What chart patterns are
Chart patterns are recurring arrangements of price that traders give names to, so they can be taught, discussed, and looked for. They cover single candles, areas price reacted at, and gaps left behind by fast moves.
Naming something makes it easier to talk about and easier to over-trust. A name implies a definite object with definite behaviour, when what is actually on the chart is a set of measurements someone chose to treat as significant.
What every pattern here has in common
Three things are true of all of them, and they matter more than the differences between them.
They read the past as evidence about the future. Each construction takes something that already happened — a rejection, a fast departure, an unfilled gap — and treats it as information about what price will do on returning. That step is an assumption, not an observation, and it is the same assumption in every case.
A person decides where they are. How long a wick counts as long, how fast a move counts as decisive, where a zone's edge sits: these are thresholds a reader chooses. Two readers with different thresholds mark different patterns on identical data, and both can defend their marks.
Published examples are chosen after the outcome is known. Almost every illustration in a course or a post is drawn on a chart where what happened next is already visible. Applying a method to past conditions to show how it would have performed is what regulators call back-testing, and investor-education material is explicit that such results are not actual performance. Selecting which past instance to show, knowing how it resolved, is a different task from selecting one in advance.
How to use this section
Read the entries with those three points in mind, and the useful question stops being whether a pattern works and becomes: whose definition, applied when, and reviewed how.
The entries are grouped by what they measure rather than alphabetically. Single-candle signals read one bar's shape. Areas and zones read a band price departed from. Imbalances and gaps read a space inside the move itself.
Mistakes that only show up at category level
Each page lists its own pitfalls. These are the ones that become visible only when you look at the whole category:
- Collecting instead of testing. Four half-applied definitions produce less usable evidence than one applied consistently for a month. This is the category's characteristic failure and the reason this reference lists few entries.
- Switching pattern after a losing run, which restarts the evidence-gathering from zero every time.
- Stacking patterns and calling it confirmation. Two discretionary marks agreeing is not independent evidence if the same person drew both.
- Assuming the families are interchangeable. A candle read over one interval and a band spanning several are different kinds of claim, and mixing them hides which one is doing the work.
What no pattern on this page can tell you
None of these constructions says how much to risk, where a position becomes wrong, or what price will do next. Those are separate decisions, and they are the ones that determine what a losing sequence costs.
Regulator guidance on this is blunt: there are no silver bullets, and programs based on past performance cannot guarantee future results. That applies to a shape on a chart exactly as it applies to a paid signal service.
In this section
Supply and demand trading
Areas price left quickly, treated as worth watching if price returns — and why the method's results depend on the rules you write rather than on the zones.
Read guide →Chart patternsHammer candlestick
A long lower wick with a small body near the top — and the same candle after a rise is called a hanging man, which is why the shape carries no direction.
Read guide →Chart patternsFair value gap (FVG)
The unfilled band between neighbouring candles inside a fast move — and the only construction here that is partly an artefact of the chart's own resolution.
Read guide →Key takeaways
- Every construction here reads a past move as evidence about future behaviour — an assumption, not an observation.
- Each one is drawn by a person, so two readers mark different patterns on identical data.
- Published examples are selected after the outcome is known, which is why they look stronger than live results.
- Depth on one written definition produces reviewable evidence; breadth across four produces none.
Frequently asked questions
What are chart patterns?
Named arrangements of price on a chart — single candles, areas price reacted at, and gaps left by fast moves — that traders look for and discuss. The name is a label for a set of measurements someone chose to treat as significant, not a description of a definite object.
Do chart patterns actually work?
The question cannot be answered in that form, because each pattern is defined by whoever draws it. What can be assessed is whether one person applied one written definition consistently, and what their own records show. Any general success rate offered without a checkable method should be treated as unsupported.
Which chart pattern should I learn first?
Whichever one you can define in numbers and then leave alone for a month. Depth on one definition produces reviewable evidence; breadth across four produces none.
What is the difference between a candlestick pattern and a price zone?
A candlestick pattern is read from one candle's shape over a fixed interval. A price zone is a band drawn from where a move began, usually spanning several candles. The zone is the more discretionary of the two, because its edges are chosen rather than given.