Chart patterns (flags, triangles, head and shoulders, double tops/bottoms)

Chart patterns are recurring shapes that show a particular balance between buyers and sellers. They're popular because they're visual and easy to spot — which is also their weakness: almost any chart looks like a pattern if you want it to. This lesson treats patterns the professional way: as specific structures with clear rules for confirmation, entry, stop, and target, which you can then test rather than trust.

What you'll learn

• The difference between continuation and reversal patterns
• Flags and pennants, triangles, head and shoulders, and double tops/bottoms
• Where to enter, where to place the stop, and how to project a measured-move target
• Why confirmation matters — and how failed patterns can become signals themselves
• How to avoid seeing patterns everywhere

1. Continuation vs reversal

A pattern only makes sense in context. A bull flag needs a prior uptrend; a head and shoulders top needs something to reverse.

2. Flags and pennants (continuation)

Structure: a sharp move (the pole) followed by a short, shallow pullback that drifts slightly against the trend (a flag, like a small channel) or tightens into a small triangle (a pennant).

Why it forms: early buyers take some profit, but sellers can't push price back far before new buyers step in.

Rules

• Entry: on a close beyond the flag's boundary in the trend direction
• Stop: beyond the opposite side of the flag
• Measured-move target: the length of the pole, projected from the breakout point

Worked example

(Illustrative.) A pole runs from 1.0800 to 1.0900 (100 pips). A flag forms, with its low at 1.0870. Price closes above the flag at 1.0885.

• Stop: 1.0865, below the flag low → 20 pips risk
• Measured target: 1.0885 + 100 = 1.0985 → 100 pips
• A realistic plan would also check for resistance before 1.0985 and might target the nearest level instead.

3. Triangles

Rules

• Entry: on a close outside the triangle
• Stop: inside the triangle — beyond the last swing on the opposite side
• Measured-move target: the height of the triangle at its widest point, projected from the breakout

4. Head and shoulders (reversal)

Structure (top): a high (left shoulder), a higher high (head), a lower high (right shoulder), with the lows between them forming the neckline. The inverse pattern at a bottom works the same way, upside down.

Why it forms: the uptrend makes one last push (the head), fails to make a higher high on the next attempt (the right shoulder — a lower high), and then breaks the neckline — which is also a break of structure.

Rules

• Entry: on a close below the neckline, or on a retest of the neckline from below
• Stop: above the right shoulder
• Measured-move target: the distance from the head to the neckline, projected down from the neckline

Worked example

(Illustrative.) Head at 1.1000, neckline at 1.0900, right shoulder high at 1.0955.

• Pattern height: 1.1000 − 1.0900 = 100 pips → target 1.0800
• Breakout entry at 1.0895, stop 1.0960 → 65 pips risk, 95 pips reward (≈ 1.5 : 1)
• Retest entry at 1.0900 after price rallies back to the neckline and rejects it, stop 1.0960 → 60 pips risk, 100 pips reward (≈ 1.7 : 1)

The retest gives a better ratio but may never come — you trade off reward-to-risk against the chance of missing the move. That's a decision to make in your rules, not in the moment.

5. Double tops and double bottoms (reversal)

Structure (top): two highs at roughly the same level, separated by a pullback low. The pattern confirms when price closes below that pullback low.

Rules

• Entry: on a close below the confirmation level (or on its retest)
• Stop: above the two tops
• Measured-move target: the height from the tops to the confirmation level, projected down

A double top that never breaks its confirmation level isn't a double top — it's just a range.

6. Confirmation and failed patterns

The most common pattern mistake is acting before confirmation — selling the second top of a "double top" before the confirmation level breaks, or shorting a "head and shoulders" before the neckline goes.

Patterns also fail. When a clearly confirmed pattern breaks out and then quickly reverses back through its breakout level, the trapped traders on the wrong side can fuel a strong move the other way. Some traders build strategies specifically around these failed patterns.

7. Seeing patterns that aren't there

Humans are excellent at finding shapes in noise. To stay honest:

• Define each pattern precisely in writing (for example: "second top within 0.3 × ATR of the first").
• Require the context — prior trend for continuation patterns, an extended move for reversals.
• Require confirmation by a candle close.
• Test it. In Module 3.4 you'll backtest your pattern rules on historical data — the only real way to know if a pattern works in your market and timeframe.

Common beginner mistakes

• Trading before confirmation.
• Forcing patterns onto charts that don't match the definition.
• Using measured targets blindly without checking for support or resistance in the way.
• Ignoring the context — continuation patterns without a trend, reversal patterns without anything to reverse.
• Assuming a pattern "always works" without testing it.

Key terms

Practice

1. Write a precise, one-paragraph definition for two patterns you want to trade, including context, confirmation, entry, stop, and target.
2. On H4 charts of your market, find five examples of each from the past year that meet your definition exactly.
3. For each, record: did it confirm? Did it reach the measured target before the stop? What was the reward-to-risk?
4. Find one failed pattern and describe what happened after the failure.

Quick recap

• Continuation patterns (flags, pennants, triangles) pause a trend; reversal patterns (head and shoulders, double tops/bottoms) signal a possible turn.
• Every pattern needs context, confirmation, an entry, a stop, and a target.
• Measured moves give a target — but check for levels in the way.
• Failed patterns can be powerful signals in the opposite direction.
• Define patterns precisely and test them before trusting them.