1.05	-0.55
7	Wed	ORB	+1.50	NONE	+1.50	0
8	Thu	FB	-1.05	NONE	-1.05	0
9	Thu	PB	+0.40	EARLY-EXIT	+1.90 (target reached)	-1.50
10	Thu	ORB	-1.50	OVERSIZE	-1.00	-0.50
11	Fri	ORB	-1.05	NONE	-1.05	0
12	Fri	ORB	+1.95	NONE	+1.95	0

Week total: -0.60R, or -$60. Had every trade followed the plan, the week would have been +2.95R. Here are the five answers.

1. Rule adherence: 8 of 12 trades tagged NONE, so 67%.

2. Mistake cost by tag: EARLY-EXIT -1.50R, OFF-PLAN -1.00R, MOVED-STOP -0.55R, OVERSIZE -0.50R. Total -3.55R. Four trades turned a positive week into a losing one.

3. Setups:

Illustrative setup summary for the week

Setup	Trades	Net R (all)	Expectancy (all)	Net R (plan-followed basis)	Decision this week
ORB	6	+1.60	+0.27R	+2.10	Keep; sample too small to judge
PB	3	-2.25	-0.75R	-0.20	Keep; losses were mostly execution
FB	2	+1.05	+0.53R	+1.05	Keep; sample too small to judge
Off-plan	1	-1.00	-1.00R	0	Not a setup; covered by question 1

The plan-followed column swaps each tagged trade for its result under the plan. On raw numbers the pullback looks worst, at -0.75R per trade; on a plan-followed basis it is close to flat. Dropping it would punish the setup for your execution.

4. Risk: one trade (number 10) was planned at 1.5R, a sizing error. One loss (number 6) exceeded -1.2R, from the moved stop. Trade 3 at -1.10R is within tolerance. The worst day was Thursday at -2.15R, inside the -3R limit.

5. One change: MOVED-STOP, OVERSIZE and EARLY-EXIT all happened with no bracket order set at entry. One rule covers three of the four tags: "Every entry goes in as a bracket with the planned size, stop and target; no manual exit before stop, target or the written time stop."

Keep, fix or drop without overreacting

A handful of trades says very little about a setup. Illustrative example: suppose a setup truly wins 45% of the time, with wins of +1.9R and losses of -1R. Its true expectancy is about +0.30R per trade, and the standard deviation of a single trade is about 1.44R. The uncertainty of an average shrinks with the square root of the sample size:

Standard error of expectancy = standard deviation per trade / square root of number of trades

With 6 trades that is about 0.59R, so a sample average anywhere from roughly -0.9R to +1.5R (two standard errors either side) is unremarkable. With 20 trades it is about 0.32R; with 50, about 0.20R. Even 20 trades cannot reliably separate a +0.30R setup from a break-even one. That is why a single week should almost never decide anything about a setup.

A workable decision rule:

• Under 20 plan-followed trades: keep trading the setup at normal size and keep logging. No verdict.
• Fix execution when all-trades results trail plan-followed results by a wide margin; the setup may be fine.
• Fix the rules (rewrite, then hold for another 20 trades) when plan-followed expectancy is near zero after fees over 20 to 30 trades.
• Drop when plan-followed expectancy is clearly negative after fees over 30 or more trades, or when you cannot write the setup's rules clearly enough to follow them.
• Keep when plan-followed expectancy stays positive after fees over 30 or more trades and the losses look like the losses you expected.

The thresholds are judgement calls; what matters is writing them down before results arrive. The opening range breakout rules guide is an example of a setup written precisely enough to be graded this way.

The monthly roll-up

Each month, put the weekly answers on one table. Setup samples start to mean something here, and you can see whether each change stuck.

Illustrative monthly roll-up (not real results)

Week	Trades	Adherence	Net R	Mistake cost	One change	Stuck?
1	12	67%	-0.60	-3.55	Bracket at entry	Mostly
2	10	80%	+1.20	-1.10	No trades after the daily limit	Yes
3	11	73%	-2.40	-1.30	Written time stop for pullbacks	Yes
4	9	89%	+0.80	-0.40	Same rule, held	Yes
Month	42	76%	-1.00	-6.35

Read it honestly: the month lost 1R, and week 3 lost 2.4R even though mistakes cost only 1.3R, which is normal variance. Adherence rose and mistake cost fell. If that holds while setup samples grow past 20 to 30 plan-followed trades, you will have real evidence. If it does not, the table shows it plainly.

Common mistakes

• Reviewing P&L only. Dollar totals mix luck, size and discipline into one number. A green week with 50% adherence is a warning; a red week with 95% adherence may be fine.
• Changing several things at once. If you change entries, stops and the setup list in the same week, you cannot tell which change did anything. One change per week, held long enough to judge.
• Skipping losing weeks. These are the weeks the review exists for. A record of only good weeks will mislead you.
• Ignoring costs. Expectancy must be measured after commissions, fees and slippage. A setup that is +0.05R before costs can be negative after them.
• Hindsight targets. "Result under the plan" uses the stop and target written before entry, not the best exit you can see on the chart afterwards.

More routines are on the process topic page. Before Monday, check the news page for scheduled releases.

Examples use round, illustrative numbers and are not actual trades or results. Education only — not investment, tax or legal advice.

More on process