Candles compress open, high, low and close into one shape. Patterns describe how that auction unfolded; they do not predict the next move by themselves. This guide shows the geometry, the context that matters and a live, reproducible crypto sample.
First: what one candle contains
Open
First traded price in the interval.
High
Highest traded price.
Low
Lowest traded price.
Close
Last traded price when the interval ends.
The body connects open and close. Wicks show the extremes rejected or explored around it. A green body only says close > open; red says close < open.
01—05
Beginner starter kit
Use this sequence every time. It separates observation from execution and prevents a familiar shape from turning into an automatic trade.
01
Start on the higher timeframe
Mark trend, range boundaries and the nearest obvious support/resistance before zooming in.
02
Wait for the candle to close
A hammer can become a full bearish candle before the interval ends. Read only completed candles.
03
Name the pattern precisely
Use an explicit geometry definition, not a visual guess. The live lab publishes the rule it tests.
04
Demand confirmation
Define what price must do next: reclaim a body, break a range, or hold a level.
05
Write invalidation first
Choose the price that proves the idea wrong, then check fees, stop distance and position size.
Live OHLC lab
Test the shape against the market in front of you
The chart uses completed Binance candles from QuantRhino's own datafeed. Probabilities are recomputed in your browser for this pair, timeframe and sample — they are not universal success rates.
Loading completed candles…
Pattern field guide
Rejection · 1 candles
Hammer
A long lower wick shows rejection below the body. It is a possible bullish reversal, not a buy order.
Geometry used by this lab
One candle; body ≤35% of range, lower wick ≥2× body, small upper wick.
Useful when
After a decline, near support, with the next candle confirming above the hammer body.
Invalid / weak when
Price accepts below the hammer low or the shape appears mid-range without prior selling.
Rejection · 1 candles
Shooting star
A long upper wick shows rejection above the body. It can warn that buyers lost control.
Geometry used by this lab
One candle; body ≤35% of range, upper wick ≥2× body, small lower wick.
Useful when
After an advance, near resistance, followed by a close back below the body.
Invalid / weak when
Price accepts above the wick high or the shape forms without a prior advance.
Two-candle reversal · 2 candles
Bullish engulfing
A bullish real body covers the previous bearish real body; control shifted during the second candle.
Geometry used by this lab
Bearish candle followed by bullish candle whose body contains the previous body.
Useful when
After a pullback or sweep, with expanding range/volume and room to the next resistance.
Invalid / weak when
The second body is tiny, appears into resistance, or loses the engulfing low.
Two-candle reversal · 2 candles
Bearish engulfing
A bearish real body covers the previous bullish real body; sellers controlled the second candle.
Geometry used by this lab
Bullish candle followed by bearish candle whose body contains the previous body.
Useful when
After an advance or liquidity sweep, especially near resistance.
Invalid / weak when
The engulfing high is reclaimed or the pattern appears directly into support.
Indecision · 1 candles
Doji
Open and close are nearly equal. It says balance/indecision — direction must come from context and the next candle.
Geometry used by this lab
One candle; real body ≤10% of the full high-low range.
Useful when
At an important level after expansion, followed by a decisive break of its range.
Invalid / weak when
In low-volatility chop; doji are common there and carry little information alone.
Three-candle reversal · 3 candles
Morning star
Strong selling, then hesitation, then a bullish recovery through at least half of the first body.
Geometry used by this lab
Large bearish body, small middle body, bullish close above the midpoint of candle one.
Useful when
After a decline, ideally at support with the third candle closing firmly.
Invalid / weak when
The third candle cannot recover the first midpoint or the pattern forms inside congestion.
Three-candle reversal · 3 candles
Evening star
Strong buying, then hesitation, then a bearish recovery through at least half of the first body.
Geometry used by this lab
Large bullish body, small middle body, bearish close below the midpoint of candle one.
Useful when
After an advance, ideally at resistance with broad confirmation.
Invalid / weak when
The third candle cannot break the first midpoint or price immediately reclaims the high.
Compression reversal · 2 candles
Bullish harami
A small bullish body sits inside the prior bearish body. Selling pressure compressed, but has not yet reversed.
Geometry used by this lab
Large bearish body followed by a smaller bullish body fully inside it.
Useful when
After a decline, with a later break above the mother candle midpoint/high.
Invalid / weak when
No upside confirmation, or the mother-candle low is broken.
A practical reading order
1 · Location
Is the pattern at support/resistance, after a sweep, or in the middle of noise?
2 · Regime
Reversal shapes fight strong trends. In a range, the same shape can mean something else.
3 · Confirmation
Wait for a completed candle and define the price that invalidates the idea.
4 · Risk
A pattern is not a position size. Fees, slippage and the stop distance decide whether the trade is viable.
Do
✓Compare the pattern rate with the market base rate.
✓Prefer patterns at a meaningful level after a clear move.
✓Record sample size and uncertainty, not only win rate.
✓Test each market and timeframe separately.
Avoid
×Trading a pattern while its final candle is still open.
×Treating a wick as proof of a reversal without confirmation.
×Using a tiny sample as a universal probability.
×Ignoring fees, slippage, liquidity and invalidation.
Small glossary
Bullish / bearish
Describes direction or pressure; it does not mean guaranteed profit/loss.
Rejection
Price traded into an area but closed away from the extreme, leaving a wick.
Confirmation
A later completed candle satisfies the condition chosen before entry.
Invalidation
The observable price condition that makes the trade idea wrong.
Base rate
How often the same outcome happened across all eligible candles, without the pattern.
Wilson interval
A range that communicates uncertainty around an observed proportion, especially when N is small.
What the research actually says
The literature is mixed, which is exactly why the lab above reports its own N, base rate and confidence interval. A 2026 large-scale crypto study found predictive information in some patterns; a 2021 crypto study found most tested patterns of little use; a robust DJIA study found no trading value after testing. Market, timeframe, definition and costs matter.
Educational research only. A descriptive probability from the latest 999 candles is not an out-of-sample strategy, and past behavior does not guarantee future results.