How to Trade Chart Patterns

Core idea

Patterns alone do not make a trader. The Encyclopedia gives an edge — three decades of statistics on how each pattern behaves — but Bulkowski's working method adds the things that statistics alone cannot supply: a bias (bottom fishing over momentum), a few favourite setups (the "diving board," pothole, swing-trade pullback, throwback retrace), a list of setups to avoid (the U-shape that traps long-side bullish patterns at the end of a horizontal top), and a context filter (trade in the direction of the market and the industry, not against them).

Author's argument: Buy when a confirmed bullish pattern appears. Sell when a confirmed bearish pattern appears. Add a rising market and a rising industry, plus a stock "aching to rise," and your odds improve materially. Everything else in this book is decoration on that skeleton.

The two halves of an edge

The statistical half (covered in the rest of the Encyclopedia) tells you which patterns have low failure rates and large average moves. The discretionary half (covered here) tells you where in the chart to look for them. Bulkowski's experience is that the same pattern — say, a double bottom — performs very differently depending on what came before it. A double bottom after a downtrend and flat base is a high-percentage long. The same double bottom at the end of a U-shape topping a long rally is a bull trap.

Why it matters

A reference book is only useful if you can convert its numbers into action. This topic is the bridge. It also reframes the cost of being wrong: the worst losses come not from bad patterns but from good patterns in the wrong location. The U-shape failure mode below explains more retail losses than any individual identification error.

What the topic buys you

• A starting screen — flip 600 charts in an hour, looking for one specific shape.
• A bias that scales — bottom fishing requires fewer trades as portfolios grow.
• Two veto setups that keep you out of common traps.
• A handful of swing setups whose entry, stop, and exit are all reducible to nearby price structure.

Key takeaways

Mental model — the working method

Copy sourceZoom

The four named setups

Diving board

A long horizontal base on the weekly chart, a sudden steep plunge, then a fast recovery back to the base of the board. The setup works across multiple years and market regimes — Bulkowski has found instances in 2013–2014, 2014–2016, and 2018, in unrelated industries. Two entries are possible: aggressive (buy near the plunge low) or conservative (wait for a close above the highest peak of the base). The stop sits just below the bottom of the diving board.

Pothole

A drop, a flat base that recovers but does not advance, then a brief secondary dip (the "pothole") that forms any bullish pattern — double bottom, triple bottom, head-and-shoulders bottom. The pothole is a bear trap: it pulls in late sellers, then reverses sharply. The defining precondition is that the flat base appears after a downtrend, not a rally. Confirm with a reason for the stock to rise (good quarter, raised guidance, sector tailwind).

U-shape trap (do not take long)

A strong uptrend, then a long horizontal pause whose length is proportional to the prior rise. A bullish pattern (double bottom, H&S bottom) forms inside the horizontal. The pattern confirms — and then busts. Bulkowski sees this setup constantly. The fact that the prior trend was bullish does not save the trade; the horizontal top is distribution disguised as continuation.

Standard pattern trade in the direction of the trend

Confirmed bullish pattern in an uptrend → buy, stop below the pattern, target by the measure rule (height of pattern added to breakout). Symmetrically for bearish. The Encyclopedia's per-pattern topics give the failure rates and average moves you need to size the trade.

Volatility surrounding entries: throwbacks and pullbacks

A throwback happens after an upward breakout: price climbs, pauses, and returns to the breakout price within ~30 days, then usually resumes. A pullback is the mirror after a downward breakout. Both occur in roughly two-thirds of trades and round-trip in about 12 days. Two practical consequences:

• Place stops with throwback geometry in mind. A stop pinned a few cents below the breakout will trigger on a normal throwback and shake you out before the move begins.
• Look for nearby overhead resistance (prior peaks, round numbers, sideways price between breakout and ~10% higher). When you see it, expect a throwback. When you don't see it, hope for the move to run.

Practical application

Example: applying the diving-board filter to a hypothetical airline

Imagine TXAR (a fictional carrier) traded sideways between $42 and $46 for 18 months on the weekly chart. A geopolitical shock takes the stock to $32 in three weeks on heavy volume. Demand for the route does not actually change — the company's earnings come in flat. Over the next eight weeks, price climbs back to $40, then chops between $38 and $42 for two months.

This is the diving-board template — flat base, plunge, recovery — but the conservative entry has not triggered yet (no close above the $46 ceiling of the prior base). Two paths:

Whichever path you choose, the industry filter applies: if the broader airline sector is rolling over, halve the size or skip.

Performance contests and the limits of catalogues

Bulkowski's last note in this topic — about performance contests — points at something honest: even the best statistical patterns lose money in some hands and make money in others. The contest winners tend to combine pattern recognition with tight, mechanical risk management. The Encyclopedia gives you the first half; the second is up to you.

Related lessons