How Do I Build Confidence in My Own Judgment When Developing a Discretionary Trading System?

Every term, a student asks me some version of the same question: how do I build confidence in my own judgment when developing a discretionary trading system? They have studied price action, read the books and watched a few hundred charts. Yet the moment a real setup appears, they freeze, second-guess, or take the trade and then panic out of it. The problem is rarely a lack of knowledge. It is a lack of evidence that their knowledge works.

That distinction matters. Confidence in trading should not be a feeling you summon with positive affirmations in the bathroom mirror. It is a conclusion you reach after collecting data on your own decisions. Done properly, that data tells you which of your instincts deserve trust and which belong in the bin.

In this article, I will walk you through a practical framework I use with intermediate traders. We will cover how to structure your discretion, how to measure it, how to separate skill from luck, and how to test your judgment against a mechanical baseline. You will also get honest warnings, worked examples, and clear next steps.

Table of Contents

• Why Confidence Must Be Earned, Not Felt
• How Do I Build Confidence in My Own Judgment When Developing a Discretionary Trading System?
• The Decision Journal: Measuring Your Judgment
• Separating Good Decisions from Good Outcomes
• Testing Your Discretion Against a Mechanical Baseline
• Critical Warnings: What to Be Careful Of
• Next Steps for Developing Your Judgment
• Frequently Asked Questions

Why Confidence Must Be Earned, Not Felt

Let us start with an uncomfortable fact. Retail traders, as a group, lose money. Brokers in the UK and EU are required to publish the percentage of retail CFD accounts that lose, and the figures typically sit between roughly 65% and 80%. Many of those losing traders felt very confident at the time.

So feeling confident proves nothing. What we want is calibrated confidence, where how sure you feel matches how often you are actually right. If you say you are 70% sure of a setup, it should work about 70% of the time across a large sample.

Why this matters for discretionary traders specifically

A mechanical trader can backtest their rules and get a number. A discretionary trader cannot fully do that, because part of the system lives inside their head. That makes self-measurement your only real form of quality control. Without it, you are an inventor who never tests the prototype and simply hopes it flies.

Calibrated confidence gives you several practical benefits:

• Consistent execution: you take valid setups without hesitating.
• Sensible position sizing: you size up only where your edge is proven.
• Faster recovery: losing streaks are expected, not catastrophic.
• Honest improvement: you know which part of your process is failing.

How Do I Build Confidence in My Own Judgment When Developing a Discretionary Trading System?

The answer begins with a principle I call structured discretion. Pure gut-feel trading is not a system; it is a mood. Instead, you fix most of the process in writing and allow judgment only at clearly defined decision points.

Step 1: Write a playbook

Your playbook describes each setup you trade in plain language. Include the market context, the trigger, the entry, the stop, the target logic and, crucially, the conditions that cancel the trade. Keep it short enough that you can read it in two minutes before a session.

For example, a playbook entry might read: “EUR/USD, H4 uptrend, pullback to prior breakout level, bullish rejection candle on H1. Stop below the rejection low. Cancel if a high-impact USD release is due within two hours.”

Step 2: Define where discretion is allowed

Now mark the exact points where your judgment adds value. Typical candidates include:

1. Assessing the quality of the trend (clean or choppy).
2. Judging whether a level is significant enough to respect.
3. Deciding whether the market context, such as session or news, favours the trade.

Everything else, especially stop placement and risk per trade, should be fixed. Discretion over risk is where accounts go to die.

Step 3: Grade every setup before entry

Give every potential trade an A, B or C grade based on your discretionary checkpoints. Then, and this is the important bit, write down a probability estimate. “A-grade, 65% chance of reaching 1R before stop.” This forces your instincts to make a testable prediction rather than a vague hunch.

The Decision Journal: Measuring Your Judgment

A trading journal records what happened. A decision journal records what you thought would happen, and why, before you knew the result. That before-and-after comparison is the engine of genuine trader confidence.

What to record

• Date, pair, timeframe and setup name from your playbook.
• Your grade (A, B or C) and probability estimate.
• A one-line reason: “Clean trend, level tested three times, London open.”
• Whether you took the trade, and if not, why.
• The outcome in R-multiples (profit or loss divided by initial risk).

Record the setups you skip as well. Skipped trades are judgment calls too, and you need to know whether your “no” decisions are any good.

A worked example of calibration

Imagine a student, let us call him Daniel, who logs 90 setups over three months. His results look like this:

• A-grade (30 setups, estimated 65%): 19 winners, a 63% hit rate.
• B-grade (35 setups, estimated 55%): 17 winners, a 49% hit rate.
• C-grade (25 setups, estimated 45%): 13 winners, a 52% hit rate.

What does this tell Daniel? His A-grade judgment is well calibrated, and he can trust it. However, his ability to tell a B from a C is essentially noise; the C trades actually did slightly better. That is not a disaster. It is extremely useful information. He should study what separates A setups and stop pretending his B-versus-C judgment adds value.

This is the scientific method applied to your own brain. Forecasting researchers such as Philip Tetlock found that people who score their predictions over time become measurably better forecasters. Traders are no exception.

Separating Good Decisions from Good Outcomes

Here is where many intermediate traders trip over their own shoelaces. They judge a decision by its result. A rule-breaking trade that wins feels clever. A textbook trade that loses feels stupid. Both feelings are misleading.

The arithmetic of losing streaks

Suppose your system wins 55% of the time. That sounds respectable. But over 100 trades, the chance of hitting at least one run of five consecutive losses is roughly two in three. In other words, a perfectly good system will probably hand you a nasty streak at some point.

If your confidence depends on recent results, that streak will convince you that your judgment is broken. You will start tinkering with a system that was working fine. I have watched students rebuild a perfectly sound strategy three times in a year for exactly this reason. It is rather like replacing a car engine because of one flat tyre.

Sample size matters

Small samples lie. Twenty trades tell you almost nothing about a 55% edge, because random variation easily swings the result by 15 to 20 percentage points either way. As a rough guide:

• Under 30 trades: treat results as anecdotes.
• 50 to 100 trades: patterns begin to emerge.
• 100+ trades per setup: you can start drawing reasonable conclusions.

Psychologists call the tendency to judge decisions by outcomes “outcome bias”, and it travels with its cousin, confirmation bias, where we notice evidence that supports what we already believe. Both erode honest self-assessment. Your decision journal is the antidote, because it scores the decision against your stated process, not against the market’s mood that day.

Testing Your Discretion Against a Mechanical Baseline

This is my favourite technique, and the one students find most revealing. It answers a blunt question: does my judgment actually improve on simply taking every signal?

How to build the baseline

Take the mechanical core of your setup, meaning the trigger, stop and target, and record every occurrence, whether you liked it or not. This becomes your “shadow system”. Then compare its results with the trades your discretion selected.

For example, over six months:

• Shadow system (all 140 signals): +0.12R average per trade.
• Discretionary selection (62 trades taken): +0.31R average per trade.

In this case, your filtering more than doubled the expectancy per trade. That is hard evidence that your judgment adds value, and you can lean on it with confidence. If the numbers had been reversed, you would have learnt something equally valuable, if less flattering.

Why this works

The baseline removes the excuse of “the market was difficult”. The market was equally difficult for the shadow system. Any difference between the two is attributable to you. That is about as close to a controlled experiment as a discretionary trader can get.

Critical Warnings: What to Be Careful Of

I would be failing as your lecturer if I let you leave without the safety briefing. Building trading judgment has some well-known traps:

1. Grading after the fact. If you assign a grade once the trade is open, or worse, closed, your data becomes worthless. Grade before entry, every time.
2. Moving the goalposts. Changing your playbook every fortnight resets your sample size to zero. Commit to a version for at least 50 to 100 trades.
3. Discretion over risk. Never let judgment widen a stop or increase size mid-trade. Confidence is earned on entries, not on hopeful stop adjustments.
4. Confusing a trending market with skill. A strong trending phase makes everyone look brilliant. Check how your judgment performs across different market regimes.
5. Ignoring costs. Spread, slippage and swap can quietly turn a small edge negative. Record results net of costs.
6. Overconfidence after a hot streak. Five wins in a row is just as statistically meaningless as five losses. Do not double your risk because you feel invincible.

Finally, remember the scale of the arena you are playing in. According to the Bank for International Settlements Triennial Survey, global FX turnover was about $7.5 trillion per day in 2022. Your edge, however real, is small relative to that ocean. Humility is part of good judgment.

Next Steps for Developing Your Judgment

Knowing the theory is pleasant. Applying it is what changes your results. Here is a practical sequence for the next three months:

1. Week 1: Write your playbook. Limit it to one or two setups.
2. Week 2: Define your discretionary checkpoints and grading criteria in writing.
3. Weeks 3 to 12: Log every setup, taken or skipped, with a grade and probability estimate. Trade small or on demo.
4. Monthly: Compare your grades against outcomes and your selection against the shadow system.
5. End of quarter: Keep the judgment calls that proved valuable. Mechanise or remove the ones that did not.

Questions to develop your thinking

• Which of my discretionary checkpoints could I describe precisely enough to mechanise? If you can, consider doing so. Discretion is most valuable where it is hardest to codify.
• Do my skipped trades perform better or worse than my taken trades? If skipped trades win more often, your filter is working backwards.
• Is my calibration consistent across sessions? Many traders judge well in London and poorly late in the New York session, often because of fatigue.

Frequently Asked Questions

How many trades do I need before I can trust my judgment?

Aim for at least 50 to 100 logged setups per playbook entry before drawing firm conclusions. Fewer than 30 trades is mostly noise, so treat early results as preliminary.

Can I build trading judgment on a demo account?

Yes, for measuring decision quality and calibration. However, real money adds emotional pressure, so move to small live positions once your demo data shows a consistent edge.

What is the difference between a trading journal and a decision journal?

A trading journal records outcomes. A decision journal records your predictions and reasoning before the outcome is known, so you can measure whether your judgment was accurate.

Should a discretionary trading system have any fixed rules?

Absolutely. Risk per trade, stop placement logic and cancellation conditions should be fixed. Discretion works best in narrow, defined areas such as assessing context and setup quality.

What if my results show my discretion adds no value?

That is useful, not shameful. Mechanise the setup, follow it consistently, and use your judgment to research new ideas rather than filter live trades.

Conclusion

So, how do I build confidence in my own judgment when developing a discretionary trading system? You treat your instincts as hypotheses, not truths. Write a playbook, limit discretion to defined checkpoints, and grade every setup with a probability before entry. Then measure your calibration with a decision journal, judge decisions by process rather than outcome, and compare your selections against a mechanical shadow system.

Confidence built this way survives losing streaks, because it rests on evidence rather than emotion. Start your decision journal this week, log every setup for the next three months, and let the data tell you which parts of your judgment have earned your trust.