Candlestick patterns and Elliott waves: shape, condition and structure
Pin bars, engulfing and inside bars under explicit OHLC conventions, with separate wave and confirmation checks.
Published: (UTC+3)
Русский · English · Deutsch · Español · Français · Italiano · 繁體中文 · 日本語 · 한국어
A candle records four prices for an interval; a candlestick pattern defines relationships between one or more candles. Our invented inputs separate these observations from a complete Elliott wave hypothesis.
A candle shape does not replace wave evidence
For a completed candle, body B=|C−O|, range D=H−L, lower shadow=min(O,C)−L and upper shadow=H−max(O,C). Before classifying it, check L≤O,C≤H, one source and one interval. Unfinished H,L,C can change; even completed OHLC does not disclose intrabar order.
Neither a pattern name nor a later chosen condition establishes the entire wave. For example, a study can specify a subsequent close above 106 after the pin bar below. That is a separate later-bar test, not a universal requirement or reversal promise. It cannot restore an original wave hypothesis whose structural condition has already failed.
| Item | Calculation or value |
|---|---|
| P | B / W− / W+ |
| E | [min(O,C),max(O,C)] |
| I | H2<H1; L2>L1 |
Does naming a candle validate the wave?
No. Check the chosen shape first, then its context, new observations and the wave’s structural requirements separately.
A pin bar under an explicit convention
For this exercise, choose a lower-tail form with B>0, lower shadow≥2B, upper shadow≤B and close in the upper third of D. O104,H106,L96,C105 gives body 1, range 10, lower shadow 8 and upper shadow 1; close 105 lies in the upper third. This is our measurable pin-bar convention, not a universal rule or a claim that every such shape is a hammer.
The hammer documentation attaches similar geometry to preceding-downtrend context and configurable factors; names and filters from different methods should stay distinct. In a separate hypothesis, this candle belongs to proposed wave two of an ordinary upward impulse from I100. Low 96 breaches origin 100, rejecting that hypothesis regardless of shape or close 105. A similar tail establishes no reversal.
| Item | Calculation or value |
|---|---|
| O/H/L/C | 104/106/96/105 |
| B/W−/W+ | 1/8/1 |
| C-L / H-L | 9/10 |
| I / L | 100 / 96 |
Does close 105 rescue wave two from low 96?
No. The relevant low matters in this hypothesis, not only the close.96 below origin 100 violates the condition being tested.
Body engulfing does not require engulfing every wick
Our bullish case uses a down first candle, an up second candle, strict O2<C1 and C2>O1. Bar 1 O105,H109,L97,C100 has body[100,105]. Bar 2 O99,H107,L98,C106 has body[99,106], covering the first body. Its full range[98,107] nevertheless lies inside the former range[97,109].
Body engulfing can therefore coexist here with an inside bar by H/L: the classifications measure different relationships. Reverse body directions and inequalities for a bearish counterpart, and examine context separately. TradingView also distinguishes body coverage from full-range coverage. Established geometry alone proves neither reversal nor a completed corrective wave.
| Item | Calculation or value |
|---|---|
| 1:O/H/L/C | 105/109/97/100 |
| 2:O/H/L/C | 99/107/98/106 |
| [min(O,C),max(O,C)] | [100,105] ⊂ [99,106] |
| [L,H] | [98,107] ⊂ [97,109] |
New high 107 is below old high 109. Does that cancel body engulfing?
No. The tested bodies are[100,105] and[99,106]. Full wick coverage is not part of this condition.
Compare the entire range for an inside bar
Choose strict Hchild<Hmother and Lchild>Lmother for this exercise; exclude equality. Mother O100,H110,L90,C104 and child O103,H108,L94,C98 pass 108<110 and 94>90. Body color and direction do not independently determine range containment.
Containment does not establish the direction of the next break. Other methods may handle equality differently, so state the convention. Even a later break does not establish a trade fill or the whole wave. Record any subsequent observation condition, its availability time and a possible failure outcome beforehand.
| Item | Calculation or value |
|---|---|
| 1:O/H/L/C | 100/110/90/104 |
| 2:O/H/L/C | 103/108/94/98 |
| H2 / H1 | 108<110 |
| L2 / L1 | 94>90 |
Would a child high exactly 110 pass this version?
No. Equality is excluded under our strict convention. That follows from the selected condition, not every definition of an inside bar.
Price action: preserve facts before interpretation
Record available OHLC, swings and a preselected zone separately in the teaching journal. Then classify the candle through explicit inequalities and propose a wave role with internal requirements and an invalidation condition. Use one source, scale and cutoff; the adjacent levels article helps preserve initial boundaries after the outcome.
Only after a new completed bar should the test result be appended. Keep failed and unresolved cases; an observation is not a market fill. A trading study would separately need order rules and costs. This process itself measures no profitability or probability and claims no automatic-trading product feature.
Can the zone be replaced after the outcome while retaining a successful old entry?
No. Preserve the original conditions and outcome; a new zone or wave role creates a revised hypothesis.