ELLIOTT WAVE STUDY
Wave patterns: structure and differences.
Identify a pattern by its role in the larger wave and the structure of its components. A similar outline is insufficient: establish degree and position before testing subdivisions and price constraints.
A map of the main families.
The numbers 5 and 3 below describe internal structures, not any five or three visible fluctuations. A “three” can be a complex correction. Reverse price directions for a downward motive sequence; structural roles remain the same.
| Family | Labels and subdivisions |
|---|---|
| Impulse | 1–2–3–4–5; 5–3–5–3–5 |
| Diagonal | 1–2–3–4–5; ending 3–3–3–3–3; leading needs a separate qualification |
| Zigzag | A–B–C; 5–3–5 |
| Flat | A–B–C; 3–3–5 |
| Triangle | A–B–C–D–E; 3–3–3–3–3 |
| Combination | W–X–Y; W and Y are complete corrections, X is a corrective connector |
Impulse, extension and diagonals.
In an ordinary impulse, wave 2 does not fully retrace 1; wave 3 passes the end of 1 and is never the shortest of 1, 3 and 5; wave 4 stays outside wave 1 territory. Wave 3 is itself an impulse. Check bar extremes and subdivisions rather than just the endpoints of five lines.
An extension develops the smaller structure of a motive component in greater detail. Do not automatically label additional turns at the same degree. A truncated fifth ends short of the third’s endpoint: establish that its own components are complete instead of mistaking the end of available history for pattern completion.
An ending diagonal occupies a final 5 or C position, normally subdivides 3–3–3–3–3 and overlaps waves 1 and 4. A leading diagonal belongs in 1 or A of a zigzag. The selected source discusses both 3–3–3–3–3 and 5–3–5–3–5 for leading diagonals; this lesson does not assert one universal internal variant. A wedge alone proves nothing: record the definition and the components actually confirmed.
Zigzag and three kinds of flat.
In a zigzag, A and C are motive fives and B is corrective; B does not exceed the start of A. In a flat, A and B are corrective and C is a motive five. Identical A–B–C labels therefore do not identify the same pattern.
Each invented downward correction below starts at 130 and ends A at 120. These endpoints illustrate shape, not proof of 5–3–5 or 3–3–5. A regular flat brings B near A’s start; an expanded flat carries B beyond that start and C beyond A’s end; in a rare running flat, B exceeds A’s start while C falls short of A’s end. Do not choose a rare variant merely to rescue a failed count.
| Candidate and internal structure | Invented B and C endpoints |
|---|---|
| Zigzag · 5–3–5 | B=126; C=114 |
| Regular flat · 3–3–5 | B=130; C=119 |
| Expanded flat · 3–3–5 | B=134; C=116 |
| Running flat · 3–3–5 | B=134; C=123 |
The five parts of a triangle.
A triangle is an A–B–C–D–E correction with five corrective components. It differs from both a five-part impulse and a diagonal: labels and larger-wave position matter. It normally precedes the last actionary component of the larger pattern, for example as 4 or B; it can also finish a corrective combination. Do not assign a standalone triangle to the whole of wave 2 of an ordinary impulse.
In a contracting triangle, A–C and B–D boundaries converge; a barrier triangle has one nearly horizontal boundary; expanding boundaries diverge. In a running variant, B may exceed A’s start. E crossing the A–C trendline does not by itself invalidate every triangle rule. Wait for E to develop and inspect its smaller components before declaring completion.
W–X–Y: two corrections with a connector.
In a double zigzag, W and Y are zigzags, joined by a corrective X. In a sideways double three, W and Y can be different permitted corrections, for example a flat followed by a triangle. In that combination the triangle finishes as Y rather than beginning as W.
W and Y name whole patterns, not replacements for A and C in any three-part move. Their internal labels belong to a smaller degree. A double zigzag typically extends the price correction, whereas a sideways combination extends its duration. More complex W–X–Y–X–Z notation exists in classifications, but should not become the default explanation for an unclear segment.
A checking sequence and exercise.
Establish context
Record the parent wave, degree, bar source and fixed boundaries.
Examine components
Inspect A and C or 1, 3 and 5 at a smaller degree. Mark unresolved parts as unconfirmed.
Apply the constraints
Name the specific model and the condition that would make you reject it.
Keep an alternative
Compare interpretations on the same data; avoid moving boundaries to force a preferred answer.
After a rise from 100 to 130, the endpoints are A=120, B=134, C=116. Is that enough to confirm an expanded flat?
Show the explanation
No. The endpoints fit its outline: B exceeds 130 and C falls below 120. You must still establish corrective A and B and motive C, giving 3–3–5, as well as degree and parent position. If A clearly subdivides into five, reconsider this particular hypothesis.
Theory references
Connect the pattern to the chart
Follow the step-by-step BTCUSDT study to separate price movement from a wave interpretation. The internal structure is only partly established on these daily bars.
Ending diagonal example: check the inside and the boundaries
This is an original, invented rising ending-diagonal exercise, not a historical chart or a current forecast. Assume it is a candidate for a larger wave 5 or terminal C; that surrounding context is not supplied by the numbers. An ending diagonal and a leading diagonal have different positions. EWI’s Diagonals describes the ending form with five three-part legs and a normally converging wedge. It says wave 4 almost always overlaps wave 1; our chosen example explicitly does.
A reproducible 3-3-3-3-3 outline
The horizontal coordinate t is an invented, evenly spaced teaching index, not a date or a bar from a price feed. The initial point is (t = 0, price = 100). Each row gives its start and three internal a–b–c endpoints; consecutive rows share their boundary. The whole outline has 15 internal legs, not five unexplained straight strokes.
| Outer leg | Times: start, a, b, c | Prices: start → a → b → c | Net point length |
|---|---|---|---|
| 1 | 0, 1, 2, 3 | 100 → 114 → 108 → 120 | 20 up |
| 2 | 3, 4, 5, 6 | 120 → 114 → 118 → 110 | 10 down |
| 3 | 6, 7, 8, 9 | 110 → 120 → 115 → 126 | 16 up |
| 4 | 9, 10, 11, 12 | 126 → 120 → 123 → 117 | 9 down |
| 5 | 12, 13, 14, 15 | 117 → 126 → 122 → 132 | 15 up |
Wave 2 ends at110 >100; wave 4 at117 >110, so neither retraces all of the preceding advancing wave. Wave 3 ends at126 >120. Actionary lengths are20,16,15: the third is not shortest. The same ordering holds in percentages:20%, approximately14.55%,12.82%. Wave 4’s117 lies inside wave 1’s100–120 territory, while the corrective lengths shrink10→9. The price advances20→16→15 contract in this particular construction.
Check the wedge on the stated linear scale
The upper boundary through outer endpoints1 and3 is U(t) = 117 + t. The lower boundary through endpoints2 and4 is D(t) = 103 + (7/6)t. Their slopes1 and7/6 differ. The width U − D = 14 − t/6 narrows from13 at t=6 to11.5 at t=15; their intersection is at t=84, after the displayed structure. Endpoint5 is132=U(15). These are converging, nonparallel lines on the linear coordinates used here. A logarithmic chart changes the geometry and must be checked separately; this example makes no semilog-wedge claim.
Do the three-part legs have smaller structure?
To make the example reproducible at one further degree, construct each a and c leg from its own start S to end E with normalized price fractions 0,0.30,0.20,0.70,0.60,1; construct each b leg with 0,0.70,0.30,1. A fraction f means S+f(E−S), including for descending legs. Space those points evenly within that leg’s time interval. Thus each row’s a–b–c has a finite5-3-5 sketch, and all five rows together contain65 smaller legs.
For example, row1’s a,100→114, becomes100→104.2→102.8→109.8→108.4→114. Its motive lengths4.2,7,5.6 make the middle leg longest; the fourth108.4 does not overlap the first104.2. The same normalized construction works after reversing direction. Each b is a three-leg correction; its smaller internal structure is deliberately unresolved. No finite table proves every deeper degree, market OHLC ordering, volume or larger context.
Completion, alternatives and rejection
At a teaching cutoff t=12, only waves1–4 are supplied; wave5 and any finished diagonal are still developing. If points13–15 are later added as invented exercise data, the outline is complete at the stated depth, but a market reversal is still not an observed fact. For this exercise, a later completed move below D(t) is a follow-up review condition chosen by the author; it is not an EWI rule that guarantees a reversal.
Reject this specific contracting example if the supplied internal legs no longer support the declared subdivisions, if wave2 fully retraces wave1, if wave4 fully retraces wave3, if wave3 is shortest, or if the two tested boundaries are parallel or diverge. A different structure needs a new case; do not relabel a rising channel as this wedge. A wedge shape alone cannot establish a completed diagonal. Throw-over and rare slight truncation are source-discussed variations, so reaching or missing the drawn line alone does not settle completion. Use the rule checklist, compare an alternative interpretation, and study the conditional fifth-wave calculation without turning its projections into a promise.
Study offline with the original PDF guide
Rules, counting, corrections and Fibonacci, with invented worked examples and an answer key. Available in this page language.