The short answer
Three white soldiers is a three-candle bullish reversal pattern, and it is one of the few candlestick setups with real statistical backing, but the backing is widely misread. The pattern itself is a sustained push of buying that often marks a genuine low, yet most traders lose with it because they treat the candle shape as the whole strategy instead of a signal that needs context.
The famous number is 82%. Thomas Bulkowski's encyclopedia-scale testing found three white soldiers breaks upward 82% of the time, which sounds like a money printer until you read what "breakout" means (thepatternsite).
It means price is more likely to close above the pattern than below it, not that buying the pattern makes you an automatic winner.
I want to separate what the 82% actually measures from the profitability most traders assume it promises, because that gap is where the money leaks out. The pattern is real, the statistics are real, and the misuse is what loses.
The wider context on candlestick charting is in the candlestick patterns hub, and this page covers one of the strongest reversal setups in the field.
What three white soldiers actually is
The anatomy is fixed and worth knowing precisely. Three white soldiers is three consecutive long-bodied bullish candles, each opening within the previous candle's body, and each closing higher than the last.
The opening-within-the-prior-body detail matters, because it shows buyers accepting higher prices each session rather than gap-and-fade. The progressive closes signal sustained demand, and the long bodies show conviction, which is why the pattern reads as a reversal rather than a random three-green run.
The ideal three white soldiers also has modest shadows, with each candle's wicks small relative to its body, because long wicks would show intraday rejection of the buying. Clean bodies with small wicks signal conviction the whole session, while tall wicks on otherwise-green candles dilute the pattern into something weaker.
The psychology behind the pattern is spent selling meeting fresh buying, which is why it reads as a reversal. Three sessions of progressive closes mean the sellers who controlled the decline have lost control, and the buyers who were waiting have stepped in, and that handoff is the signal the candles record.
The pattern is the bullish mirror of three black crows, the three-candle bearish reversal, and both belong to the Japanese candlestick body of work popularised in the West by Steve Nison in the early 1990s. The names and the structure come from classical candlestick analysis, not from modern internet trading lore.
I always check the three-candle structure before I call the pattern, because three random green candles that gap up and fade are not three white soldiers, and the difference is whether each session opened inside the prior body.
The 82% stat, and what it really means
Bulkowski's number is the most-cited figure on this pattern, and reading it correctly is the whole game. He found three white soldiers breaks upward 82% of the time, meaning a close above the top of the highest candle is more likely than a close below the lowest (thepatternsite).
What that does NOT mean is that buying the pattern wins 82% of trades, or that it returns a profit 82% of the time. Breakout direction is the raw probability of the next move, before spreads, before entries, before stops, and before the trader's behaviour, and all of those erode the headline number.
The honest translation is that the pattern tilts the odds upward, which is valuable, but it does not remove the need for an entry, a stop, and a target. A tilt is an edge, and an edge is not a guarantee, and most beginners read the 82% as the latter.
I treat the statistic as a reason to take the pattern seriously, not as a reason to abandon risk management, because the gap between an 82% breakout rate and a profitable trade is exactly where disciplined trading lives.
Bulkowski built the figure from thousands of historical examples across US equities, which is worth knowing because the sample is large but not forex-specific. The 82% direction rate is a robust empirical average, and it is reasonable to expect a similar tilt in liquid forex pairs, with the same context caveats applied.
The other drag on the headline rate is cost, because spreads, commissions, and slippage are not in the 82%. A pattern that breaks up 82% of the time can still be a net loser if the average win is small and the average loss is large, which is why the risk-to-reward framing matters more than the win rate most beginners fixate on.
Why context decides everything
If the 82% is the headline, context is the small print, and the small print is where the profit is. The same candlestick shape is far more reliable at a major support level with rising volume than it is in the middle of a range, because the level and the volume confirm the buying (IG).
Multi-candle patterns like three white soldiers are more reliable than single-candle ones in general, and higher timeframes produce fewer false signals than lower ones, per the same analysis. So a three-white-soldiers on the daily chart at a daily support is worth far more than one on the five-minute chart in no-man's land.
The most reliable version stacks the factors: the pattern sits at a major support level, forms on rising volume, and follows an extended down move where sellers are finally spent. Strip any of those and the pattern's edge shrinks, which is why the same shape pays in one place and fails in another.
I will not trade the pattern without at least a level and a timeframe behind it, because the candle shape on its own is a coin with a slight edge, and the context is what turns the slight edge into a real one.
How to trade it honestly
The honest trade is to wait for the pattern at a major support level, confirm it, and then enter with a defined stop and target. The entry typically comes on the close of the third candle or on a small pullback the next session, not mid-formation on hope.
The stop goes below the low of the three-candle pattern, because a move beneath the lowest candle invalidates the reversal premise. Sizing that stop from your risk per trade is non-negotiable, and the method is covered in the guide to volatility-based position sizing.
The target is the next resistance level above, which gives the trade its reward relative to the risk. A three-white-soldiers reversal off a daily support can run for hundreds of pips if the level holds, and the reward is the distance to the next ceiling, not a fixed pip count.
I enter on confirmation, never on the second candle mid-formation, because a pattern that has not finished is not yet a pattern, and front-running it turns a probability into a guess.
A more conservative entry is the pullback after the third candle, which waits for a small retracement toward the middle of the pattern before buying. The pullback entry pays a slightly worse price for a much better location, and it lets the impulsive buyers reveal whether the reversal has real follow-through.
When three white soldiers fails
The pattern fails in specific conditions, and naming them prevents the larger losses. The first is the ranging or choppy market, where three green candles are just noise inside a band and the so-called reversal runs straight into the top of the range.
The second is the overbought extension, where three white soldiers appear after a long advance rather than at the end of a decline. A bullish reversal pattern with no decline to reverse is a continuation, and buying it is buying the top of an extended move.
The third is the missing-volume pattern, where the three candles form on falling interest and the buying has no force behind it. Volume is the fuel, and a reversal printed on declining volume is a reversal nobody is actually committing to.
I skip the pattern in ranges, after extended up-moves, and on low volume, because those are the three contexts where the 82% tilt reverses, and trading into them is how a good pattern becomes a bad loss.
A specific failure to watch is the dark fourth candle, where the session after the pattern opens and reverses hard. The pattern is only confirmed by what follows it, and a single bearish engulfing candle right after three white soldiers erases the setup and often marks the real top.
Three white soldiers vs three black crows
The bearish mirror is worth a paragraph, because the two come as a pair and most traders learn them together. Three black crows is three consecutive long-bodied bearish candles, each opening within the prior body and each closing lower, signalling a sustained push of selling.
The logic is identical, just inverted: three black crows at a major resistance, on volume, after an extended advance, is a real bearish reversal. The same shape in a range or on low volume is noise, exactly like its bullish twin.
Bulkowski's testing covers both, and the reliability factors transfer, so if you understand the soldiers you understand the crows. The mirror is not a different strategy, it is the same strategy seen from the short side.
I treat the two patterns as one concept applied in both directions, because the skill of reading sustained conviction in three candles is the same whether the conviction is buying or selling.
Common mistakes, and the fixes
The errors that lose money on this pattern are predictable, and naming them is most of the defence. The first is trading the candle shape alone, ignoring level, timeframe, and volume, which turns a context-dependent signal into a blind bet.
The second is entering before the third candle closes, on the assumption the pattern will complete. An incomplete pattern is not a pattern, and the front-run gets stopped when the would-be third candle reverses into a bearish one.
The third is overtrading, taking every three-green run as three white soldiers. Three random green candles that gap and fade do not qualify, and overcalling the pattern fills the account with low-quality signals that drain by a thousand cuts.
The fourth is ignoring the broader trend, buying a bullish reversal inside a relentless downtrend where the pattern gets swallowed. A reversal pattern needs a decline to reverse, and trading it against an unstoppable trend is a known way to lose.
I made the first and third mistakes early, calling the pattern anywhere I saw three green candles, and the fix was a checklist: real decline, major level, rising volume, confirmed close. The checklist is boring, and it is the entire edge.