Rising and Falling Three Methods, the Continuation Pattern

Most candlestick patterns get taught as a picture with a name attached, which makes them sound more exotic than they are. Strip the name off the three methods candlestick pattern and what is left is a trend that took a breather without giving anything back. One long candle in the direction of the trend, then three small candles that pause inside that candle's range, then another long candle that picks the trend back up and breaks out. Rising three methods is the bullish version. Falling three methods is the bearish mirror. The reason the pattern is worth knowing is not the name or the five-candle shape. It is the containment rule sitting in the middle of it, which is one of the few things in candlestick lore you can actually check rather than argue about. This post covers the rules, the near-misses that get mistaken for the real thing, where it quietly stops working, and how AI grading handles a structure like this one off a static screenshot.

Quick Answer

What the Three Methods Pattern Actually Is

The formation comes out of the Japanese candlestick tradition and reached Western charting mostly through Steve Nison's Japanese Candlestick Charting Techniques, which is where most of the pattern names in the modern candlestick pattern catalogue came from. In the original framing it sits in the continuation family alongside the mat hold, which is a close cousin with a gap in it. That family distinction is worth holding onto, because it is the single most useful thing about the pattern and the thing most cheat sheets bury under the drawing.

Here is the structure in plain terms for the rising version. Candle 1 is long and green, printed inside a move that was already going up. Candles 2, 3 and 4 are small, usually red, and they drift gently lower. Critically, none of them trades above candle 1's high or below candle 1's low. Candle 5 is long and green again and closes above candle 1's close, and in the cleaner examples it takes out the high of the entire formation. The falling three methods flips every colour and direction: long red candle inside a downtrend, three small green candles contained in its range, then a long red candle closing below candle 1's close.

What the pattern is telling you is about failure, specifically the failure of the counter-move. In a rising three methods, sellers had three complete bars to push price somewhere after a big green candle, and the best they managed was a drift that never escaped the range of that one candle. That is meaningfully different information from a random pullback, because the boundary is defined and the counter-move respected it. This is the same logic that makes a shallow flag more interesting than a deep one, and it belongs in the same mental shelf as the rest of the shapes covered in the full guide to reading and grading candlestick patterns.

One correction to the textbook before going further. The "three" is a convention, not a law. Two small candles or four small candles produce the same message, and real charts print the idea with counts that do not match the diagram far more often than they print the exact five-bar version. Insisting on three means throwing away good structure and, worse, it pushes people to count bars instead of checking the range. Containment is the test. The candle count is a description.

Row four is the one that does the work. The other six rows are quality filters, and you can trade a slightly imperfect version of most of them. Break containment and you do not have a weaker three methods, you have a different chart.

Rising Three Methods vs Falling Three Methods

Structurally the two are mirror images, and the drawing below shows them side by side. The rising version needs an uptrend, a long green candle, a contained red pause and a green breakout. The falling three methods needs a downtrend, a long red candle, a contained green bounce and a red breakdown. Same rules, opposite colours. The long candles at either end are frequently near-wickless, which means you are often looking at a marubozu doing the pushing at both ends of the formation, and that is a good sign rather than a coincidence. A full-body candle says one side ran the whole bar.

The asymmetry worth knowing is behavioural, not structural. Downtrends move differently from uptrends. They tend to be faster, they cover more ground per bar, and they produce violent counter-trend bounces that look identical to a contained pause right up until the bounce keeps going. So a three-candle green drift inside a downtrend is doing double duty in your head: it might be the pause before continuation, and it might be the first leg of a real bottom. The falling version therefore leans harder on candle 5 than the rising version does. Until that long red candle actually closes below candle 1's low, you are looking at a bounce with a story attached. Checking what the higher timeframe says before you trade the lower one matters more on the bearish version for exactly that reason.

The Rules That Separate a Real One From a Near-Miss

Work through a concrete rising three methods so the rules stop being abstract. Illustrative numbers, not a trade record. A stock has been grinding up all morning and prints a 5-minute bar from $48.20 to $49.60 on 1.8M shares, closing near the high. That is candle 1, and its range is $48.20 to $49.60. The next three bars are small and red: they close at $49.32, $49.15 and $49.21, and the lowest tick across all three is $49.04. Volume on those bars runs 480k, 340k and 260k. Nothing left the range, and participation dried up on the way down. Then a bar opens at $49.24 and closes at $50.10 on 2.2M shares, clearing the $49.60 pattern high. That is the whole pattern, and every rule is checkable against numbers you can read off the chart.

The volume shape in that example is not decoration. Contracting volume through the pause and expanding volume on the break is what separates a rest from a distribution. If those three small bars had printed on volume equal to or heavier than candle 1, the small bodies would mean something different: heavy two-sided trade going nowhere, which is a fight rather than a pause. Confirming that volume actually backs the move matters more here than on most patterns, because the pause candles are small by definition and the body size will not tell you anything on its own.

Location is the other filter. A three methods that forms with the pattern low sitting on a level price has already respected several times is a different proposition from one floating in the middle of nowhere, which is why mapping your support and resistance levels before the session does more for this pattern than any amount of candle-naming. And the breakout has to be real. Candle 5 closing back inside candle 1's range is a failed attempt, not a slightly weaker success, and it very often turns into the opposite trade.

• Take it
  Established trend, long candle 1, every pause candle contained inside its range on shrinking volume, candle 5 closing through the pattern extreme on expanding volume.
• Skip it
  No trend going in, a pause candle that escapes the range, pause volume as heavy as candle 1, or a candle 5 that closes back inside the range.
• Never
  Enter at the close of a wide candle 5 with the stop under the pattern low. That is the full pattern of risk taken at the worst price of the bar.

Three Methods vs Bull Flag vs Three White Soldiers

The fastest way to place this pattern is against the two continuation shapes it lives next to. The bull flag describes the same event with looser rules: a strong move, a pullback that gives back very little, a resumption. The rising three methods is that idea specified tightly enough to be checked, with the pullback confined to a single candle's range and the count conventionally fixed at three. That tightness makes it objective and also makes it rare. The three white soldiers formation is a different animal entirely, because nothing pauses. All three candles push the same way, so it is a statement about relentlessness rather than about a counter-move failing.

There is a fourth relative that belongs in this conversation and does not fit the table. The harami, or inside bar, is containment at its smallest scale: one candle sitting entirely inside the previous one. The three methods is that same logic stretched across a handful of bars instead of one, which is why the two patterns tend to appear on the same charts and why traders who already respect inside bars pick this one up quickly.

Where the Three Methods Pattern Fails

The most common failure is not a bad break, it is a missing trend. A long candle followed by three small ones followed by another long candle is a shape that sideways chop produces constantly. Without a directional move going in, there is nothing to continue and the formation is decoration. This is the failure mode that costs beginners the most, because the picture looks textbook-perfect while carrying no information. Establishing that the chart has a direction before reading any individual bar is the first job, which is the whole argument behind getting trend direction settled first.

The second failure is liquidity. Three small candles on a thin name can be an absence of prints rather than a controlled pause. Containment across bars that traded a few hundred shares each is not evidence that sellers failed, it is evidence that nobody showed up. The pattern needs a name with real two-way flow for the message to mean anything, and it degrades badly on illiquid tickers and in the dead middle of the session. The third failure is position in the move. A three methods that forms after price has already run a long way from its moving averages with no meaningful pullback is much more likely to be the last pause before the top than the middle of the trend, which is the same question of whether the chart is overextended before you enter that ruins a lot of otherwise clean breakout entries.

Then there is the arithmetic trap on candle 5, which is inherited straight from any wide breakout bar. The logical stop for a rising three methods long sits below the pattern low, which is also candle 1's low. Enter at the close of a long candle 5 and you are risking the entire pattern range on a trade taken at the highest price of the sequence. In the example above, entering at $50.10 with a stop under $48.20 means risking $1.90 per share for a target that did not move. The fix is to treat the break as the trigger for a plan rather than a reason to hit market at the close, and to size from the stop instead of from enthusiasm. Frequent intraday trading carries a substantial risk of loss on its own, as FINRA's guidance on day trading spells out, and a pattern with a memorable name does not change that arithmetic.

Last, be honest about reliability. There is no win-rate figure for this pattern that survives scrutiny, because every published test picks its own definition of containment, its own candle count, its own timeframe and its own market, and the results move around accordingly. The mechanism is defensible without a statistic. A counter-move that gets several bars and cannot escape one candle's range has told you something. That is worth using as one input among several rather than as a signal you size into on its own.

How AI Grading Reads a Three Methods Setup

Worth being precise about what a machine can and cannot do with this one. When you upload a static chart screenshot, the analysis reads the structure visible in that image, and the three methods happens to be unusually legible for that job. The whole pattern is geometry: the vertical extent of one candle, the highs and lows of the next few sitting inside it, and a final candle leaving that band in the original direction. No tick data required, no gap history, nothing happening off the visible chart. That puts it in a different bucket from patterns that depend on pre-market context or overnight gaps, which cannot be read reliably from a picture at all. The general version of that boundary is covered in the breakdown of what pattern detection can genuinely do on a chart image, and the neutral overview lives at AI chart analysis.

The fit with the grading engine is closer here than with most candlestick patterns, and it is worth saying why. The engine only grades momentum continuation. A long is a pullback inside an established uptrend resuming higher, and a short is a counter-trend rally inside an established downtrend resuming lower. It deliberately does not call reversals or counter-trend setups. A rising three methods is a pullback inside an uptrend resuming higher. A falling three methods is a counter-trend rally inside a downtrend resuming lower. Those are the two shapes the engine was designed around, so the pattern lands inside its remit rather than at the edge of it, which is not true of the reversal candles it will decline to trade no matter how textbook they look. If you want the setup-specific version of that read, there is a dedicated page for grading a rising three methods from a screenshot, and the trading approach it sits inside is the momentum playbook built around trading with the established direction.

The limits are the same as always and they matter. It does not scan the market for forming three methods patterns and it will not alert you when one completes, because it reads the screenshot you hand it and nothing else. It does not predict candle 5, since the bar that decides whether the pause resolves up or down has not printed and nothing can read it early. It does not name the pattern and hand you a green light. What it gives you is a consistent second read on the chart in front of you, applying the same standard to a contained pause at 3pm that it applied at 9:45. That consistency is the point, because the question that decides this trade is never "is this a three methods". It is whether the trend, the containment, the level and the volume all agree, and that question is easy to answer honestly on a fresh chart and hard to answer honestly on the fourth one after lunch.