Your edge dies at the click.
Knowing the setup isn't the hard part — executing it identically on trade 40 as on trade 4, after two losses, with money on the line, is. This page covers the failure loops, the biases that cause them, the pre-commitment devices that beat willpower, a full trading-plan template, and the QA-style improvement loop that turns your journal into a research pipeline.
1 · Separate process from outcome ↑ top
Trading gives you feedback that lies. A rule-breaking trade can win; a perfect trade can lose. If you learn from P&L, you learn the wrong lesson roughly half the time. Grade the decision, and let the money take care of itself over the sample.
2 · The loss spiral, and where to break it ↑ top
Willpower is not a plan
Every trader who has ever blown an account intended to be disciplined that morning. Discipline fails under emotional load, so the fix must be structural: limits enforced by the platform, size set before the session, a checklist that physically must be answered. Design the system so the tilted version of you can't do much damage — because that version will show up.
Every trader who has ever blown an account intended to be disciplined that morning. Discipline fails under emotional load, so the fix must be structural: limits enforced by the platform, size set before the session, a checklist that physically must be answered. Design the system so the tilted version of you can't do much damage — because that version will show up.
3 · The biases, in their trading costumes ↑ top
| Bias | How it shows up at the screen | Counter-measure |
|---|---|---|
| Loss aversion | Cutting +0.5R winners fast, holding −2R losers hoping for a return to entry | Pre-set partials; broker-side stop placed with the entry order |
| Confirmation bias | Adding indicators until one agrees with the trade you already want | Fixed indicator set; write the bear case before entry (prompt 06) |
| Recency bias | Abandoning a good system after three losses; oversizing after three wins | Judge systems on 30+ trade samples only; fixed size regardless of streak |
| Sunk-cost fallacy | "I've held this for two hours, I can't close it now" | Time stop: if the thesis hasn't played out in N bars, it's wrong — exit |
| FOMO | Chasing a move already 80% complete, entering with no stop level | Rule: no entry more than X × ATR from the setup zone. Missed moves cost nothing |
| Overconfidence | Size creep after a good week; skipping the checklist "because it's obvious" | Size changes only monthly, based on journal stats — never intra-week |
| Narrative bias | Building a macro story that makes it impossible to accept the stop | Levels decide, not stories. The stop is the story's falsification test |
| Gambler's fallacy | "Four losses in a row, the next one must win" → doubling size | Each trade is independent. Streaks are expected; size is not a function of them |
4 · The trading plan ↑ top
A plan you haven't written down is a mood. Copy the template below, fill it in once, and read it before every session. It should fit on one page — if it doesn't, you have too many setups.
Template
One-page trading plan
=== MY TRADING PLAN === 1. MARKETS & HOURS I trade: [e.g. GER40 08:00–10:30 UTC, US500/US100 14:30–17:00 UTC] I do NOT trade: [midday chop, 5 min around high-impact news, instruments not on this list] 2. MY SETUPS (maximum 3 — named, so I can tag them) A) [name] — regime: [ ] · location: [ ] · trigger: [ ] B) [name] — regime: [ ] · location: [ ] · trigger: [ ] C) [name] — regime: [ ] · location: [ ] · trigger: [ ] If it isn't A, B or C, it is not a trade. 3. RISK Risk per trade: [1]% · Max open risk: [3]R Daily stop: −[3]R → platform closed for the day Weekly stop: −[6]R → done for the week, review only Drawdown rule: at −[10]% from equity high, halve size 4. ENTRY CHECKLIST (all five must be YES) [ ] Regime matches the setup [ ] Price at a level I marked before the session [ ] Trigger candle has CLOSED [ ] R:R to nearest real obstacle ≥ 1.5, size calculated from the stop [ ] I am calm and within my daily limit 5. MANAGEMENT Partial: [50]% at [1.5]R, then stop to breakeven Trail: [behind each new structural swing / Supertrend] Time stop: if not working after [N] bars, exit I never move a stop away from price. Ever. 6. REVIEW Daily: log every trade (R, setup tag, process box, one note) Weekly: expectancy + rule-adherence % by setup Monthly: size review — only after 30+ trades of data 7. MY PERSONAL LEAK (be honest) My most expensive habit is: [e.g. re-entering immediately after a loss] The structural fix I have committed to: [e.g. 10-minute timer, platform closed]
Read it aloud before the open. The 30 seconds this takes is the highest-ROI activity in the trading day.
5 · Journaling that actually changes behaviour ↑ top
What to log — the minimum viable journal
Most journals fail because they're too heavy to maintain. Log these columns and nothing else; you can always screenshot the chart for detail.
| Field | Example | Why it's there |
|---|---|---|
| date / time | 2026-08-13 09:12 | Reveals which session windows pay you |
| market | GER40 | Per-instrument expectancy — often wildly different |
| setup tag | C1-pullback | The most valuable column. Enables per-setup stats |
| direction | long | Many traders are systematically worse in one direction |
| result (R) | +1.8 | The normalised outcome |
| process box | deserved-win | From Fig P1 — separates skill from luck |
| rule broken | none / "entered before close" | Adherence % comes from this |
| one note | "Volume confirmed, textbook" | One sentence. More than that and you'll stop doing it |
Eight fields, about 20 seconds per trade. Paste the whole log into prompt 17 monthly for the analysis.
The improvement loop — treat your trading like a system under test
- Baseline first. You cannot improve what you haven't measured. Thirty logged trades under current rules is your control group.
- Change one variable. Stop distance, or entry timing, or a market removed — not all three. Otherwise attribution is impossible.
- Pre-register the success criterion. Decide before the test what result would make you keep the change (e.g. "expectancy up ≥ 0.15R over 30 trades"). Deciding afterwards is how you fool yourself.
- Respect sample size. Under ~30 trades, differences are noise. Resist concluding from five.
- Keep a decision log. Record what you changed, when and why. Six months on, this is the most valuable document you own.
The progression — sim → small → scale (and the honest timeline)
| Stage | Goal | Advance when… | Typical duration |
|---|---|---|---|
| 1 · Simulation | Learn the platform, prove you can follow a checklist | 30 trades with > 90% rule adherence — outcome irrelevant | 2–6 weeks |
| 2 · Minimum size | Feel real emotion at trivial stakes | 50 trades, adherence > 90%, expectancy not negative | 1–3 months |
| 3 · Half size | Test whether the edge survives real pressure | 100 trades with positive expectancy across ≥ 2 market conditions | 3–6 months |
| 4 · Full size | Run the business | Ongoing — scale down whenever adherence or expectancy slips | Ongoing |
The two traps in this table
Skipping stage 2 — simulation teaches mechanics but not emotion; the jump from sim straight to meaningful size is where most edges evaporate.
Advancing on P&L instead of adherence — a profitable month with 60% rule adherence means you got paid for gambling. Advance on process metrics; the money follows the process, not the other way round.
Skipping stage 2 — simulation teaches mechanics but not emotion; the jump from sim straight to meaningful size is where most edges evaporate.
Advancing on P&L instead of adherence — a profitable month with 60% rule adherence means you got paid for gambling. Advance on process metrics; the money follows the process, not the other way round.
6 · The two-minute rituals that hold it together ↑ top
Before
Pre-session
Read the plan aloud · state today's bias in one sentence · confirm the risk budget and daily stop · note your own state (slept badly? halve size).
During
Between trades
After every close: log it in 20 seconds. After every loss: hands off the mouse for 10 minutes. Before every entry: the five-question checklist, out loud if needed.
After
Post-session
Screenshot each trade · assign its process box · write the one note · close the platform. Never re-open "just to look" — that's how flat days become losing days.
Where to go from here
Read Fundamentals for the mechanics and Risk for the maths, then pick one setup from the combinations and trade only that for 30 logged trades. Breadth is what beginners collect; depth in a single repeatable setup is what actually pays. The rest of this site is a reference library you consult as questions arise — not a syllabus to finish before you start.
Read Fundamentals for the mechanics and Risk for the maths, then pick one setup from the combinations and trade only that for 30 logged trades. Breadth is what beginners collect; depth in a single repeatable setup is what actually pays. The rest of this site is a reference library you consult as questions arise — not a syllabus to finish before you start.