Home › Comparisons

Algorithmic vs discretionary trading: code or judgment?

Some traders write rules into code and let a computer place trades. Others look at the chart and decide. Most good trading sits somewhere in between: written rules, followed by a person.

Comparison · 7-minute read · Updated 2026-10-21

The short answer

Algorithmic trading means every rule is written in code. The computer checks the market and places the orders. You build and test the rules; the machine follows them.

Discretionary trading means a person makes each decision. You may have a plan, but you judge each chart in the moment.

There's also a middle path, often called systematic or rules-based trading. The rules are written down in plain words, and you follow them by hand. Most beginners start here.

Side by side

How algorithmic trading works

You write rules like: "If price closes above the last turning point, buy. Stop below the prior low. Risk 1%." The code runs on every candle. When the rule fires, it sends the order.

On TradingView, you can test rules like this with a Pine Script strategy and the Strategy Tester. See Pine Script strategy backtest.

How discretionary trading works

You look at the chart and ask questions. Is the trend clear? Is this level clean? Is there news in an hour? Then you decide.

Good discretionary traders still have rules. They just leave room for judgment, like skipping a setup on a messy day.

Algorithmic: strengths and limits

The good parts:

• No hesitation. The code doesn't get scared or greedy.
• Testable. You can run the rules on years of past data.
• Fast and tireless. It can watch many markets at once.

The limits:

• Overfitting. It's easy to tweak rules until they fit the past perfectly. Then they often fail on new data.
• Tech risk. Bugs, bad data, lost connections and broker outages can all cause losses.
• Test results are hypothetical. They leave out real fills, slippage and changing markets.
• Still needs a person. Someone has to watch it, fix it and decide when to stop it.

Discretionary: strengths and limits

The good parts:

• Adapts. You can skip a trade when the market looks odd.
• No code needed. You can start with a chart and a pen.
• Sees context. News, holidays and odd days are easy for a person to notice.

The limits:

• Feelings sneak in. Fear, boredom and the urge to win it back affect choices. See trading psychology for beginners.
• Hard to measure. If every trade is a judgment call, it's hard to tell what's working.
• Inconsistent. The same setup may get different treatment on different days.

The middle ground most people need

Write your rules down first, in plain words:

• When do I look for a trade? (For us: trend on the bigger chart, a level where price turned before.)
• What confirms it? (A candle body closing past the level, not just a tail.)
• Where is the stop? (Past that level.)
• How big? (Risk about 1%, size from the stop with the position size calculator.)
• When do I stop for the day? (After 2 losses.)

Then follow them by hand, on past charts first. See how to write a trading plan and pre-trade checklist.

Once the rules are clear and tested, coding them is a natural next step if you're curious. Our course includes the full source code of the StructureX chart tool to learn from. It's source code, not a ready-made trading bot.

Which fits you?

1. New? Start rules-based by hand. You learn why each rule exists.
2. Enjoy coding and testing? Algorithmic trading can fit, as long as you test on data the rules have never seen.
3. Experienced, with a long journal? More judgment can fit, because your records show where it helps.

Common mistakes

• Trusting a perfect backtest. If it never loses, it's probably fit to the past.
• Overriding your own rules mid-trade. That turns a rules-based plan into guessing.
• Buying a bot you can't see inside. If you can't read the rules, you can't judge the risk.
• Skipping the forward test. Run the rules on live paper trades before real money.

Questions