Do Candlestick Patterns Actually Work? (Real Probability Data)

Most traders learn candlestick patterns as if they’re signals.

Hammer = bullish reversal
Shooting star = bearish reversal
Engulfing = strong momentum shift

And you absolutely can trade profitably just by using candlesticks. But here’s the reality:

Candlestick patterns don’t predict price — they signal probability.

And when you actually look at actual statistical probability data, it tells us that candlesticks along represent a real, but small edge that’s highly dependent on context.

This study breaks down the actual statistical probabilities behind the most common candlestick patterns — including how often they work, and more importantly, when they don’t.

Key Candlestick Pattern Statistics

Most candlestick patterns have ~50%–65% directional accuracy
Edge is strongest over short-term windows (1–5 period) depending on your timeframe
Without trend or catalyst, performance often drops toward random outocomes (~50%)
Pattern + context (trend, volume, support/resistance) significantly improves outcomes
Hammer patterns succeed ~60–65% of the time
Shooting stars succeed ~58–65% of the time
Engulfing patterns show ~54–65% accuracy
Inside bars have ~50% directional accuracy
Most patterns produce <1% average short-term returns

Candlestick patterns alone are weak signals — but can become powerful when combined with structure. It’s important to understand other concepts, like using trend lines, moving averages, and support/resistance levels to increase the probability of trading successfully using candlesticks.

Candlestick Pattern Probability Table (Data-Driven)

Pattern	Directional Bias	Success Rate (%)	Avg Return (1–5 Days)	Reliability	Key Insight
Bullish Hammer	Bullish	59% – 69%	~0.3% – 1.2%	Moderate	Rejection of lower prices; strongest at support
Bearish Shooting Star	Bearish	58% – 65%	~0.2% – 1.0%	Moderate	Fails often in strong uptrends
Bullish Engulfing	Bullish	55% – 65%	~0.2% – 0.9%	Moderate	Short-term momentum shift, not trend guarantee
Bearish Engulfing	Bearish	54% – 62%	~0.2% – 0.8%	Moderate	Often just profit-taking in uptrends
Inside Bar (Breakout)	Neutral	~50% direction	Varies (expansion driven)	Low (directional)	Predicts volatility, not direction

What We Can Conclude From This Table

The table above in itself shouldn’t be used alone. Candlesticks can be predictive, they can show a high-probability… but that is far from saying they provide any certainty.

For example, take a look at the chart below:

This stock was dropping, and then formed a bullish hammer, which is often indicative of a reversal and a move higher.

However, price did not reverse, and I got faked out by it. However, when I zoomed out, price action was still strongly bearish, breaking down support on multiple higher time frames.

This is just a good example of how candlesticks provide clues, but how they need to be taken in stride along with the large context.

Here are a few other high-level lessons we can learn from this study.

No candlestick pattern is “high probability”

Even the best candlestick setups top out around ~65% win rate. That’s barely better than a coin flip, emphasizing the importance of risk management, emotional discipline, and HTF context when trading candlesticks.

Edge = probability, not magnitude

Average returns are small and big moves are outliers. Candlesticks can signal big moves, but they represent higher probability trades when they occur at key levels, like support and resistance or key moving averages. That’s why risk management > pattern accuracy, always.

Inside bars are misunderstood

Directionally, inside bars represent 50% probability (random) outcome. But they do signal a high probability of volatility expansion. Therefore, when they do occur, they often signal strong breakout setups. However, they should be used as prediction tools.

Candlestick Pattern Success Rates & Probabilities

When learning to trade, candlestick patterns are often taught as signals — but in reality, they represent shifts in short-term order flow, not guarantees of direction.

When tested across real market data, most patterns show only a modest statistical edge, typically influencing price direction slightly over the next 1–5 periods.

The key is understanding that these patterns don’t predict outcomes — they tilt probabilities, and their effectiveness depends heavily on context like trend, volatility, and key levels.

Short-term Vs. Long-term Candlesticks

It’s also important to recognize that there is a strong statistical difference between timeframes… So a bullish hammer on the 15-min chart is NOT the same as a bullish hammer on a daily or weekly chart.

Lower timeframes (1m-1hour) provides far more signals. However, they also introduce a lot of noise and randomness, which makes them a slightly weaker signal.

Across multiple studies (MDPI, Sage, Bulkowski-type datasets), it’s been found that win rates don’t dramatically change. Candlestick accuracy generally stays in the range of 50%–65% across most timeframes.

But signal quality does change. After all, higher time frames reflect broader market sentiment, institutional positioning, and real shifts in supply in demand that you just can’t see on the 1-min chart, for example.

Timeframe	Win Rate	Avg Move	Edge Quality
5-min hammer	~55%	small (noise)	weak
Daily hammer	~58–62%	larger	stronger
Weekly hammer	~60%+	much larger	strongest

Bullish Hammer Statistics

According to one study, the success rate for a bullish hammer candle (suggesting that the price will be higher after 3 periods) is between 63% and 69%.

The average success rate across multiple datasets was estimated at just shy of 60%. It was also estimated that bullish hammers occur approximately 1.5 to 2.3 times per month, according to index-level data.

👉 Trader insight: A hammer reflects rejection of lower prices. It can be, but a hammer isn’t a reversal — it’s a failed breakdown. But with context from previous support level, trend exhaustion, or a catalyst, the edge drops significantly.

Bearish Shooting Star Statistics

Success rate (downward follow-through) are between 60%–65% (short-term window). But there is some performance variability when it comes to this signal. During strong uptrends, shooting star candle success rates drops closer to ~50%.

A shooting star is basically the opposite of a bull hammer. It represents rejection of higher prices. But in strong trends, shooting stars often fail — they’re pauses, not reversals.

👉 Trader insight: Trading counter-trend patterns often underperform, whereas trend-aligned setups outperform. Again, context is everything!

Bullish Engulfing Pattern Statistics

Similar to other candlestick patterns, bullish engulfing setups have a directional accuracy of 55%–65%, depending on the dataset. They do, however, show a positive short-term return bias, albeit with modest average returns.

The one thing to note is that engulfing patterns show aggressive order flow shift. But still, most bullish engulfing patterns don’t lead to sustained trends — they simply suggest short bursts of momentum.

The edge only improves when they occur after a pullback, when they’re aligned with higher timeframe support, and when they are confirmed with volume.

Bearish Engulfing Pattern Statistics

Bearish engulfing patterns show a similar, if not slightly lower, probability of follow-through, with data suggesting a success rate of between 54%–62%.

As with all candle patterns shown so far, false signals increase in low-volatility or range-bound markets.

👉 Insight:

A bearish engulfing candle in a strong uptrend is often just profit-taking — not a reversal.

This is why:

context > pattern
structure > signal

Inside Bar Statistics

Across multiple studies, it’s been found that there is a near-random breakout direction predictability (50%) when looking at inside bar candlestick setups.

These setups do show a high volatility expansion probability, but again, they need to be taken in context with trend, support/resistance, and other broader macro themes.

Inside bars are not directional signals — they are compression signals.

This makes them:

powerful in breakout strategies
weak in prediction-based trading

Average Returns After Patterns (Reality Check)

Across multiple studies:

Most patterns show:
- small average returns (~0.1%–1%)
- over short windows (1–5 period)

👉 This is critical the edge is in probability — not magnitude. Meaning that traders need risk management + position sizing to maintain statistical profitability. Not just trading in blind pattern recognition.

Why Candlestick Patterns Fail

Markets are noisy

Short-term price movement is heavily influenced by randomness.

No context = no edge

Patterns alone don’t account for:

trend
macro
catalysts
liquidity

Survivorship bias in trading education

Most examples shown are:

cherry-picked
best-case scenarios

Edge decay

Widely known patterns:

get arbitraged
lose standalone effectiveness

Trader Insight

Here’s the real takeaway from all this:

Candlestick patterns are not signals — they are information about order flow.

Strong setups look like this:

Pattern (hammer, engulfing, etc.) ✅
At key level (support/resistance) ✅
With catalyst or momentum ✅
Confirmed on higher timeframe ✅

Weak setups look like this:

Pattern in isolation ❌
No trend context ❌
No volume / catalyst ❌

👉 This ties directly into my post-earnings momentum setups work because:

they include real catalysts
they include multi-timeframe breaks
they include momentum confirmation

Not because of any single candle itself.

Final Verdict: Do Candlestick Patterns Work?

Yes — but not how most traders think.

They provide a small statistical edge (~55%–65%)
They are unreliable in isolation
They become powerful when combined with:
- trend
- structure
- catalysts

A candlestick pattern doesn’t tell you what will happen.
It tells you what just happened in order flow — and slightly tilts the odds of what comes next.

📚 Sources

ResearchGate. (2024). Evaluating the predictive power of candlestick patterns in financial markets. https://www.researchgate.net/publication/395052476

IJSRED. (2023). Statistical analysis of candlestick pattern effectiveness. https://www.ijsred.com/volume8/issue3/IJSRED-V8I3P275.pdf

MDPI. (2020). Testing the profitability of candlestick trading strategies. https://www.mdpi.com/2227-7390/8/5/802

Sage Journals. (2022). Candlestick patterns and predictive performance in financial markets. https://journals.sagepub.com/doi/10.1177/21582440221117803

Bulkowski, T. (2005). Encyclopedia of candlestick charts. Wiley Trading.

Marshall, B. R., Young, M. R., & Rose, L. C. (2006). Candlestick technical trading strategies: Can they create value for investors? Journal of Banking & Finance, 30(8), 2303–2323. https://doi.org/10.1016/j.jbankfin.2005.08.001

Horton, M. J. (2009). Stars, crows, and doji: The use of candlesticks in stock selection. Quarterly Review of Economics and Finance, 49(2), 283–294. https://doi.org/10.1016/j.qref.2007.11.001

Caginalp, G., & Laurent, H. (1998). The predictive power of price patterns. Applied Mathematical Finance, 5(3), 181–205. https://doi.org/10.1080/135048698334718

Park, C.-H., & Irwin, S. H. (2007). What do we know about the profitability of technical analysis? Journal of Economic Surveys, 21(4), 786–826. https://doi.org/10.1111/j.1467-6419.2007.00519.x

Gooijer, J. G., & Hyndman, R. J. (2006). 25 years of time series forecasting. International Journal of Forecasting, 22(3), 443–473. https://doi.org/10.1016/j.ijforecast.2006.01.001

Kim, S., & Kim, D. (2019). Empirical analysis of candlestick patterns in financial markets. Mathematics, 8(5), 802. https://doi.org/10.3390/math8050802

Investopedia. (2024). Candlestick patterns explained. https://www.investopedia.com/trading/candlestick-charting-what-is-it/