Timeless Markets.Org
Educational only — not financial advice. Chart patterns carry only a small, context-dependent edge; advertised win-rates are usually misleading.
Concept · Definitive Guide

Chart Patterns

The oldest technical analysis — and the most over-sold. Here's the honest version.

Overview

Chart patterns are recurring shapes in price — head-and-shoulders, triangles, flags, double tops and bottoms, cups-with-handles — that traders believe tend to precede a characteristic move. They fall into two families: reversal patterns that mark a turn, and continuation patterns that mark a pause in a trend.

They are the oldest and most popular form of technical analysis — and also the most over-sold, so this is a topic where an honest look at the evidence matters more than usual.

Origins & history

How it works

A head-and-shoulders reversal pattern with a neckline
A reversal pattern — head & shoulders: a failed attempt at a new high, confirmed when price breaks the neckline.
An ascending triangle continuation pattern
A continuation pattern — a triangle: a pause that resolves, usually, in the direction of the prior trend on the breakout.

Most patterns share the same three parts: a formation (the shape), a trigger (a breakout of the pattern's boundary, ideally on rising volume), and a measured-move target (a projected objective based on the pattern's height). The breakout — not the shape alone — is what turns a picture into a trade.

Market psychology & mechanics

A good pattern is a picture of the supply-and-demand battle. A head-and-shoulders top, for example, is the story of buyers making a final push to a new high (the head), failing, and then being unable to better the previous peak (the right shoulder) — a visible shift from accumulation to distribution. A triangle is a coil of narrowing disagreement that must eventually resolve. The shapes matter because they encode who is winning.

Honest assessment

The evidence — read this carefully

There is real signal here. The academic study by Lo, Mamaysky & Wang (2000) applied automated pattern recognition to decades of US stocks and found that several patterns — including head-and-shoulders and double-bottoms — do carry incremental information. Patterns are not pure noise.3

But the popular "success rates" are misleading. The headline "this pattern wins 80–90% of the time" numbers are cherry-picked from textbook-perfect examples and depend entirely on how you define success. Bulkowski's own large-sample statistics are far more sober — for many patterns only about half reach their measured-move target, with meaningful failure rates.2

Evidence rating: small, genuine statistical content — but highly dependent on context, volume confirmation, and an objective definition. The marketing win-rates are not real.

Weaknesses & failure modes

Professional uses vs. retail misuses

How professionals use them

  • As context, confirmed by volume and a real breakout — never the shape alone.
  • With a defined invalidation (the level that kills the pattern) and measured-move target.
  • Weighted by Bulkowski-style statistics, not folklore.

Common retail misuses

  • Trading the shape before the breakout.
  • Believing the "90% win-rate" marketing numbers.
  • Drawing the pattern they hope to see.

Going deeper

The families: reversals (head-and-shoulders, double/triple tops and bottoms, rounding tops) and continuations (flags, pennants, triangles, rectangles, cup-with-handle). Tools: the measured move, volume confirmation, and Bulkowski's performance rankings. Related: patterns are really just support/resistance and trend structure drawn as shapes.

Practice

Quiz 1 — What actually triggers a chart-pattern trade — the shape, or something else?

The breakout of the pattern's boundary, ideally on rising volume. The shape is only a setup; the breakout (and its confirmation) is the trigger. Trading the shape early is a common mistake.

Quiz 2 — A site advertises an "85% win-rate" head-and-shoulders. Should you believe it?

Be skeptical. Such numbers come from cherry-picked textbook examples and definition-dependent measurement. Large-sample statistics (Bulkowski) are far more modest — often only about half reach target.

Quiz 3 — Do chart patterns contain any real signal at all?

Yes — a small one. Lo, Mamaysky & Wang found head-and-shoulders and double-bottoms carry incremental information. But the edge is small and context-dependent, not the easy money the marketing implies.

This concept in the knowledge graph

PrerequisitesHow to read a chart, Support & resistance
UnlocksCup-with-handle, breakout trading
RelatedTrend structure, Volume, candlestick patterns
Opposing viewThe randomness critique — most "patterns" are pattern-matching on noise

Resources

References (primary where possible)

  1. Origins — Richard Schabacker (named the patterns) and Edwards & Magee, Technical Analysis of Stock Trends (1948).
  2. Thomas Bulkowski, Encyclopedia of Chart Patterns — measured performance statistics — thepatternsite.com.
  3. Lo, Mamaysky & Wang, "Foundations of Technical Analysis," The Journal of Finance, 2000 — Wiley.