Some candles record a decision. A long-legged doji records a session in which a decision was attempted in both directions and refused in both. Price ran well above the open, then well below it, and finished where it started. It is one of the most dramatic-looking candles on a chart and, measured on its own, one of the least directional.
A long-legged doji is a candle whose open and close are at or near the same price, with long upper and lower wicks: the session ranged widely in both directions and ended unchanged. It records indecision, not direction. It means something only at a location, such as after an extended move, and the next close beyond its range resolves it.
That makes it the clearest member of a family of indecision candles: the doji, the long-legged doji, the spinning top, and their directional cousins, the gravestone doji and the dragonfly doji. This post sorts them by the one measurement that separates them, and covers what each needs before it means anything.
What is a long legged doji, in measurable terms?
Every single candle can be described by its body-to-range ratio: how much of the session’s full high-to-low range was covered by the move from open to close. The how to read a candlestick post covers why that one number turns a shape into a category.
A doji is a candle where that ratio is effectively zero: the open and close are within a few ticks of each other. Bulkowski’s identification guideline for the long-legged doji adds one condition: long shadows. His worked example adds that the shadows need not be the same length, only longer than recent shadows on other candles. When the open and close sit near the middle of an unusually tall range, some sources call it a rickshaw man, which Bulkowski treats as its own entry.
The drawing is the standard doji. The long-legged version is the same shape stretched: a range noticeably wider than the candles around it, with the body still a thin line.
What does the long-legged doji actually record?
Both sides had control at some point in the session, and both lost it. Buyers pushed price well above the open and could not hold it. Sellers pushed it well below and could not hold that either. The close back at the open says the session ended with neither side able to claim the move.
That is real information, but it is information about the session, not about the next one. Bulkowski’s tests make the point bluntly. Every percentage in this post comes from his candlestick tests, and each page describes its figures as based on hundreds of trades without stating the market or period. On those terms, the long-legged doji acts as a bullish continuation 51% of the time, which he describes as random, and ranks 37th of 103 candle patterns for overall performance. His rickshaw man reports a continuation rate of 51%, and his high wave candle, the same idea with a small body instead of none, a reversal rate of 51%. The indecision label is accurate. The direction is a coin toss on the candle alone.
When does indecision mean something?
At a location. An indecision candle in the middle of a sideways range is one more indecisive session in a market that has been indecisive for weeks. The same candle in two other places is worth noticing:
- After an extended move. A long-legged doji after a long advance records the first session in which buyers pushed higher and could not keep it. That is not a reversal, but it is a change in behaviour from the sessions before it.
- At a tested level. At a support or resistance level that has held before, a session that probed through the level in both directions and closed back at the open records the level being contested.
Bulkowski adds a finding that fits the mechanism: long-legged doji candles with shadows taller than the median moved 50% further after the breakout than those with shorter shadows. A wider contest seems to be followed by a wider resolution, though not in a direction the candle tells you.
Confirmation and invalidation for an indecision candle
Indecision cannot be wrong on its own terms. What can be wrong is the read you attach to it, and that read can be stated and falsified:
- Needs: a location, either an extended prior move or a tested level. Without one, there is no read to make.
- Resolved by: the next close beyond the doji’s high or low. That close, not the doji, supplies the direction. The confirmation candle post covers why a close and not a wick is the standard.
- Invalidated by: that resolving close landing on the side of the prior move. After an advance, the read is “the advance has stalled”: a close below the doji’s low supports it, and a close above the doji’s high ends it, because the advance resumed.
The rickshaw man and the high wave candle are read the same way: same location requirement, same resolving close, same invalidation.
How does a spinning top differ from a doji?
By the body. A spinning top has a small but visible body with long wicks on both sides: there was a real difference between open and close, just a small one relative to the range. A doji has effectively no body at all.
In practice, the distinction matters less than people think. Bulkowski’s white spinning top page reports a reversal rate of 50%, right alongside the doji family’s figures. All of them record a contested session. What separates a useful one from a useless one is where it formed, not the exact height of its body.
- Needs: a location, as above.
- Resolved by: a close beyond the spinning top’s range.
- Invalidated by: a close beyond its range in the direction of the prior move.
The directional cousins: gravestone and dragonfly
Two doji variants do carry a lean, because their wicks are on one side only.
The gravestone doji opens and closes at or near the low of its range, with a long upper wick: the session rallied and gave all of it back. After an advance, the traditional reading is a rejection of higher prices.
- Needs: an advance or a level price has failed at before.
- Confirmed by: the next candle closing below the doji’s low.
- Invalidated by: a close above the doji’s high.
The dragonfly doji is the mirror: open and close at or near the high, with a long lower wick, after a decline.
- Needs: a decline or a level that has held before.
- Confirmed by: the next candle closing above the doji’s high.
- Invalidated by: a close below the doji’s low.
Even here, the measured lean is slight or absent. Bulkowski reports the gravestone doji acting as a bearish reversal 51% of the time and the dragonfly doji as a reversal 50% of the time in his tests. The one-sided wick is a more specific record of what happened inside the session, and it still needs the next close.
Where indecision candles mislead
Thin markets. A doji on a low-volume instrument, or in the quiet middle of a trading day, can simply mean little traded. That is an absence of participants, not a standoff between them.
Holiday and half-day sessions. A shortened session produces small, doji-like candles for reasons that have nothing to do with indecision.
Reading the doji as the reversal. The most common error. The doji records a stall; the next candle records which way the stall resolved. The three-candle patterns post covers the morning and evening stars, where a small middle candle is followed by a third candle that does the resolving, which is the doji idea with the confirmation built in.
What this post is not
It is not a list of turning points. It is a way to describe an indecision candle accurately: a contested session, worth noticing only at a location, resolved by the next close, and invalidated by the prior move resuming. If you can name all four for the candle on your chart, you have read it. For the rest of the single-candle shapes with their context and invalidation, see the candlestick pattern library.
Sources
Every figure above was checked against these before publishing.
- Bulkowski, Long Legged Doji
- Bulkowski, Rickshaw Man
- Bulkowski, High Wave
- Bulkowski, White Spinning Top
- Bulkowski, Gravestone Doji
- Bulkowski, Dragonfly Doji