During-session routine

It is 11:05 ET (23:05 ICT, winter). Your NQ plan was written before the open, your levels are marked, and you are down 1.5R after two breakout attempts that did not work. A third setup is forming, and it looks clean. Do you take it, at what size, and who decides: the person who wrote the plan at 08:30 ET, or the person sitting here now? This lesson is about the hours between the plan and the review, and the few rules that stop the second person from rewriting the first person’s work.

The job, in one sentence

During-session routine (quy trình trong phiên): the fixed list of things you do, never do and check between the open and the close Definition. From the first bar to the close, your job fits in one sentence: execute the plan’s setups, log each trade within 5 minutes of exit, re-check the plan at fixed times, and otherwise watch without touching House rule. Boredom is the job; excitement is the warning light.

A state change (thay đổi trạng thái) is a change in the facts your plan was written for: where price trades relative to prior value, whether the auction is balancing or moving, what is on the calendar in the next hour, and your own P&L and condition. A scheduled re-check (kiểm tra lại theo lịch) is a time written into the plan when you look at those facts on purpose, rather than whenever a candle scares you House rule.

Chart first: one session, three states

Scheduled re-checks: catching state changes on purpose

Your plan was written for a state. The session can leave that state, and the routine’s job is to notice at known times instead of by feel. Write three to five re-check times into the plan, in ET with your local time in brackets House rule. A reasonable start:

• 10:30 ET (22:30 ICT, winter): the initial balance (biên độ giờ đầu, the first hour’s range, M2.07) is complete.
• 12:00 ET (00:00 ICT): midday.
• 15 minutes before any scheduled release on your calendar, for example 13:45 ET on an FOMC day (M5.12 covers the release window).
• 15:45 ET (03:45 ICT): close-only starts.

At each re-check you answer the same four questions, out loud or on paper:

1. Where is price against the plan’s map? Inside prior value, accepted above or below it (M4.01 has the acceptance checks), or back inside after a failed break Context. This is a reading of where price sits, not a forecast.
2. Which branch of the plan does that state point to? If the state matches a branch you wrote (for example “if accepted above value, only pullback longs”), switch to it. If it matches none, there is no trade until the next re-check.
3. What is on the calendar in the next hour?
4. Where are you? Day P&L in R against your limit, trades taken against your count, and the self-check from your pre-session routine (M7.01).

The re-check changes branches; it never writes new ones. A setup you invent at 11:20 because the day “looks different” has no written invalidation and no risk math, so it fails M0.06’s four questions by definition. Whether a given state predicts the rest of the day is a separate question, graded in the lessons that own those states (M2.07, M4.01), and mostly the answer is “no edge shown”. The re-check exists so that you trade the plan you wrote and graded, not because the state forecasts the day.

The never-list

The never-list (danh sách không bao giờ): write yours, but start here House rule.

• Never trade without M0.06’s four questions answered (context, trigger, invalidation, risk math).
• Never move a stop away from price (M6.08).
• Never exceed the re-entry count (M6.10).
• Never trade outside the sessions in your plan.
• Never add to a loser.

Averaging down (trung bình giá xuống), precisely: adding to a losing position lowers your average entry but raises the money at risk on a thesis that is currently failing Definition. That is arithmetic: two contracts against the same stop lose twice as much. Scaling into planned entries, with both prices, both sizes and the cancel rule written before the session (M6.06), is execution; adding into red P&L is hope.

Circuit breakers

Circuit breakers (ngắt mạch): the session ends immediately when any of these fires House rule.

1. Your daily loss limit is hit. This course’s house example is −2R; M7.08 and the Risk course show how to set yours from your own drawdown numbers.
2. Two revenge signatures from M6.10 in one session (bigger size, a widened stop, no setup, a skipped cooldown, an unplanned instrument).
3. You notice you are tilted: the self-check collapsed.
4. Platform, data or connection problems you cannot fix in a minute.

“Done” means flat and closed, not “one more to make it back”. Breakers are written before the open because in the moment you will negotiate, and the person who set the breaker (calm, morning) outranks the person who wants to override it (tilted, at night).

Why expect the negotiation? Because it is measured, though not on you. Professional Treasury-bond futures traders regularly took above-average risk in the afternoon after losing money in the morning MeasuredOther market. In controlled choice experiments, people took more risk after a gain (the “house-money” effect) and, after a loss, were drawn to bets that offered a chance to break even MeasuredOther market. So wins can tilt you too. Whether these effects are as strong in your own NQ trading is plausible but not measured PlausibleOther market.

Between trades and at the close

Between trades: the reset. After every exit, log the trade, then stand up, drink water and take two minutes with no screens House rule. After a stop-out the longer 15-minute cooldown from M6.10 applies. The next trade gets a fresh head, not the previous trade’s residue, and that includes a win: the house-money result above is the reason a big winner gets the same reset as a loser.

Shutdown ritual (nghi thức kết thúc): in the last 15 minutes of your planned session, close-only (no new positions), flatten everything and confirm flat in the execution platform, not only on the chart House rule. For the regular session that is 15:45–16:00 ET (03:45–04:00 ICT, winter); if your planned session ends earlier, the ritual moves with it. Overnight holds are a separate strategy with separate planning, never an accident of not exiting.

Prop-firm note. On an evaluation or funded account the firm’s daily loss limit and its close-out time are firm rules, separate from your house breaker; set your breaker inside the firm’s limit so the firm’s rule never fires first. The Prop Firm course shows each firm rule in a dated card. If you place orders in one platform and read order flow in another (for example TopstepX for orders, ATAS X for the footprint), do the “confirm flat” step in the order platform.

Worked example

Synthetic day, illustrative numbers. 11:05 ET (23:05 ICT, winter). The trader is −1.5R after a full stop (−1R) on a breakout long and a −0.5R exit on the second attempt. A third setup forms: a rejection short at the session high, which is a branch in today’s plan. Checklist: planned setup? Yes. Risk? Standard 1R is $50. Re-entry count? Different setup, so allowed. Cooldown taken? Yes, 15 minutes. Room to the breaker? 0.5R. A full 1R stop would take the day to −2.5R, past the limit, so the plan’s rule applies: size the trade so a stop cannot breach the limit. The trader takes it at half size (0.5R risk), and the 2:1 target is hit: +1R. The day ends at −0.5R with the breaker never breached.

Alternate timeline: after the second loss the trader skips the log, skips the cooldown, and takes the third trade at double size “to get back”. It stops out: −1.5R − 2R = −3.5R, the daily limit blown through on one trade. Same market, same setups. The routine was the whole difference, which is why it is a lesson and not a suggestion.

Common mistake: Treating the never-list as “guidelines” that bend when a setup looks really good. The setups that tempt you to break the rules are exactly the ones the rules exist for; feeling sure is a fact about you, not about the market (M7.05). The rule bends for nothing, or it is not a rule.

Common mistake: Judging the breaker on one day. In the chart above price rallied after the −2R stop, and it is tempting to call the breaker a mistake. One day’s path after a stop is an anecdote; the breaker’s job is to cap the bad days, and you judge it over many sessions in your journal (M7.03), not on the one where it cost you.

Common mistake: Treating a re-check as a forecast. “Accepted above value at 10:30” tells you which written branch applies; it does not tell you the day will trend. Reading a state as a prediction turns a house rule into a lore claim.

Drill: write your re-check card

Before your next session, write a card with (1) your re-check times in ET with your local time in brackets, (2) the four questions, (3) for each re-check, the states you expect and the plan branch each one points to, and (4) your four breakers with numbers. During the session, tick each re-check when you do it, and note any trade taken outside a written branch.

Recap

• In-session job: execute planned setups, log within 5 minutes, re-check the plan at fixed times, otherwise watch without touching (a house rule).
• A re-check reads the state and switches to a branch you already wrote; it never invents a new one, and the state is context, not a forecast.
• The never-list (no trade without the four questions, no stop moved away, no extra re-entries, no unplanned sessions, no averaging down) plus four circuit breakers that end the session immediately.
• Breakers are pre-committed because losses and gains measurably change risk-taking in other markets; how much they help your own results is not yet measured.
• After every exit, a reset; at the end, close-only and confirm flat. “Done” always means flat and closed.

Further reading

• Coval, J. D. and Shumway, T. (2005), “Do behavioral biases affect prices?”, Journal of Finance 60(1), 1–34: futures market makers and afternoon risk after morning losses.
• Thaler, R. H. and Johnson, E. J. (1990), “Gambling with the house money and trying to break even”, Management Science 36(6), 643–660.
• Our lab: no measurement yet of in-session routines, re-check schedules or daily loss limits (search date 2026-11-09). The auction states you read at a re-check are graded in their own lessons (M2.07, M4.01).

Next: M7.03 turns today’s log into the review that closes the loop, after M7.01 built the plan this routine executes.

Educational content only — not financial advice. Futures trading involves substantial risk of loss.