Concept
Double Top/bottom
Double Top/bottom, also known as Adam & Eve variants, is a Chart & Candlestick Patterns concept. The Library holds 1 implementation — a working definition you can pull into Quant.
Top Double Top/bottom indicator
The top custom implementation, built on the original standard Double Top/bottom formula.
1 total
From studying Double Top/bottom to trading it: take the implementation below into Quant and backtest it instantly.
.png&w=3840&q=75)
What is a Double Top/bottom?
A double top is two swing highs at about the same level with a pullback between them; a double bottom is the mirror image at the lows. The pattern is a failed second test: the retest of resistance (or support) cannot attract follow-through, and the traders who chased the second push are left offside. Nothing is confirmed at the second touch. The classic completion rule is a break of the interim swing between the two tests (the neckline); only then has a failed retest turned into a change of behavior.
The pattern anchors the classical reversal canon, catalogued from the earliest chart-pattern taxonomies onward, and it anchors the measurement tradition too: counting studies have repeatedly found that strict double tops confirm far less often than casual chart reading assumes, and that outcomes hinge on the confirmation rule chosen. The 'Adam and Eve' shape labels come from that measuring literature, which found the sharp-versus-rounded distinction worth cataloguing.
Tolerance matters: the two extremes rarely match to the tick, and a second peak that slightly overruns the first often just clears the stops resting above equal highs before turning. Pattern references also distinguish shape: an 'Adam' test is sharp and spiky, an 'Eve' test wide and rounded, giving the Adam & Eve combinations. The measure rule projects the pattern's height from the neckline for a first objective.
Two readings of the second test now coexist. The classical read wants a clean hold below the first extreme, demand simply failing to reach the old price; the liquidity read prefers the overrun, a wick through equal highs that collects the resting stops and closes back inside, treating the sweep as the stronger evidence because the level's fuel was spent. Both readings converge on the same completion discipline: the neckline decides, and until it breaks the structure is just a range whose upper boundary has been tested twice.
How to identify a double top or bottom
The double-top case at highs is described; mirror every element for double bottoms.
- 1Require a trend into the pattern: the first peak should cap a meaningful advance, since the structure is a reversal claim.
- 2Demand a real interim pullback: a distinct swing low between the tests, deep enough that the two peaks are separate auctions rather than one level test.
- 3Apply tolerance at the second test: equal to the tick is rare, and both the slight undershoot and the stop-clearing overrun qualify if the level rejects.
- 4Grade the second test's character: rejection candles, fading volume versus the first peak, or a sweep-and-reclaim all strengthen the failed-retest read.
- 5Wait for the neckline: a decisive close through the interim low completes the pattern; before that, the structure is a range, not a reversal.
- 6Set the references: measure-rule target projected from the neckline, invalidation on acceptance back above the double extreme.
How traders use it
- As a confirmation-based reversal trade: entry on the neckline break (or its retest), stop beyond the double extreme, first target from the measure rule. The second touch alone is a level test, not a completed pattern.
- As a liquidity read: a second high that wicks above the first and closes back inside is treated by many traders as a stop sweep that strengthens the case, while a clean hold below the first high is the more conventional double top.
- As a volume check: classic references look for lighter volume on the second test and expansion on the neckline break; a quiet, low-participation break is easier to fade back into the range.
- With candle triggers at the second test: an engulfing bar or pin bar rejecting the retest gives aggressive traders a defined early entry, with the neckline break held back as the add or the confirmation.
- In scanners with honest statistics: the geometry codes easily (two peaks within tolerance, interim retracement minimum), and measured confirmation rates per market keep the pattern's real base rates, rather than its reputation, driving the sizing.
Double Top/bottom vs related structures
Ascending/descending/symmetrical Triangle: A descending triangle also presses a flat boundary, but its second side compresses (falling highs) rather than swinging freely. Many failed double bottoms evolve into triangles as the rotations shrink, which is why chartists re-classify rather than force the first label.
Two-bar Reversal: The two-bar reversal is the bar-scale expression of the same failed-retest logic: commitment one way, immediate rejection back. A double top's second test often contains one at its extreme, the micro trigger inside the macro pattern.
Hikkake: The hikkake is a trap pattern built from an inside bar's false break; the double top is a trap built from a level's false retest. Different scales, same engine: positioning committed to a break that fails, then fueling the reversal.
Concept family
Chart & Candlestick Patterns
84 concepts mapped · 84 in the Library
Double Top/bottom FAQ
How far apart do the two tops or bottoms need to be?
There is no fixed rule. Classic chart-pattern references want two distinct swings separated by a meaningful pullback, not two candles side by side, and many screeners require the interim dip to retrace a noticeable fraction of the first advance. The closer together the two tests sit, the more the structure reads as a single level test or a small range rather than a double top.
When is a double top or double bottom confirmed?
The common rule is a close beyond the interim swing between the two tests, the neckline. Until that break, price is simply holding a range between support and resistance, and the structure can resolve as continuation instead. Confirmed patterns still fail, so most traders pair the break with a stop beyond the double extreme rather than treating the signal as settled.
What do Adam and Eve tops mean?
Shape labels from the pattern-measurement literature: an Adam test is sharp, narrow, and spiky, often a single wick; an Eve test is wide and rounded, built from many bars. Combinations (Adam-Adam, Adam-Eve, Eve-Eve, Eve-Adam) describe the two tests in order, and the catalogued statistics differ modestly across them, mostly as a reminder that not all double tests are the same auction.
Is a second peak that exceeds the first still a double top?
Within tolerance, yes, and in the liquidity reading it is the preferred version: the overrun sweeps the stops resting above the first high and closes back inside, spending the level's fuel. What disqualifies the pattern is acceptance above the first peak, closes holding beyond it, which converts the structure into a breakout and retires the reversal claim.
How often do double tops actually confirm and work?
Measured far less often than folklore implies: counting studies find most twin-peak structures resolve as ranges or continuation, with outcomes highly sensitive to the tolerance and confirmation rules used. The honest takeaway is procedural, trade the confirmation rather than the picture, keep the stop beyond the extremes, and let per-market statistics rather than the pattern's fame set expectations.
What is the difference between a double top and a triple top?
One more failed test of the same area, with the same completion logic at the neckline. Triple structures overlap with rectangles (the boundary has now been proven three times), and the extra test cuts both ways: more evidence the level matters, and more time for the structure to resolve as a range instead. Most playbooks treat the distinction as descriptive rather than a different trade.
Turn Double Top/bottom into a trading strategy.
Take the implementation from this page into Quant, then build on it, backtest it on real data, and keep refining it with AI.