Double Bottom (W Pattern) — How to Trade It

The W pattern — the double bottom — is the mirror image of the double top and one of the most commonly recognized end-of-decline setups. The market tests the same level twice, defends it twice, and buyers take control. That's the theory. What about the measurements? Thomas Bulkowski studied 4,376 double bottoms from 1991-2025, and the results are decent: an average 37% rise with 16% failures. But the most interesting part is something else: the "ugly" version of the pattern — with a clearly higher second low — beats the classic one (+41% with 15% failures). Textbook aesthetics loses to measurement once again.

Step by step: how to recognize the pattern, what the numbers actually say, and how to trade it without falling into the usual traps.

How to Recognize a Double Bottom

The pattern consists of two lows at a similar level, separated by a clear bounce, after a downtrend.

Validity conditions (per Bulkowski's guidelines):

1. A downtrend before the pattern. Price needs something to rise from — the pattern reverses a decline. The more substantial the prior decline (with a caveat below), the more room for the move afterward.
2. Two clear lows at a similar level. The average price difference between the lows is about 1%, but deviations are allowed. The key question: do both lows look like a test of the same zone?
3. A bounce between the lows of at least ~10%. Two lows separated by a shallow wobble is consolidation, not a W.
4. Time gap between the lows — usually a few weeks (the median in the database is 16 days). Two lows on adjacent candles are noise.
5. Volume is usually higher on the first low, with a downward volume trend across the pattern. A second test of the low on quiet volume means supply didn't deliver.
6. Confirmation: a close ABOVE the highest peak between the lows. A hard requirement. Bulkowski states it directly: buying before confirmation gives you a 48% chance that price keeps falling. Without confirmation, it isn't a pattern, just two lows.

Just as with peaks, Bulkowski splits lows into Adam (narrow, spiky, often with a one-day spike) and Eve (wide, rounded). That gives four variants; Eve&Eve performs best — the lowest failure rate and the highest average rise. A practical way to tell them apart: ask whether the lows look similar or different, and whether they stay narrow all the way down (Adam) or widen toward the top (Eve).

What the Numbers Say (Bulkowski, Encyclopedia of Chart Patterns)

Terms these statistics rely on:

• Break-even failure rate — how often price, after confirmation, didn't even rise 5% before the move died.
• Average rise — from the breakout price to the highest peak before a decline of at least 20%, calculated on perfect trades — so it's a comparison tool, not a forecast.
• Target % — how often price reaches the measure-rule target.
• Throwback — a return to the breakout level within 30 days of an upside breakout. It doesn't invalidate the pattern as long as price doesn't close back below it; but patterns with a throwback statistically end up performing worse.

The numbers (bull market, US stocks, n = 4,376 patterns from 1991-2025):

What this tells us:

• The double bottom works, but it's not a star of the catalog. Rankings around 23-26 are solidly mid-pack — weaker than the inverse H&S (13/39) or the triple bottom (12/39). The pattern's popularity comes from how easy it is to spot, not from exceptional statistics.
• 48% with no confirmation is the most important number in this article. Buying "because the second low is holding" is a strategy with the success rate of a coin flip. All of the pattern's edge sits in the confirmation.
• A throwback in ~2/3 of cases — after the breakout, price most often returns to the breakout level. Anyone entering a full position on the breakout candle is signing up for a rough ride; anyone waiting with part of the position for the return usually gets a better price.

The standard caveat: US stocks, daily timeframe, bull-market bias in the sample. The W pattern shows up constantly on BTC/ETH (ETH's June 2022 bottom around $880-1,000 is often described as a stretched-out double bottom), but these specific numbers were measured elsewhere.

How to Trade a Double Bottom

Entry. Long on a candle close above the peak between the lows. A second variant: adding to (or entirely building) the position on a throwback — price returns to the broken resistance and defends it as support. Bulkowski also describes a handle variant: after confirmation, price builds a flat shelf instead of rising — a breakout from that shelf often kicks off the real move.

Stop loss. Just below the lower of the two lows. If the pattern is tall and the stop ends up too far from entry, Bulkowski's advice is direct: raise the stop (e.g., based on volatility) or skip the setup. An overly wide stop isn't "safety," it's a bad risk-reward ratio.

Target — the measure rule. Height = the highest peak between the lows minus the lower low. Multiply by the target-achievement rate (e.g., 73% for Adam&Adam) and add it to the breakout price. Treat any resistance along the way — especially multi-top zones — as the first profit target.

Setup selection — what statistically improves results:

• Big W — a double bottom with a tall, steep left flank (a sharp drop with no consolidation). Price often runs back toward the start of the decline.
• A flat base on the weekly chart, where the pattern looks like a pothole in the road — these setups tend to produce the biggest rises.
• Breakout volume above the 30-day average improves results.
• Avoid patterns sitting under multi-top resistance, and reversing "against" a strong higher-order downtrend — a bounce that's really just a correction in a bear market tends to end in a return to the decline.
• Higher-order trend context. From Bulkowski's lessons: a double bottom acting as a correction within an ongoing UPTREND (price pulls back, draws a W, and returns to the trend) is a statistically much friendlier setup than the same W trying, on its own, to reverse a months-long bear market. In the first case, the pattern trades with the current; in the second, against it — the chart looks identical, the result usually isn't.

A numerical example. A W on BTC: lows at $90,000, the peak between them at $98,000, confirmation on a close above $98,000. Height $8,000; realistic target 98,000 + (8,000 × 0.73) = about $103,800. A stop below the lower low ($89,500) gives a risk of $8,500 against a potential of about $5,800 — a poor R:R, so in practice it's more useful to enter on a throwback closer to $98,000 or find a tighter stop. An illustration of the mechanics, not a signal.

Ugly Double Bottom and Busted Patterns — Uglier Means Better

The ugly double bottom is a Bulkowski discovery from 2006: a W pattern where the second low is 5-15% HIGHER than the first. The textbook would tell you to reject it — the lows "should" be equal, after all. The measurement says otherwise: an average 41% rise versus 37% for all double bottoms, with 15% failures versus 16%.

The logic holds up: a higher second low isn't a flaw, it's earlier proof that supply is weakening — sellers couldn't even touch the previous low. Conditions: the second low sits 5-15% higher, with no lower low in between, and confirmation on a close above the highest peak between the lows (without that, as usual, there's no pattern). Bulkowski's best-case context: a multi-month decline turning into a steep, panicky drop (45-60 degrees), a low, a bounce, a higher low, and a strong volume breakout within about 11 days.

The other side: the busted double bottom. If, after confirmation, price rises less than 10%, reverses, and closes BELOW the pattern's low, the pattern is broken and becomes a bearish signal (trapped buyers turn into supply). A stop hit below the low is information, not an invitation to average down.

A myth to bust: "a perfect, even W is the best setup." The measurements say: the version with a higher second low performs better, the Eve&Eve variant beats the rest, and the outcome still depends most on whether you waited for confirmation. Chart aesthetics is not a statistical edge.

FAQ

When is a double bottom confirmed? When price closes above the highest peak between the lows. Buying before confirmation gives 48% odds of a further decline — a coin flip, not an edge.

What is an ugly double bottom? A W pattern with a second low 5-15% higher than the first. It performs better than the classic version: +41% average rise with 15% failures. A higher second low is earlier proof of weakening supply.

Which variant is best? Eve&Eve (two wide, rounded lows) — the lowest failure rate and the highest average rise across variants. But no variant saves an entry made without confirmation.

FAQ

Rafał — Krypto Bez Ściemy

Trader and founder of Strefa Tradingu. He’s been breaking crypto down on YouTube for years — no hype, no signals, with a focus on market structure and risk management.

kryptobezsciemy.plInstagramYouTubeKlub Zero