A trader spots a pullback that just looks right and takes it without being able to say exactly why, beyond a feeling built from thousands of hours of screen time. It works, three times in a row. The fourth time it does not, and afterward the trader cannot explain what was different, because nothing about the process was ever written down in the first place. That gap, between a real, working instinct and a rule someone else, or a future version of yourself, could actually follow, is what turns a discretionary read into a repeatable system.
Where pure discretion breaks down
Discretion is not fake. Experienced traders genuinely see things a checklist misses, a slightly different candle shape, a subtle change in how fast price is moving, a level's history that a rule cannot fully capture. The problem is consistency: a purely discretionary trader makes a slightly different judgment call under stress than they do when calm, takes the same setup differently on a losing day than a winning one, and cannot easily tell, months later, whether a stretch of results reflects a real edge or a run of favorable moods. Discretion without any written structure is real skill with no way to audit it.
Turning judgment calls into written rules
The fix is not eliminating judgment, it is writing down the parts of the judgment that repeat. Take the vague looks right pullback: watching enough examples usually reveals that the version that works has a few consistent features, perhaps a pullback that retraces less than half the prior swing, a candle that closes in the top third of its range, and a location at or near a level with real history. Writing those three conditions down turns an unexplainable feeling into a rule that can be tested, taught, and checked against, even if it started as pure intuition.
What should stay discretionary
Not everything should become a rule, and forcing it usually makes the system worse, not better. Broad context, whether today is a trend day or a range day, whether a news event changes the read for the next hour, whether overall volatility argues for smaller size than usual, tends to resist rigid rules and rewards experienced judgment instead. The workable split: rules govern the mechanical parts, entry trigger, stop placement, target logic, position size, while judgment governs the context those rules get applied within. A system with no discretion left in it usually just moves the failure point from bad judgment to a rulebook too rigid for markets that never repeat exactly.
A worked conversion
Take the earlier pullback idea and finish converting it. Rule: only take the pullback if it retraces less than 50 percent of the prior swing and holds above the 20 EMA on the working timeframe. Trigger: a bullish candle closing in the top third of its range on the pullback low. On EURUSD, say the prior swing ran from 1.0790 to 1.0850, and the pullback holds at 1.0825, a 42 percent retracement, with a strong closing candle back above 1.0830. Entry: buy on that candle's close. Stop: below the pullback low at 1.0818, a 12 pip risk. Target: the prior swing high at 1.0850, plus a trailed extension if momentum continues. Invalidation: a close back below 1.0818, or a retracement deeper than 50 percent before the trigger candle even appears, which means the setup never qualified in the first place.