Double bottom pattern in crypto: how to read and confirm a W

Identify the two bottoms and the neckline

A double bottom is a reversal candidate that can follow a decline. Its two troughs form a W around an intervening rebound, but the shape alone does not confirm it.

These shapes can look alike while price is still moving. Use the sequence, not the number of touches alone:

The neckline is drawn through the intervening reaction high. If there are several minor highs, choose the swing point your method defines and use that rule consistently. A sloped neckline changes with time, so evaluate its value where the proposed breakout occurs. A horizontal line at the intervening high is a useful simplification only when the chart supports it.

Some studies classify the shape of the troughs. In Bulkowski’s terminology, an Adam bottom is relatively narrow and V-shaped, while an Eve bottom is wider and more rounded. Adam-and-Adam (A-A) and Adam-and-Eve (A-E) are different subtypes; their results should not be merged into one number for every W-shaped crypto chart. Shape labels are descriptive, and borderline examples can be hard to classify in real time.

Separate the second low from confirmation

The pattern is only a candidate while price is making the second low. One confirmation rule is a completed candle close above the neckline. Another is a move above the neckline followed by a retest in which the former resistance holds as support. A wick above the line that closes underneath is not the same event as a completed close above it. The close rule can reduce ambiguity, but it enters later; the retest rule can wait still longer and miss moves that do not return.

A breakdown below both troughs weakens or invalidates the bullish interpretation. After an upside neckline break, a quick move back under the line is evidence that price did not retain the breakout on that interval. Whether to exit on the first close, a retest failure or a deeper structure level is a strategy choice. Keep the chart description (“pattern invalidated”) separate from an account-level stop and position-size decision.

Volume may add context: some traditional descriptions look for lower activity during the second decline and increased activity on the breakout. Crypto volume is venue- and instrument-specific, however, and activity alone does not establish net buying or a higher probability. Pair the volume observation with the price close and use the same feed in any subsequent test.

Calculate a synthetic measured move

Suppose a synthetic chart has lows near 90, with an intervening high and neckline at 100. The pattern height is 100 − 90 = 10. If a later candle closes above the neckline at 102, a traditional projection adds the height to the breakout area: 100 + 10 = 110 using the neckline, or about 112 if the method anchors from the 102 close. Choose one convention in advance; neither projected price is a guaranteed target.

Before treating the projection as a possible objective, check resistance above the current price, prior highs and volatility. If a resistance area sits at 106, it may interrupt the move well before 110 or 112. Also consider the distance from the planned entry to invalidation. A measured target twice as far away as a stop can still lose money often enough to be unsuitable once costs and failed breakouts are included.

For a meaningful outcome review, record every detected W, including ones that never confirm. Write down how far apart the two low prices may be and the minimum number of candles between them. Also record how you draw the neckline, what counts as a breakout, the entry, stop and target, how long you track the outcome, and execution costs. Without the pre-confirmation candidates, a study of only completed patterns can hide how often attractive W shapes never break upward.

Read historical double-bottom statistics by subtype

Study scope: the figures in this section come from ThePatternSite’s bull-market results for historical stocks. Bull market describes the source’s market regime; it does not mean that every individual pattern or breakout was bullish. These figures do not combine all market regimes.

ThePatternSite’s current Adam-and-Adam page reports 1,154 historical stock “perfect trades”: 16% break-even failure, 39% average rise, 67% throwback rate and 73% meeting the page’s price target. Its Adam-and-Eve page reports 1,020 samples: 12% break-even failure, 43% average rise, 67% throwback and 69% target attainment. The site defines break-even failure as failure to rise at least 5% after an upward breakout; a throwback is a return toward the breakout level.

Those are different endpoints: the 5% favorable-move threshold is not the same as reaching the measured target. The pages analyze historical stock patterns, not cryptocurrency; they do not validate a detector on 5-minute, 1-hour or daily crypto candles. The figures do not include a specific trader’s entry, stop, exchange costs, leverage or funding, so calling them crypto “success rates” would overstate what the sample establishes.

The data shows why the subtype and the outcome being measured matter. Its percentages do not give the odds for a live setup. To assess a crypto rule, collect all qualifying candidates from a named exchange and period, including delisted or inactive markets where relevant. Report both favorable price moves and returns after costs, with clear assumptions about entries, exits and order execution.

altFINS reports about 82% for double bottoms flagged by its engine and defines guide-wide success as a forecast target hit within one pattern length. This is a platform-specific target event, not the historical stock 5% move or net return. Its public summary does not state the sample size or study period for double bottoms.

Common failures and practical tradeoffs

The second bottom undercuts the first. A small wick below the first low can be part of the second test, but a close and continued trade beneath both troughs undermines the reversal thesis. Decide whether your test uses lows, closes or a volatility tolerance.

The neckline never breaks. A W-looking shape can remain a range or roll over. Do not count it as a confirmed double bottom simply because the chart later rose from the right edge of a screenshot.

Price breaks out and returns below the neckline. This is a failed breakout under a close-and-hold rule. Waiting for a close or retest may avoid some brief spikes but can also reduce the reward-to-risk of a successful move. This is a tradeoff to evaluate, not a free confirmation filter.

The target is blocked by nearby resistance. A geometric projection ignores the rest of the chart. Map prior highs and likely liquidity areas before using a measured move as an exit reference. The multi-timeframe guide describes how to inspect context without treating overlapping charts as independent proof.

Frequently asked questions