Module 2H

The session write-up: the plan, and the score against it

Builds on 2G. You can produce a direction. This turns it into something you committed to before the open.
Leads into Phase 3, where this form is revealed to be the five-layer read in disguise.

This curriculum's own. The architecture follows the working journal this method is actually run on.

The cardinal rule, before anything else

Two halves, filled at two different times, on purpose. The plan is written before the open, while price cannot pressure you. The score is written after the close, against what happened.

The gap between those two halves is where your edge or your leak lives, and you cannot see that gap unless both halves exist in the same structure every session.

If you cannot fill the pre-market half with specifics before the window opens, you do not take the trade. Not "you probably should not." You do not.

Three kinds of field, and the difference matters

BINDING. Written before the open, and it cannot be changed afterwards. If you find yourself wanting to revise a binding field mid-session, that impulse is the data. Log it as a deviation instead of editing the field.

OBSERVED. Written after the close, describing what actually happened. Never predictive.

COUNTED. A behaviour tallied without being forbidden. These exist so you can find out whether something is costing you rather than assuming it is. Do not treat a counted field as a rule you broke.

Same fields every session. That sameness is the entire point, because it makes sessions comparable, and patterns invisible in any one trade become obvious across twenty.

Part one: before the open

1. Context

Session date
Position against prior session value: above / inside / below
Condition: balanced / imbalanced up / imbalanced down
Daily structure: higher highs and lows / lower highs and lows / ranging
4H structure: higher highs and lows / lower highs and lows / ranging
Anchor direction, from the 4H - BINDING: up / down / no clear anchor
Does the 4H agree with the Daily: Y / N. If N, the 4H decides direction and you wait for it rather than overriding it
Price in the daily range: premium / equilibrium / discount
Levels in play, as actual prices. Prior day high and low, prior point of control, value area high and low, overnight high and low, VWAP, untouched gaps, unmitigated order blocks
Day type call - BINDING: rotation / trend
Bias - BINDING: long / short / neutral
If neutral, the single condition that resolves it
High-impact event inside or within thirty minutes of the window: Y / N

If that last answer is Y, the model is suspended. Stop here. This is a no-trade session and it scores as a complete one. The rule is binary, with no exceptions for "it might be fine" or "it was already priced in."

2. Location and playbook

Playbook - BINDING: reversal / continuation / none today
Why this playbook, in a sentence or two
The draw, meaning the side price is being pulled toward
Trade direction aligns with that draw: Y / N
The location, as a price. Where the question actually gets asked
Confluences required at that location. Minimum two, and they must be two distinct observations. One price wearing two labels counts once
The trade in one sentence, described as it will play out

Locations are map features, never triggers. Arriving at one is not permission. The trigger is the flow read.

3. The viability gate

Target, as a price. Target logic: fixed / named level / prior point of control
Target distance in points. Under ten stands the trade down. No exceptions
Invalidation, as a price. Stop distance: four points. Dollar risk at one contract: $20
Clear path to target, no opposing level or thin boundary in between: Y / N
Extension zone, and it is only valid if written now: ___ or none

Writing an extension zone before the open is what makes extending the target a plan rather than an improvisation. A blank here means you exit at the target.

Path extension is not available until you have demonstrated discipline at the fixed target, so most readers leave this blank for a long time. That is correct.

4. Stand-aside lines

I stand aside entirely if:
The thesis is wrong if price:

The stamp, at the open

Two fields. Do not revise them later.

Opening type - OBSERVED: auction / rejection-reverse / drive / test-drive
What told you: range shape, cumulative delta, price relative to prior value

Opening type	Favours
Auction	Standby. Most auction days produce no qualifying trade
Rejection-reverse	Reversal. The morning already contains the dynamic
Drive	Continuation. Direction is established
Test-drive	Continuation. The test already cleared liquidity

Does the opening type match the playbook you committed to: Y / N

If N, the honest options are to stand down, or to switch playbook and log the switch as a deviation. Both are legitimate. Quietly proceeding as though it matched is the one that is not.

2E is blunt about why this field exists: the wrong combination kills more setups than the wrong analysis.

Part two: after the close

5. Confirmation, the flow read

This is a confluence threshold, not a checklist you complete in order. Requiring one specific candle event as a gate collapses a multi-factor stack into a single-factor dependency, which is the same error as treating a gap as a trigger.

Reversal playbook. Four phases, and phase three is where the threshold applies. The phases are ordered. The factors inside phase three are not.

1. The drive into opposing liquidity. Which pool
2. The sweep. Occurred Y / N. Volume at the level: high with size transacting / thin. Volume is the tell. High plus reversal is a genuine sweep. Thin plus continuation was a real break
3. The flow turn. Tick every factor you actually observed, not what you assumed: cumulative delta flipped or diverged · delta swings on the reversal side · absorption · volume and size at the level · compression after the sweep · tape decelerated then re-accelerated · auction exhausted · structure shift printed. Total aligned: ___ of 8
4. The expansion. Followed Y / N

Three is the floor. Fewer than three and the sweep was just a sweep. No flow turn, no trade.

On the two-factor unlock. Two aligned factors can be enough, and the skill is knowing which two are doing the work. That is a statement about judgement, not counting. Three stays your floor until you can state, before the trade, which two you are relying on and why they are sufficient at that specific location. If you cannot, the count is still doing the work for you and the floor still applies.

The structure shift is one factor in the list, never a gate. Three factors with no shift can be valid. The shift alone never is.

Reversal invalidators, tick any that occurred: delta made a new extreme with the sweep · no absorption and no compression · continuation volume exceeded the rejection · multiple sweeps both sides · higher timeframe context agreed with the sweep, meaning wrong playbook · price was accepted beyond the level and built value there.

Continuation playbook. Four flow events, and this one is sequential.

1. Momentum approach. Aggression driving, delta aligned, cumulative delta making new extremes with it
2. Volume spike at the level. The next thirty to ninety seconds decide whether this is continuation or reversal
3. The trap. A pullback that looks like failure. Shallow in price, weak delta, cumulative delta pausing without reversing, volume lower than the approach
4. Acceleration. Tape jumps, delta makes new extremes, aggression eats through resting orders rather than being absorbed

Furthest step reached: 1 / 2 / 3 / 4
Entry type: anticipatory during the trap / confirmation on the acceleration / none

Use confirmation entries until the trap read is reliable. Anticipatory gets a better price and needs a confident read you have not yet demonstrated.

Continuation invalidators: delta diverged on the approach · volume on the pullback exceeded the approach, so the counter-aggression was real · absorption appeared on your side · multiple shallow pullbacks with no acceleration · congested path · higher timeframe contradicted the direction.

Then, either playbook: time the read completed · was it inside the window · trade taken Y / N · if not taken: hesitated / too late / size fear / distracted / doubted the read / correctly stood aside.

That last option is not a consolation prize. A read that did not reach three factors, and a trade you correctly did not take, are the same outcome, and it is a good one.

6. Entry

Read these off your platform. Do not estimate. Entry price, entry time, direction, stop price, lag from sequence completion to fill, and whether direction matches the bias you locked before the open.

7. What the trade actually did

Furthest adverse excursion, in points
Furthest favourable excursion, in points
Target reached. Extension zone reached. Extension taken
COUNTED: deepest retracement after +5 points

That counted field exists because the region between a few points of open profit and the target is where a plan most often gets abandoned, and nobody can tell you what it costs them without measuring it.

It is recorded, not policed. There is no rule here, and you should not invent one for yourself before you have twenty sessions of your own data.

8. Did the read hold

Did price reach the location you wrote · confluences that actually printed, out of how many required · did it play out as your one sentence described, yes / partial / no · did the reason for the playbook hold · deployment: as planned / switched playbook / off-plan / stood down · actual day type OBSERVED and whether it matches the binding call · bias resolved correct / wrong / no resolution · the level that actually mattered.

9. Management

The four valid exits. There are no others. Any exit outside this list is a process breach rather than a judgement call.

Stop loss, the structural invalidation set at entry
Fixed target
Extended target, and only if the extension zone was written before the open
The discretionary trail
none of these, and if you tick this, the exit was not a plan

Was a target order resting the entire time: Y / N. Stop modified: no / trailed to entry minus two / moved for another reason.

"None of these" and a "no" on the resting target order are the two most diagnostic answers on this form. Both mean the exit was a decision made under pressure rather than a plan executed. Neither is scored as failure. Both are counted.

The trail is discretionary, and that word is load-bearing. Around eight points you may trail to entry minus two, which cuts remaining risk from four points to two while protecting most of the unrealised gain. It is only justified when you can actually read the expansion failing: delta flipped against you · absorption on the opposing side · compression near the target · tape died on your side · price entered a high volume node that was not on the map. Cues observed: ___ of 5.

If you ticked zero cues and still trailed, that is anxiety wearing the costume of a rule. Log it as a deviation. The form counts it either way.

Circuit breakers. Three consecutive losing days triggers a review before the next session. Five losses in a rolling week triggers one full day off. A second trade in one session costs you the next full day, whatever either trade did. Record which, if any, triggered today.

10. Decision quality

Execution quality, the binary: clean or compromised.

Clean means the full process was followed: write-up done, bias formed, entry only on a confirmed read, stop honoured, no early exit. Compromised means any deviation at any point, regardless of whether the trade made money.

This binary is separate from the score below and both are needed. The score is granular and per-session. The binary is what makes aggregation possible, and the monthly review depends entirely on this one field being answered honestly.

A profitable trade where you skipped the write-up is compromised, not clean.

If this was a loss, classify it: execution error, process not followed · model loss, process followed and the trade failed · regime mismatch, the setup did not fit the day type.

These demand opposite responses, which is why the field exists. Treating a model loss as an execution error is how a working process gets abandoned.

Decision quality, 1 to 5. Five is executed exactly to plan or correctly stood aside. Three is partial, hesitant or late. One is no plan, or violating the plan you wrote.

Deviation, from this closed list only: none · entered without the sequence completing · traded the other playbook's sequence · traded outside the window · stop moved · target pulled or never placed · exited without a valid reason · exited a trade that had been well in profit · improvised setup · chased an extended move · skipped a setup that met every condition · second trade after the first.

Where did it break, answered on every session including wins: context and bias / location and playbook / target and viability / the trigger / the entry / the management / nothing, it worked.

Answering this on winning sessions is what makes the form worth filling. A win with a broken step is a warning that arrived disguised as a success, and it is the most commonly ignored signal in trading.

Then two lines. What I did not expect. And the behaviour to repeat.

What this form is measuring

Not profit. Three fields decide whether you are improving: decision quality, whether the setup appeared as written, and whether you stood aside when it did not.

A no-trade session, fully documented, scores a five. A profitable trade you improvised does not.

This is not encouragement. It is arithmetic. You are profitable at a win rate around a third, which means being wrong roughly two times in three is the normal operating condition of a working process. A system that only feels acceptable when you win cannot survive its own base rate.

Working alone

Nobody is reviewing this. That changes what the review cadence has to be.

The weekly mini-review, about fifteen minutes

How many sessions were clean and how many compromised. What the compromised ones had in common. One behavioural target for next week, stated concretely.

The monthly full review

Needs twenty to thirty logged sessions to be worth doing, so early on you run the weekly and skip this. Four questions, always in this order.

When do I execute cleanly and when do I not? A behavioural audit, not a performance one. Look for the conditions that produce breakdown rather than the trades where it happened. Day of week, time of the deviation, whether a win or a loss came immediately before, whether the write-up was skipped. The conditions are what you can change
What is my win rate on clean executions only? Calculate the two groups separately. If clean clearly beats compromised, the execution is the problem and not the model, which is the most useful thing this review can tell you. If clean executions are also losing at a high rate, that is a different and more important signal
What pattern appears in my losing trades? For each, find the earliest point of deviation. Most people find their losses cluster around one or two failure points
What does my journal language tell me about my state? Read the last few weeks of pre-market halves without looking at the outcomes. Hedging language tends to precede compromised execution. Absolute language tends to precede overconfidence errors. Clean conditional language tends to precede clean execution

Question four is the one that works best alone, because it needs no mentor and no outcome data. You are reading your own words back with the results hidden. It is the closest thing in this curriculum to an objective look at your own state.

The behavioural pattern log

A separate running document, updated monthly. Not a list of trades. A list of your own tendencies, each with when it appears, what it produces, and its current status.

This is what makes improvement cumulative rather than repetitive. Without it, every month's review rediscovers the same pattern from scratch.

After any deviation

Open the next session by reading that deviation before you fill in context. Not as punishment. Deviations recur because they are never re-read, and this costs nothing.

What this review is not

Not a P&L review. A profitable week from compromised execution teaches worse lessons than an unprofitable week from clean execution.

Not a trade critique. The unit of analysis is the pattern across trades. Obsessing over one losing trade is how people avoid looking at the pattern connecting them.

Session open

Self-check

Answer without looking back up the page

1. Why are the two halves filled at different times?

A correct answer names the state you are in for each, and what the gap between them reveals. Before the open there is no pressure and no position. After the close there is an outcome you already know. Those produce different writing from the same person, and the difference between them is the actual measurement.

2. What does a binding field mean, and what do you do when you want to change one?

A correct answer says you log the impulse rather than edit the field, and explains why: the wish to revise is itself the data, and editing destroys the only record that it happened. A form you can revise is a record of what you believe now, not of what you decided then.

3. Your target is eight points away and every confluence has printed. What happens?

A correct answer is one word: nothing. The viability gate stands the trade down under ten points and it does not consult how good the setup looks. If you wanted to make an exception, that is exactly what the gate is for.

4. A trade wins, but you entered before the sequence completed. What is the decision quality score, and what does "where did it break" say?

A correct answer separates outcome from process entirely. The score is low, the break is at the trigger, and the trade was profitable. All three are true at once, and holding them together is the whole skill this form teaches.

5. Why is deepest retracement after a few points of profit counted rather than made into a rule?

A correct answer explains the difference between measuring a behaviour and forbidding one, and why the order matters. A rule imposed before measurement removes the behaviour you were trying to study, so you never learn what it was costing you. Measure first, then decide.

If you could not answer 3 or 4

Do not trade yet, and reread the field types and the viability gate before you do anything else. Those two questions are the entire logic of this form and neither of them requires any chart-reading skill. They are about whether you can hold a rule that costs you a setup, and whether you can score a winning trade badly. Everything else here is detail.

One rep before you continue

Five sessions, both halves, then read them together

Do not begin by journalling everything. That is how this becomes a chore you abandon in three weeks.

Run the full form on five replay sessions, both halves, exactly as written. Fill the pre-market half before playing the session forward. Fill the post-close half afterwards.

Then read all five together and name the first pattern you can see.

One observation. That single sentence is the entire purpose of collecting the data, and if five sessions produce it, you have learned what the form is for far better than fifty sessions of dutiful filling would have taught you.

That completes Phase 2

You can now read a condition, read structure, anchor to a timeframe, name the locations, classify an open, read a window, form a bias, and commit it to paper before anything moves.

Phase 3 takes this form and reveals what it has been all along: a layer-by-layer walk through the five-layer execution read. Context, location, confirmation, execution, management. The sections you have just filled in are those five layers wearing different names.

From that point on this stops being an exercise and becomes the record of every session you trade. Which matters more than it sounds, because Phase 4 works on this record and on nothing else.

Log the module

You have finished Phase 2

Not a test, and nothing is marked. This form is the thing you would actually run every day, so whether it looks sustainable to you is the most useful thing you can tell me. About a minute.