Rising and Falling Three Methods: Continuation Structure
Rising three methods is a five-candle continuation pattern in an uptrend: a long bullish body, three small bearish bodies inside the first candle's high-low range, then a long bullish candle closing above the first close. Falling three methods mirrors it in a downtrend, with three small bullish bodies between two long bearish candles and a final close below the first close. Completion describes the observed sequence; it does not guarantee that the next candle will continue the move.
Key Takeaways
- Both directions use five candles, including a three-candle pause. Three white soldiers and three black crows have no intervening pause.
- Body containment and full high-low containment are different tests. State which one your detector uses.
- A final close beyond the first close is different from a close beyond the first high or low.
- Freeze body-size thresholds and prior-trend rules before evaluating a sequence.
- An invalid shape and a valid pattern followed by a losing trade are different outcomes.
Read the rising and falling structures together
TradingView publishes the two directions as mirrored five-candle patterns. Its descriptions specify middle bodies inside the first candle's range and a final close beyond the first close. They do not state that every middle wick must stay inside that range. The table separates those published conditions from optional restrictions. Number candles from 1 to 5, oldest first.
| Condition | Rising three methods | Falling three methods |
|---|---|---|
| Prior context | An existing uptrend | An existing downtrend |
| Candle 1 | Long bullish body: close > open | Long bearish body: close < open |
| Candles 2, 3, and 4 | Three short bearish bodies | Three short bullish bodies |
| Middle-body boundary | Each open and close inside candle 1 low to high | Each open and close inside candle 1 low to high |
| Candle 5 | Long bullish body; close 5 > close 1 | Long bearish body; close 5 < close 1 |
| Optional stricter breakout | Close 5 > high 1 | Close 5 < low 1 |
Choose strict or loose rules before searching a chart
For the examples below, define a long body as more than 2 price units and a short body as no more than 1.5 units. These illustrative limits are not TradingView detector defaults or recommended settings. In a reusable study, replace the fixed units with a stated measure such as the mean absolute body over the 20 bars before candle 1, held constant throughout the pattern. Use inclusive boundaries for containment: touching the first high or low is allowed. Keep the final-close comparison strict: equality does not complete the pattern. Measure trend before candle 1; one possible context filter compares the preceding close with the close ten bars earlier. That is a chosen research rule, not proof of a sustained trend.
- Body-contained version: every middle open and close is between low 1 and high 1. A wick may extend outside.
- Strict wick-contained version: every middle low is at least low 1 and every middle high is at most high 1. Apply the same body and color rules.
- Tolerance version: allow the selected boundary to extend by a fixed number of minimum price ticks. Record the tolerance and whether it applies to bodies, wicks, or both.
- A two- or four-candle pause, a mixed-color pause, or a doji allowance changes the detector. Label it a variant and report its results separately from the five-candle version.
- A close beyond high 1 or low 1 adds a breakout filter. Do not silently substitute it for a close beyond close 1.
Rising three methods: a five-candle OHLC example
Assume an uptrend before candle 1. The first real body is 106 - 100 = 6 units, and the final body is 109 - 102 = 7. Each middle body is 1 unit. All three middle ranges fit within the first range of 99 to 107, so both body containment and strict wick containment pass. The final close of 109 exceeds the first close of 106 and the optional breakout level of 107. These are invented prices for checking the rules. The chart does not show the preceding trend or establish that a trade would be profitable.
| Bar | Open | High | Low | Close |
|---|---|---|---|---|
| 1: advance | 100 | 107 | 99 | 106 |
| 2: pause | 105 | 105.5 | 103.5 | 104 |
| 3: pause | 104 | 104.5 | 102.5 | 103 |
| 4: pause | 103 | 103.5 | 101.5 | 102 |
| 5: completion | 102 | 110 | 101.5 | 109 |
Hollow bodies close at or above the open. Filled bodies close below the open.
Hypothetical OHLC prices. Long body > 2 units; short body <= 1.5 units. Trend context is assumed outside the diagram.
Falling three methods: the same checks in reverse
Assume a downtrend before this sequence. Candle 1 has a bearish body of 120 - 114 = 6 units, and candle 5 has a bearish body of 118 - 111 = 7. The three bullish middle bodies each measure 1 unit. Their full ranges stay within 113 to 121. The final close, 111, is below both the first close, 114, and the first low, 113. This is the arithmetic mirror of the rising example around a price of 110: subtract each price from 220 and swap the transformed high and low. The mirror makes it possible to audit whether a detector accidentally applies a bullish inequality to bearish prices.
| Bar | Open | High | Low | Close |
|---|---|---|---|---|
| 1: decline | 120 | 121 | 113 | 114 |
| 2: pause | 115 | 116.5 | 114.5 | 116 |
| 3: pause | 116 | 117.5 | 115.5 | 117 |
| 4: pause | 117 | 118.5 | 116.5 | 118 |
| 5: completion | 118 | 118.5 | 110 | 111 |
Hollow bodies close at or above the open. Filled bodies close below the open.
Hypothetical OHLC prices using the same body thresholds. This is a mirrored rule example, not an observed market event.
Boundary breaches and missing continuation are different failures
Change one condition at a time to find out which rule rejects a sequence. The first four rows below are edits to the rising example; the last row describes a later outcome after its original five candles have completed. A wick below the boundary rejects strict containment but does not, by itself, reject body containment or prove a reversal.
| Change | Result | Reason |
|---|---|---|
| Candle 3 low becomes 98.8; its body stays 104 to 103 | Body-only passes; strict wicks fail | 98.8 is below low 1 = 99. A separately specified 0.2-unit wick tolerance would admit equality. |
| Candle 4 becomes O 99.5, H 100, L 98.5, C 98.8 | Both containment versions fail | The close lies below 99 even though its bearish body is only 0.7 units. |
| Candle 5 closes at 106 with other prices unchanged | No completed rising pattern | Close 5 equals close 1 instead of exceeding it; an intrabar high of 110 is insufficient. |
| Candle 5 closes at 106.5 with other prices unchanged | Basic close rule passes; extra breakout filter fails | 106.5 exceeds close 1 = 106 but remains below high 1 = 107. |
| After the original completion at 109, a later bar closes at 98 | A valid historical pattern followed by failure | A study using a close below the pattern low of 99 as invalidation would now reject the continuation thesis. |
Turn the recognition rules into a testable workflow
Wait for candle 5 to close before recording completion. TradingView documents that realtime indicator values can update before the bar closes; an intrabar marker may disappear. Closed-bar evaluation handles that source of change, but does not establish that every data request in a script is free of repainting. A detector supplies event times. To test trades, add an entry rule, an exit rule, position size, costs, and a fixed evaluation window. For example, an assumed fill at 109.2 and a stop trigger at 98.9 creates 10.3 units of planned price risk per unit held, before costs or adverse execution. It does not create a guaranteed maximum loss or a pattern-derived profit target. Compare body-only and wick-contained detectors on the same ordinary price candles. Hold entries and exits constant, count the events each version excludes, and reserve a separate period for evaluation. Neither a stricter rule nor an added volume filter is automatically better. This page reports no win rate, best timeframe, or tested performance.
Pineify next step
Specify both directions in the Pineify AI Coding Agent
Ask the Pine Script AI Coding Agent to draft a detector with a three-bar pause by default, explicit body thresholds, a body-only versus wick-contained switch, and tolerance measured in minimum price ticks. Request separate labels for variants with a different pause length and signals only after the final bar closes. Review the generated conditions, request a syntax check, and compare chart markers with the examples above. Code generation does not establish a trading edge.
Sources and verification
Last verified:
Educational information, not investment advice. Prices, thresholds, and execution assumptions are hypothetical. Pattern completion does not guarantee continuation or profit. Trading involves risk of loss, and stop orders can fill beyond their trigger prices.