Double bottom and its four shapes

The definition, and the confirmation without which nothing exists

A double bottom is made of two bottoms at a comparable level, separated by a rally. Price falls, bounces, falls back to the same place, and bounces again. In Bulkowski's identification guidelines the pattern follows a decline, and the rally between the two bottoms measures at least 10%, with variations allowed.

The point most presentations forget: the pattern is only confirmed once price closes above the peak between the two bottoms. Before that close, there is no double bottom, there are two bottoms. The distinction is not academic, it is the difference between a pattern and a hope.

It is also what makes the pattern honest to trade: confirmation is a precise level, known in advance, that everybody can see in the same place. Unlike the triangle, there is nothing to draw and therefore nothing to interpret.

The bullish version is the double bottom, the bearish version the double top. They work as mirrors.

The four shapes, and the one that wins

The studies separate bottoms by their shape, and that is the most useful distinction in this lesson. A narrow, pointed bottom is called Adam. A wide, rounded bottom is called Eve.

Four combinations follow: Adam and Adam, Adam and Eve, Eve and Adam, Eve and Eve.

The Eve and Eve combination performs best: an average rise of 50%, against 43% for Adam and Eve, 42% for Eve and Adam and 39% for Adam and Adam, and 5th place out of 39 patterns, while the other three come 17th, 20th and 26th. On the bearish side, the Eve and Eve double top also carries the lowest failure rate of the four, but by a hair: 20%, against 21% to 25%.

In other words, two wide rounded bottoms are worth more than two spikes. A double spike is far more spectacular on screen, but the rounded shape is the one Bulkowski describes as the classic double bottom. One possible reading, which his statistics do not measure: a rounded bottom describes long absorption, a spike a panic low followed by a mechanical bounce.

Failure rate: 12% for Adam and Eve, Eve and Adam and Eve and Eve, 16% for Adam and Adam. Bulkowski counts as a failure a confirmed pattern that does not rise by at least 5%, which is not a rate of losing trades. That is more than twice the 5% of the cup and handle.

These figures come from the four pages Thomas Bulkowski devotes to each shape, Adam and Adam (1,154 perfect trades, his term), Adam and Eve (1,020), Eve and Adam (759) and Eve and Eve (952), consulted in October 2026. Two limits before these percentages travel. They were measured on stocks, on daily charts and in bull markets, and the average rise runs from the open of the day after the breakout to the highest price reached before a drop of at least 20%: it measures what price did, not the result of trades taken with a stop. And his pages give no equivalent for forex, indices, crypto or intraday timeframes: treat the numbers as unestablished there.

If you have read 4% or 5% elsewhere, that number is his as well: his overview of the four types, which refers to the second edition of his book and is illustrated by a chart dated 2010, still shows those failure rates, with average rises of 35% to 40%. The pages by shape refer to the third edition and were updated in 2025: theirs are the numbers this lesson uses.

How to trade it

The entry is at confirmation, meaning on the close above the intermediate peak, never on the second bottom. Entering on the second bottom means betting the pattern will form, and Bulkowski puts a number on that bet: in 48% of cases price continues lower without ever confirming. It is a different trade and it does not carry these numbers.

Two times out of three that entry gets shaken: price returns to the breakout level within thirty days in 65% to 67% of cases depending on the shape, and Bulkowski notes that these returns hurt performance.

The stop goes below the lower of the two bottoms. And here is the practical constraint: the taller the pattern, the further that stop sits from the entry, so the smaller the position must be. A beautiful wide pattern is an expensive pattern, like a hammer with a long shadow.

The target is the height of the pattern, from the lower of the two bottoms to the intermediate peak, projected from the breakout point. It is reached in 73% of cases for Adam and Adam, 72% for Eve and Adam, 69% for Adam and Eve and 65% for Eve and Eve: the best ranked shape is also the one that reaches its target least often. Bulkowski suggests a more cautious target, the height multiplied by that percentage before it is projected.

One simple filter follows from the shapes: if you have to choose between two double bottoms, take the one whose bottoms are wide and rounded rather than pointed.

What it gives in your own journal

The measurement specific to this pattern is obvious and almost nobody does it: tag the shape of your two bottoms, rounded or pointed, at the moment you take the position. Three trades per category will say nothing, thirty will start to speak.

If the announced gap shows up for you, you have just turned a decent pattern into a selection criterion, and it is free: it needs no indicator and no tool, just a label at the time of the trade.

The second measurement is the timing of your entry. Record whether you entered at confirmation or before it. If your early entries show a markedly lower win rate, you have the answer to a question many traders ask themselves without ever measuring it.

Tradoshi lets you combine two labels in the journal filter, so shape and entry timing read together.

Frequently asked questions

Must the two bottoms be at exactly the same price? No, at a comparable level. Bulkowski records an average gap of 1% for Adam and Adam, and gaps usually between 0% and 4% for Eve and Adam and between 0% and 6% for Eve and Eve. His criterion is visual: the two bottoms should appear to bottom near the same price.

When is the pattern confirmed? On the close above the peak between the two bottoms. Before that, there is no pattern.

Adam or Eve, how do you decide? Narrow and pointed for Adam, often a single one-day spike; wide and rounded for Eve, which widens as it rises. The sorting is not always clear-cut: in 2025 Bulkowski reclassified one of his own examples. Eve and Eve is the best performing combination.

What about the triple bottom? Same logic with one more bottom, confirmed by a close above the highest peak between the bottoms. Bulkowski calls it somewhat rare and gives it a page of its own: a 13% failure rate, a 46% average rise and 12th place out of 39, across more than 2,500 perfect trades, numbers close to those of the double bottom.

Is there a minimum time gap between the two bottoms? Bulkowski sets no minimum, he describes what he observes: bottoms several weeks apart. The median is 16 days for Adam and Adam and 23 days for Eve and Adam, the average is almost two months for Adam and Eve, and for Eve and Eve the gap mostly falls in the two to seven week range.