Reading candles without the cargo-cult patterns.
Spend an hour on TikTok and you'll come away with a list of about thirty named candle formations and the strong impression that finding one is the same as having a trade. Both halves of that impression are wrong, in the same way. Candles encode something real — four numbers per bar, the open, high, low, and close, compressed into a glyph. The encoding is genuine. What's folklore is the idea that the glyph alone predicts the next bar. Patterns are the alphabet. Context is the sentence. Most traders memorize the alphabet and never learn to read.
What a candle actually encodes
A candle is four numbers and a color. The body is the rectangle drawn between the open and the close. If close is above open, the candle is bullish — drawn green or hollow. If close is below open, it's bearish — drawn red or filled. The wicks (sometimes called shadows or tails) are the thin lines extending above and below the body. The upper wick reaches to the bar's high; the lower wick reaches to the low. Together: open, high, low, close. Four numbers. One glyph.
What the glyph carries is an honest summary of the bar's price action. Where did the bar open. Where did it close. How far up did it travel before sellers showed. How far down before buyers showed. The body magnitude is conviction — a long body is a decisive directional close, a short body is indecision. The wicks are rejection — territory the bar visited but couldn't hold. The wick is the more interesting half of the candle, because it shows where the market failed.
The body, the wicks, the lie of "patterns"
Most retail candle education focuses on the names. Hammer. Shooting star. Engulfing. Three white soldiers. Evening star. Doji. The names are real — they describe specific shapes — but the focus on names hides the more useful frame: read the components. A "hammer" is just a small body high in the day's range with a long lower wick. The name is shorthand for "buyers absorbed a large attempt to push price down and closed near the high." The shape carries the same information whether you call it a hammer or not. The name is a label. The shape is the data.
Once you read the components instead of the names, you stop needing to memorize a glossary. You see a candle with a tiny body and a long lower wick and you know what happened in the bar. You see a long bearish body with no upper wick and you know sellers were in control from the open. You see a doji with balanced wicks and you know the day was a draw. The reading is forensic, not pattern-matching.
Volume changes the meaning
The candle is one half of the bar's data. Volume is the other half. The same candle on heavy volume vs. thin volume tells two different stories. A long bullish marubozu on three-times average volume is institutional conviction — large players agreed and pushed price all session. The same candle on a third of average volume is a half-empty room. The price moved because almost nobody was selling, not because anyone particularly wanted to buy.
Volume is what separates a candle that matters from a candle that happened. The framework treats volume as a first-class input alongside the candle structure. Pillar checks for entries weight high-volume confirmation candles materially more than low-volume ones with identical shape. Same glyph, different signal, because volume is the audience reaction to the price action and the audience reaction is most of the information.
The cargo-cult patterns
The retail candlestick canon is roughly thirty named patterns, most of them multi-bar formations: three white soldiers, three black crows, evening star, morning star, dark cloud cover, piercing pattern, harami, abandoned baby, tweezer top, tweezer bottom. They sound technical because they have specific definitions and required bar counts. They feel powerful because they're rare, and rare-feeling things feel high-signal.
They mostly don't work. Specifically: most multi-bar exotic patterns show negligible standalone edge in honest backtesting. The pattern's predictive value either evaporates entirely or shrinks to within noise once you control for the obvious confounders — overall trend direction, sector regime, volume confirmation, distance from structural levels. The pattern looks predictive in your eye because the brain rewards pattern-matching. The math doesn't agree.
This is the cargo-cult problem. Cargo cults observed planes landing during WWII and built bamboo runways to summon more planes. The runway looked like the cause of the plane. It wasn't — the plane was caused by the war machine that built the original airstrip. Most candle patterns are bamboo runways. They co-occur with directional moves often enough that the brain notices, but they don't cause the move; they're a downstream artifact of the underlying flow.
The patterns that earn their keep (in context)
A short list of single-bar and two-bar reads that hold up — but only with confluence:
- Hammer at structural support, on volume. The shape (small body high in range, long lower wick) describes absorption: sellers tried, buyers won the close. At a confluence support level — a pivot, a moving average, a volume node — the shape carries genuine signal. Mid-range, it's noise.
- Shooting star at structural resistance, on volume. The mirror image. Buyers tried, sellers won the close. At resistance with confluence: meaningful. Mid-range: noise.
- Bullish engulfing after a confirmed downtrend, on volume. Two-bar pattern where today's bullish body engulfs yesterday's bearish body. The "after a confirmed downtrend" qualifier is doing most of the work. The pattern in chop is meaningless.
- Bearish engulfing after a confirmed uptrend, on volume. Same logic, mirrored.
- Doji at confluence levels, especially after extended trends. The doji is indecision. At a level the prior trend was supposed to break through, indecision is information. Mid-range, indecision is just chop.
Notice the pattern: every working pattern has a where qualifier. The shape is necessary; the location does most of the work. Patterns at random places on the chart are folklore. Patterns at confluence levels are signal.
Pattern survival table
| Pattern | What retail says | What honest backtests show |
|---|---|---|
| Hammer at support | "Reversal signal" | Modest edge with confluence + volume. Mid-range: noise. |
| Shooting star at resistance | "Reversal signal" | Same. Edge depends on the level, not the shape alone. |
| Bullish engulfing | "Trend reversal" | Edge only with prior trend confirmation + volume. Otherwise too common to be informative. |
| Doji at confluence | "Trend exhaustion" | Mild edge as a marker of indecision. The doji doesn't tell you the direction, just that consensus broke. |
| Three white soldiers | "Powerful continuation" | Negligible standalone edge. By the third bar, the move is mostly priced in. |
| Evening star | "Bearish reversal" | Mild edge at major resistance, negligible elsewhere. The story is the resistance, not the three-bar shape. |
| Harami | "Trend pause" | Almost no measurable edge. The pattern is too common; "small bar inside previous bar" describes a third of all bars. |
| Three black crows | "Strong continuation down" | Same as three white soldiers, mirrored. By bar three, the move is mostly behind you. |
The honest reading: the table is mostly "the pattern works only when paired with context that's already doing most of the predictive work." Which is another way of saying the pattern isn't predictive. The level is.
Context, not patterns
Here's the lesson the cargo cult misses. The candle is a clue. The context is the answer. A hammer at confluence support after four weeks of downtrend with a volume spike is a real signal. A hammer mid-range in a chop is noise. Same candle, same shape, opposite signal value, because the context determines whether the absorption story it's telling is meaningful.
The discipline is to read the candle and the location together. If your eye finds a hammer, the next question isn't "is it a hammer." The next question is "where on the chart did it print." If the answer is "on confluence support after a measured pullback," the candle has earned the right to be considered. If the answer is "in the middle of a four-week range," the candle is a glyph and nothing more.
What the framework does with candles
The Trap & Structure Coach reads candle structure but only at confluence levels. The dashboard does not fire on patterns alone — it fires on patterns that print at structural levels with volume confirmation. The 10-source S/R confluence merging (introduced in lesson 4) is the gate; the candle read is what fires through the gate. No confluence, no fire, regardless of how textbook the candle looks.
This is why the framework can show "no setups today" on a chart that's full of beautiful candle shapes. The shapes aren't enough. Most days, the candles print between the levels, not at them. Most days, the right move is to wait.
The real lesson
Patterns are the alphabet. Context is the sentence. Most traders memorize the alphabet — the names, the shapes, the multi-bar exotics — and never learn to read. The reader of the chart is asking three questions every bar: where did it open, where did it close, what got rejected. That's three numbers and two wick lengths. You don't need a glossary. You need a structural framework that tells you which bars to pay attention to in the first place. Patterns are downstream. Levels are upstream. The bar is the page; the level is the question; the candle is whether the answer arrived.
Related: Lesson 4 — R:R isn't a number (where the levels come from) · Lesson 8 — HH/HL trend structure · Trap & Structure Coach (docs)