Doji candle detector and classifier

A doji is the candle that says "neither side won." Open and close land at almost the same price — the body is a hairline — but the wicks tell you the bar's range. The shape carries information: the market tried to move, fought itself, and finished where it started. That's why doji candles get singled out from the regular noise of small-body bars. They mean something different.

The trouble is that "doji" gets used loosely. Any small-body candle gets called one in casual writing, and the distinct shapes — standard, long-legged, dragonfly, gravestone — get blurred together even though each one tells a specific story. Drop the four OHLC numbers from a candle into the box below and the tool tells you whether it qualifies as a doji at all, and if so which of the four subtypes it is, with the exact thresholds it used.

What the four subtypes actually mean

A doji's body is the giveaway — when open and close land within ~10% of the bar's full range, the candle says "we tried, it didn't stick, we ended where we started." That alone is just noise. The wicks are where the meaning lives. They show where price actually went during the bar before being pulled back to the close.

A standard doji has balanced wicks above and below a tiny body. It's the classical "cross" shape in the candlestick literature. The interpretation is straightforward indecision: roughly equal pressure from both sides, no resolution. Standard dojis at the top of a strong uptrend or the bottom of a strong downtrend are textbook reversal candidates. Mid-range, in a quiet session, they're just bars.

A long-legged doji is a standard doji on steroids — both wicks are long. The market ranged hard in both directions during the bar before tying out. This is heightened indecision, and it usually shows up after a strong directional move when the prevailing trend's traders are starting to take profit and counter-trend traders are finding it worth their while to fight back. Long-legged dojis tend to mark the moments where conviction shifts even when the trend doesn't immediately reverse.

A dragonfly doji has open, close, and high stacked at the top of the bar, with a long lower wick reaching down toward what would normally be a low. The narrative: sellers pushed price down sharply during the bar, but buyers absorbed it and dragged the close all the way back to the high. That's a buying response at scale. At a confirmed support level after a downtrend, dragonflies are one of the cleaner bullish reversal patterns. At intraday inflection points they're common, but the higher the timeframe, the more weight the pattern carries.

A gravestone doji is the mirror — open, close, and low at the bottom, long upper wick. Buyers ran price up, sellers slammed it back down to the open. At resistance after a rally, gravestones often print right before the turn. The pattern's name is morbidly literal: it tends to appear at the end of attempts to break through.

Why context matters more than the pattern

Doji candles get over-credited by retail traders because they're easy to spot, easy to name, and they appear constantly. On a typical liquid pair you'll see several daily dojis per month and dozens on the 1-hour. The pattern is information, not a signal. What separates a meaningful doji from chart-noise is location — the candle has to print at a level where reversal makes structural sense.

The cheap version of "context" is just trend direction: a dragonfly during a strong uptrend doesn't mean much (the move was already going up), but a dragonfly at the bottom of a multi-day downtrend at a prior support level is a different conversation. The same pattern at different locations carries entirely different weight. The magnetic levels post walks through the levels worth pre-marking — they're the candle's frame of reference.

The expensive version is volume and follow-through. A reversal doji on above-average volume that gets confirmed by the next bar moving in the new direction is the rare clean trigger. A reversal doji on dead volume that's followed by another doji is the market telling you the indecision is just lethargy, not a turn.

What the tool isn't doing

It isn't checking confirmation. The next bar's behaviour determines whether the doji actually marked a reversal, and there's no algorithmic substitute for waiting and seeing. It also isn't applying timeframe weighting — a daily doji on BTC counts for far more than a 5-minute one, and the tool doesn't know which timeframe you're feeding it. The classification math is the same regardless, but the actionability isn't.

The 10% body-to-range threshold is the most common one in the literature, but you'll find practitioners who use 5% (stricter, fewer matches) or 15% (looser, more matches). The tool sticks with 10% for the same reason most do — it catches the textbook patterns without flagging every small-body bar. If you want a stricter check, just compare the body % the tool prints against your own rule.

For a related candle-by-candle question — how to tell whether two adjacent bars form a real engulfing pattern — the engulfing pattern checker runs the same kind of structural test on pairs. And for the broader thread on why traders read more into chart shapes than the shapes deserve, the confirmation bias on charts post covers the cognitive pattern that makes pattern-recognition harder than it looks.

FAQ

Sibling tools

• Engulfing pattern checker — the two-candle structural test. Same OHLC math, different pattern.
• Position size calculator — once a doji at a meaningful level confirms the next bar, the question is how much to risk.
• R-multiple calculator — log the trade outcome in R-multiples once it closes; the unit that's actually comparable across setups.