Home › Guides › Reading charts
Do candlestick patterns actually work?
Candlestick patterns describe a fight inside a bar or two. They can help you notice a change in price behavior, but a shape alone cannot tell you what happens next.
The short answer
A candle pattern is a clue, not a forecast. Its meaning depends on where it appears, what came before it and what you count as a result. You can test whether a pattern helps your own decisions, but you cannot infer a reliable win rate from seeing a few good examples on a chart.
A hammer after a fall at an old support zone is different from a hammer in the middle of a quiet range. Both have the same shape. They do not have the same surroundings.
Learn the shape, then ask the harder question
A candlestick shows an open, high, low and close. The body shows the distance between open and close. The wicks show where price traded but did not finish.
A hammer has a small body near the top and a long lower wick. An engulfing pattern has a larger candle body that wraps around the previous body. An inside bar fits within the prior candle's full range. A doji opens and closes at nearly the same price.
The four-pattern guide shows those shapes. This page asks a different question: when the shape appears, does it improve a decision?
Same candle, two places
Picture two made-up charts that both show a hammer. On Chart A, price has fallen into a zone where it previously turned up twice. On Chart B, price is halfway between the last swing high and low, with no nearby level.
On Chart A, the hammer tells you sellers pushed through a known area but could not hold the low by the close. It may be worth watching the next candle. On Chart B, the same wick may simply be another random move inside a range.
Neither chart promises a bounce. The first has a reason you can state and test; the second has only a shape. The support and resistance guide explains why location matters.
Define the test before you count wins
Pick one candle pattern and one market and time frame. Do not mix hammers, dojis and engulfing bars into the same score. Write the rules before opening replay.
- Appearance. What exact size and wick rule makes a candle count? How do you handle equal prices?
- Location. Is the candle at a level you marked before seeing it, or away from one?
- Entry point. Do you act at the close or wait for the next candle? Pick one.
- Outcome. For a drill, you might count whether price moves one full candle range in the pattern's direction within the next three candles before moving that far against it.
- Every sighting. Log all qualifying patterns, not only the ones that worked.
That outcome is a study definition, not an implied trading return. It ignores spreads and account risk. If you later test a real strategy, include costs, stops and fills.
A filled-in practice worksheet
Here is an invented twenty-sighting example. It is not backtest data. Ten patterns appeared at a pre-marked level and ten away from one.
| Group | Sightings | Met the drill outcome | Did not meet it |
|---|---|---|---|
| At a pre-marked level | 10 | 6 | 4 |
| Away from a level | 10 | 3 | 7 |
| Total | 20 | 9 | 11 |
Six of ten is 60% in this made-up worksheet. Three of ten is 30%. The totals make the two groups easy to compare, but ten observations per group are far too few to call a reliable edge. Different markets or different time periods can change the picture.
Now copy the empty version:
| Candle date | Pattern and rule | Location pre-marked? | Outcome met? | What I missed |
|---|---|---|---|---|
| ____ | ____ | Yes / No | Yes / No | ____ |
| ____ | ____ | Yes / No | Yes / No | ____ |
| ____ | ____ | Yes / No | Yes / No | ____ |
Record losing and failed cases with the same care as successes. If you already know the next three candles, go to another historical period you do not remember.
The three traps
Hindsight levels. Drawing a level after a bounce makes every pattern look like a clue. Freeze the levels before the candle prints.
Changing the outcome. If one pattern moves up for one bar and another for five, calling both wins without a fixed rule breaks your sample. One definition, every row.
Ignoring the bad fills. A body that closes at $50 does not mean you can always enter at exactly $50. The bid, ask, spread and any gap change an order's result. That is why a candle worksheet is not a live trading performance report.
You can practice with chart replay and keep the results in a journal.
How this fits into a decision process
Treat the candle as one piece of a larger setup: bigger trend, level, trigger, invalidation price and size. If the candle is not at an area you had a reason to watch, there may be no trade at all.
StructureX Foundations uses simulated practice charts to train decisions. Those charts are not historical market data and cannot be used as proof of a pattern's real-market win rate. For the test above, use actual historical bars you can replay without looking ahead.
The goal is to stop collecting pattern names and start making testable decisions.
Questions
Which candlestick pattern has the highest win rate?
There is no universal answer. The outcome depends on the market, timeframe, entry and exit rules, costs and where the shape appears. A named shape alone has no fixed win rate.
Does a hammer always mean price will rise?
No. It shows that price traded lower and recovered by the close. It does not include future price data.
Why does a pattern work on one chart and fail on another?
Context differs. A pattern at a pre-marked level during a trend is not the same case as the same shape in a choppy range.
How many patterns should I study first?
One or two are enough to start. Define them well and log every sighting before adding more names.
Can I test patterns without trading money?
Yes. Use historical replay and a worksheet. Keep the test outcome separate from any claim about live fills or profit.
Follow along