Candlestick Patterns: How to Read Candles Without Guessing the Market

What does a single Japanese candle show?

Official materials from CME Group and TradingView describe the same basics: each regular candle contains open, high, low, and close. If the close is above the open, the body is usually colored as rising. If the close is below, it usually appears as falling. The color can be changed in the settings, so the position of the prices is more important than the usual palette.

A long body indicates a noticeable movement from open to close. A long upper shadow shows that the price went higher but did not retain the entire gain by the end of the period. A long lower shadow reflects a downward movement followed by a return. A small body indicates that the open and close are close to each other, although the range within the candle could have been wide.

One hourly candle represents one hour, a daily candle represents the trading day of the selected data source. On a 24-hour market, the day boundaries depend on the time zone and the quote provider. Therefore, the same asset on different charts may show a slightly different shape.

Candlestick patterns: how to read them in context?

The working sequence does not start with the name. Mark the trend or range, the nearest support and resistance levels, the usual candle range, and the position of the current price. After that, the shape gains meaning.

For example, a long lower shadow after a sharp decline shows that sellers pushed the price down, but part of the movement was bought back. The same shadow inside a narrow sideways movement can be ordinary noise. A price return does not yet prove the start of growth: the next closed candle can cancel the reaction.

A convenient reading formula consists of five questions:

1. Where did the pattern appear: at a level, after a trend, or in the middle of a range?
2. What was happening before it: acceleration, slowdown, or sideways movement?
3. Did the candle close and did the shape maintain?
4. Did the next price confirm the expected direction?
5. Where does the scenario stop being valid?

This order reduces the desire to fit the chart to a familiar name.

What do the hammer and shooting star mean?

A hammer has a small body located closer to the upper part of the range, and the lower shadow is noticeably longer. After a decline, such a candle marks an episode when the price went lower and then returned. The name does not indicate whether the next test of the level will hold.

A shooting star is the mirror image: the small body is closer to the lower part, with a long shadow remaining on top. After a rise, the pattern shows that the upward movement did not hold until the close. If the next candle again closes above the high, the bearish hypothesis loses its basis.

The study by Jamaloodeen, Heinz, and Pollacia tested the Hammer and Shooting Star over more than sixty years of S&P 500 data. The result depended on which future price was used to measure continuation: reliability based on closes was weak, while based on extremes it looked different. This is a good example of why the rule needs to be defined in advance.

How to read the engulfing and doji?

In a bullish engulfing, the body of the second candle covers the body of the previous falling candle. The bearish version is arranged in a mirror way. This is a description of the opening and closing prices, not a mandatory coverage of the shadows. Different sources use different quantitative thresholds, so automatic scanners may mark a different set of cases.

In a doji, the opening and closing are close to each other. Wide shadows with a small body indicate struggle within the period, but the final price has almost returned to the start. After a strong movement, this is a reason to watch for a reaction. In a calm range, a doji occurs often and may change nothing.

If you want to match the candlestick reaction with trend, levels, volume, and momentum, you can open a free 7-day Midas demo.

Why can't an unclosed candle be considered a finished pattern?

The highest and lowest prices may already be preserved, while the close is formed up to the last trade of the period. A change in the close alters the size of the body, its color, and the ratio with the shadows. In a two-candle pattern, the current candle may also stop absorbing the previous one.

For discipline, note two times: the moment of candidate detection and the moment of actual closing. If an early scenario is needed, describe it separately and do not mix statistics with rules that wait for the closed candle.

On lower timeframes there are more patterns, but the share of market noise is usually higher. Compare the same pattern only at the same interval and data source.

What confirmations help weed out random shapes?

The first filter is the level. If the market has already reacted in this area, a new candle has a clear point of reference. The second filter is the next price: breakout of the local structure, retention of the retest, or return to the range. The third is participation, which can be assessed using a volume indicator.

It is convenient to compare the current amplitude with ATR. A shadow that looks large can be normal for this volatility mode. The moving average helps indicate the direction, but it lags and does not replace the structure of highs and lows.

Record the cancellation condition in advance. For a bullish hypothesis, this could be a close below the pattern's minimum. For a bearish one – a fixation above the maximum. The cancellation point is needed before entry, otherwise any unsuccessful pattern can easily be declared 'not yet played out'.

How does a cryptocurrency chart differ from a stock session?

Some classic descriptions of the morning or evening star assume a gap between candles. In a market that trades without a daily close, full gaps occur less frequently. The sequence of bodies can still indicate a change in momentum, but the pattern's criteria need to be adapted and honestly recorded.

Liquidity of exchanges, participant composition, and the cutoff hour of the daily candle also differ. Verification on one data source does not guarantee the same pattern on another. If the system is executed on a specific exchange, history and signals should be evaluated using its data.

How to test a pattern on historical data without overfitting?

The study by Chen, Bao, and Zhou of eight two-day patterns on the Chinese stock market showed inconsistent results: some patterns had predictive power, some gave contradictory conclusions, and the effect decreased as the horizon increased. There is no general answer "candles work" for all definitions.

The minimal log looks like this:

Order of verification:

1. Formulate numerical criteria before viewing the result.
2. Choose one market, source and timeframe.
3. Collect consecutive cases, not skipping inconvenient ones.
4. Separate the data for training from the data for testing.
5. Calculate the result after costs and compare it with a simple baseline scenario.

Five nice examples only confirm the ability to find images retrospectively.

What mistakes most often interfere with reading candlesticks?

1. The title replaces the question. The trader sees a hammer and immediately looks to buy, without checking the preceding decline and the level.
2. The shape is still changing. The decision is made before the candlestick closes.
3. Any shadow is considered significant. The size is not compared to the usual market volatility.
4. The data source is ignored. The pattern is searched for on one chart, and execution is evaluated on another.
5. Confirmation is selected retrospectively. After each episode, the rule changes.
6. Only successful screenshots are kept. Missed cases disappear from evaluation.
7. There is no cancellation point. A losing hypothesis remains open without measurable justification.

Risks and limitations

CME Group directly separates objective OHLC data from the subjectivity of chart pattern analysis. Results are also influenced by liquidity, news, volatility and execution costs. Automatic recognition saves time but inherits thresholds set by the developer.

The market may break the hammer's low or continue rising after a shooting star. Even a confirmed hypothesis sometimes ends in loss. Position size and acceptable risk are determined before entry, not after an adverse move.

The material is for informational purposes and is not individual investment advice. Past results do not guarantee future outcomes.

Sources

• CME Group: Chart Types - Candlestick, Line, Bar
• CME Group: Technical Analysis
• TradingView: Chart types available on TradingView
• TradingView: Automatic candlestick pattern detection
• Jamaloodeen, Heinz, Pollacia: A Statistical Analysis Of The Predictive Power Of Japanese Candlesticks
• Chen, Bao, Zhou: The predictive power of Japanese candlestick charting in Chinese stock market
• Updated 12.08.2026 · verifiable context rules, confirmations, and backtesting added.

Frequently Asked Questions

Which candlestick patterns are best for a beginner to study?

Start with the mechanics of a single candle: body, upper and lower shadows, open and close. Then move on to the hammer, shooting star, engulfing, and doji. These patterns are enough to master the main habit: linking the pattern to the trend, level, closing price, and reversal point. A large dictionary of names without practical verification rules does not add value.

On which timeframe do candlestick patterns work best?

There is no universal timeframe. On short intervals, there are more cases, but the influence of noise and costs is higher. On higher timeframes, there are fewer candles, but each includes more trades and time. Choose an interval according to your horizon, use one higher timeframe for context, and do not combine results from different scales into one statistic.

Do you need to wait for the next candle after a pattern?

If the rule requires confirmation, it is waited for after the full closure of the pattern. Confirmation can be a break above the maximum, a return above the level, or a change in local structure. Early entry is only possible as a separate, pre-defined strategy with its own risk. Early and confirmed entries cannot be evaluated as a single rule.

Is it possible to find candlestick patterns automatically?

Yes. TradingView offers indicators for automatic recognition of candlestick patterns and allows setting up alerts. The algorithm merely applies the predefined criteria and does not evaluate the entire context the way a human does. Before using it, check the definitions, timeframe, data source, and the behavior of the labels on an unclosed candle.

Why does the same candle give different results?

The shape shows only four prices within the period. Before it, the market could have been rising, falling, or moving sideways; a strong zone or an empty area of the chart could be nearby. Volatility, volume, and liquidity differ. Therefore, the value arises from the combination of shape and context, and the outcome of a single trade remains uncertain.

Is it possible to combine candlestick patterns with Midas?

Yes. The candlestick reaction can be used as a price trigger, and Midas - to check the trend, levels, volume, momentum, and money flow. The free 7-day demo includes Midas Up and Midas Down. This combination does not replace your own risk assessment and does not turn the signal into a guaranteed result.