The double bottom ("W") and double top ("M") are classic counter-trend reversal patterns. A double bottom marks the potential end of a downtrend with two roughly equal lows separated by an intervening rally; a double top mirrors it at the end of an uptrend with two roughly equal highs separated by an intervening pullback. They reflect a simple narrative: price tests a level twice, fails to break it, and momentum flips. For a swing trader, the appeal is a clean structure with a well-defined entry, stop, and target — but the pattern only "exists," in any tradable sense, once confirmed.

The setup

The pattern has two parts: the twin extremes and the confirmation.

• Twin extremes. Two lows (double bottom) or two highs (double top) at approximately the same price. They need not be identical — within a few percent is normal — but a wildly uneven pair weakens the read. The two extremes are usually separated by days to weeks for a swing-relevant pattern.
• The intervening level = the "neckline." Between the two lows sits the intervening rally peak; between the two highs sits the intervening reaction trough. This peak/trough is the line that defines confirmation.
• Trigger / confirmation. The pattern is only valid once price closes through the neckline — above the intervening peak (double bottom) or below the intervening trough (double top). Per StockCharts and Bulkowski, a wick through the line is not confirmation; wait for the close. Volume expansion on the break adds conviction.
• Entry. On the confirmed neckline break (breakout-close entry), or on a retest of the neckline from the new side. The retest entry gives a tighter stop but may never come.
• Stop. Beyond the second extreme — below the second low for a double bottom, above the second high for a double top. That level invalidates the reversal thesis.
• Target (measured move). Project the pattern height — the distance from the extremes to the neckline — from the breakout point in the breakout direction.

Base rates & evidence

Bulkowski's measured statistics (thepatternsite.com) are the most-cited base rates here. They are not promises — they describe historical "perfect-trade" samples and vary by sub-type (Adam/Eve variants).

• Double bottom ("Big W"): Bulkowski reports a throwback rate of about 64%, roughly 74% meeting the measured-move price target, and a break-even failure rate near 9% (over ~2,100 perfect trades).
• Adam & Adam double top: Bulkowski reports a pullback rate of about 64%, about 64% meeting the price target, and a break-even failure rate near 25% (~1,114 perfect trades).

Two honest caveats: tops in his data fail materially more often than bottoms (the ~25% break-even rate on Adam & Adam tops is high), and "meeting target" is the full measured move — partial progress is common, full reach is not guaranteed. Bulkowski also notes that patterns which throw back/pull back tend to underperform those that don't, and that shallow throwbacks precede the best outcomes. Treat all figures as directional priors, not edges by themselves.

Strengths & limitations

Strengths. A defined structure: objective trigger, a logical invalidation level just past the second extreme, and a mechanical target. Favorable reward-to-risk when the pattern height is large relative to the stop distance.

Limitations. It is a counter-trend trade — you are betting against the prevailing move, which is structurally lower-probability than continuation setups. The #1 misuse is anticipation: entering at the second low/high before the neckline breaks. An unconfirmed "double bottom" is just a retest that may keep falling; most apparent patterns never confirm. Require the close through the neckline. Secondary failure modes: uneven extremes, no volume on the break, and immediate failed throwbacks/pullbacks.

System relevance

In the Augustus framework, reversals are treated as confirmation-required, lower-conviction setups. A double bottom/top contributes signal only after the neckline break is confirmed (close-through, ideally with volume), never on the second-touch anticipation. Given the elevated failure rate on tops in particular, the system weights these below trend-continuation setups and leans on the measured-move target and second-extreme stop for sizing rather than on the pattern's mere presence.

Sources

• Thomas Bulkowski, thepatternsite.com — "Big W Double Bottom" (throwback ~64%, target met ~74%, break-even ~9%) and "Adam & Adam Double Tops" (pullback ~64%, target met ~64%, break-even ~25%).
• StockCharts.com — Double Top/Bottom: neckline definition, close-through confirmation, measured-move target.
• Bulkowski, Encyclopedia of Chart Patterns — pattern taxonomy and measured-move methodology.