Chart patterns and Elliott: five testable outlines
Head and shoulders, double extrema, flags, pennants, wedges and cups with handles: original numbers, cutoffs and separate test branches.
Published: (UTC+3)
Русский · English · Deutsch · Español · Français · Italiano · 繁體中文 · 日本語 · 한국어
All prices and outlines are invented. t is a diagram-point index, not a bar or date; outer turns are supplied hypotheses, while full OHLC, volume and pivot-availability timing are unestablished. Each plot stops at its cutoff; later closes are revealed in separate experiments. Documentation describes pattern classes; our thresholds and datasets do not replicate an indicator algorithm.
Head and shoulders: outline and neckline
In 100→110→104→116→104→111, shoulders 110/111 lie below head 116; shoulders need not be equal. Intermediate lows 104 define horizontal neckline N=104. This is a teaching top hypothesis after an advance, not an established reversal. At cutoff t 5 the right shoulder remains provisional and future bars are not plotted.
Preselect a later completed-close test C<104. In the next experiment C103 passes; an independent failure branch C117>116 contradicts the assumed completed top with head 116. Neither close was knowable at t 5. The outline and test cannot assign Elliott numbers: internal structure, degree and context remain necessary, and no fill price is established.
| Outline / branch | Point / stage | Price | Condition / status |
|---|---|---|---|
| Head and shoulders | t 0 | 100 | Supplied conditional point |
| Head and shoulders | t 1 | 110 | Supplied conditional point |
| Head and shoulders | t 2 | 104 | Supplied conditional point |
| Head and shoulders | t 3 | 116 | Supplied conditional point |
| Head and shoulders | t 4 | 104 | Supplied conditional point |
| Head and shoulders | t 5 | 111 | Supplied conditional point |
| Head and shoulders · test | After cutoff, separately | 103 | C<104 |
| Head and shoulders · failure branch | After cutoff, separately | 117 | C>116 |
Does seeing right shoulder 111 establish a neckline break?
No. At t 5 there is no subsequent completed close below 104. Even that later test would not validate the entire Elliott count.
Double top and mirrored double bottom
Use 100→110→104→109.8 for a top and 110→100→106→100.2 for its mirrored bottom. The paired highs or lows differ by 0.2, within a preselected tolerance 0.3. This is our tolerance, not a market rule. At t 3 the intervening boundaries are 104 for the top and 106 for the bottom; near-equal extrema alone establish no reversal.
After the cutoff, test strict C<104 for the top or C>106 for the bottom: separate C103 and C107 meet their conditions. Independent failures C111 above 110 or C99 below 100 reject the selected completed-pair assumption, not every later bearish or bullish model. Similar lows do not automatically identify wave two. Preserve full bars, pivot availability and the previous hypothesis version.
| Outline / branch | Point / stage | Price | Condition / status |
|---|---|---|---|
| Double top | t 0 | 100 | Supplied conditional point |
| Double top | t 1 | 110 | Supplied conditional point |
| Double top | t 2 | 104 | Supplied conditional point |
| Double top | t 3 | 109.8 | Supplied conditional point |
| Double top · test | After cutoff, separately | 103 | C<104 |
| Double top · failure branch | After cutoff, separately | 111 | C>110 |
| Double bottom | t 0 | 110 | Supplied conditional point |
| Double bottom | t 1 | 100 | Supplied conditional point |
| Double bottom | t 2 | 106 | Supplied conditional point |
| Double bottom | t 3 | 100.2 | Supplied conditional point |
| Double bottom · test | After cutoff, separately | 107 | C>106 |
| Double bottom · failure branch | After cutoff, separately | 99 | C<100 |
Why is difference 0.2 accepted here?
Because tolerance 0.3 was specified before the outcome. Enlarging it afterwards to save a preferred pair changes the experiment.
Flag and pennant: parallel or converging
Both independent cases begin with a 100→120 pole. The flag continues 115→118→113→116→111: U(t)=121−t and L(t)=117−t are parallel, width 4. The pennant continues 113→119→114→118→115: U(t)=120.5−0.5t and L(t)=112+0.5t give width 8.5−t, equal to 2.5 at t 6. These are different pause geometries after an advance.
No breakout is supplied at cutoff t 6. For a separate t 7, choose close above U: flag C115>114, pennant C118>117. Independent C109<110 and C114<115.5 breach the lower boundary of the chosen upward hypothesis. These are exercise conditions, not continuation or fill guarantees. A pennant outline alone is not a validated Elliott triangle: its subdivisions and permitted parent position need independent evidence.
| Outline / branch | Point / stage | Price | Condition / status |
|---|---|---|---|
| Flag | t 0 | 100 | Supplied conditional point |
| Flag | t 1 | 120 | Supplied conditional point |
| Flag | t 2 | 115 | Supplied conditional point |
| Flag | t 3 | 118 | Supplied conditional point |
| Flag | t 4 | 113 | Supplied conditional point |
| Flag | t 5 | 116 | Supplied conditional point |
| Flag | t 6 | 111 | Supplied conditional point |
| Flag · test | After cutoff, separately | 115 | t 7: C>U(7)=114 |
| Flag · failure branch | After cutoff, separately | 109 | t 7: C<L(7)=110 |
| Pennant | t 0 | 100 | Supplied conditional point |
| Pennant | t 1 | 120 | Supplied conditional point |
| Pennant | t 2 | 113 | Supplied conditional point |
| Pennant | t 3 | 119 | Supplied conditional point |
| Pennant | t 4 | 114 | Supplied conditional point |
| Pennant | t 5 | 118 | Supplied conditional point |
| Pennant | t 6 | 115 | Supplied conditional point |
| Pennant · test | After cutoff, separately | 118 | t 7: C>U(7)=117 |
| Pennant · failure branch | After cutoff, separately | 114 | t 7: C<L(7)=115.5 |
Can any narrowing pause be called this flag?
No. Our flag boundaries are parallel; the pennant boundaries converge. Elliott classification is a separate review.
A falling wedge is not an established diagonal
Take 120→110→116→108→112→106 through t 5. Upper points follow U(t)=120−2t; lower points follow L(t)=111−t. Both lines descend and width 9−t narrows from 8 at t 1 to 4 at t 5. That supplies falling-wedge geometry; both boundaries slope upward in a rising wedge. The later breakout direction is not already observed at this cutoff.
For a separate t 6, choose C>U(6)=108; C109 passes. Independent C104<L(6)=105 rejects the chosen upward hypothesis. This is neither statistics nor a reversal promise. An Elliott diagonal additionally needs appropriate position and internal components; consult the existing diagonal example rather than transferring its status to any wedge. Full bars and execution are absent here.
| Outline / branch | Point / stage | Price | Condition / status |
|---|---|---|---|
| Falling wedge | t 0 | 120 | Supplied conditional point |
| Falling wedge | t 1 | 110 | Supplied conditional point |
| Falling wedge | t 2 | 116 | Supplied conditional point |
| Falling wedge | t 3 | 108 | Supplied conditional point |
| Falling wedge | t 4 | 112 | Supplied conditional point |
| Falling wedge | t 5 | 106 | Supplied conditional point |
| Falling wedge · test | After cutoff, separately | 109 | t 6: C>U(6)=108 |
| Falling wedge · failure branch | After cutoff, separately | 104 | t 6: C<L(6)=105 |
Do converging inclined lines establish an ending diagonal?
No. They establish geometry only. Wave position, subdivisions and remaining requirements are unproven.
Cup and handle: rounding and a smaller pullback
The advance 90→110 precedes the cup and is outside its depth measurement. The cup at t 1–t 7 is 110→106→102→100→102→106→110, with equal rims 110 and depth 10. Handle 108→109→107.5 then pulls back 2.5 from the rim,25% of cup depth. This is only a schematic U followed by a smaller pause. t 0–t 10 are points, not 11 actual bars: no specific detector’s minimum duration or pivot settings are claimed satisfied.
Future data is unknown at t 10. Our chosen later test is stricter than merely crossing a sloped handle line: completed C>fixed rim 110. Separate C111 passes; independent C99 below base 100 rejects the preserved-cup hypothesis. Depth 10 is no promised target, and the cup is not automatically wave two without internal evidence. Geometry, Elliott validation and trade execution remain separate records.
| Outline / branch | Point / stage | Price | Condition / status |
|---|---|---|---|
| Cup and handle | t 0 | 90 | Supplied conditional point |
| Cup and handle | t 1 | 110 | Supplied conditional point |
| Cup and handle | t 2 | 106 | Supplied conditional point |
| Cup and handle | t 3 | 102 | Supplied conditional point |
| Cup and handle | t 4 | 100 | Supplied conditional point |
| Cup and handle | t 5 | 102 | Supplied conditional point |
| Cup and handle | t 6 | 106 | Supplied conditional point |
| Cup and handle | t 7 | 110 | Supplied conditional point |
| Cup and handle | t 8 | 108 | Supplied conditional point |
| Cup and handle | t 9 | 109 | Supplied conditional point |
| Cup and handle | t 10 | 107.5 | Supplied conditional point |
| Cup and handle · test | After cutoff, separately | 111 | C>110 |
| Cup and handle · failure branch | After cutoff, separately | 99 | C<100 |
Does depth 10 guarantee another 10 of upside?
No. It measures the outline, not a promised price or return. This exercise tests only the chosen later condition.