Part 2: One- and two-candle formations · Chapter 3 of 14 · 9 min
Dojis
The doji (open and close almost equal) after a rise or a fall appears 151 times on H1 gold, as often as in chance (158). Price reverses in 34% of cases and continues the trend in 50%; the trades give −0.19 ± 0.17 R and the confirmation of the next candle does not change the picture (0.00 ± 0.31 R).
The rules
A doji is a candle where open and close are almost equal. The rules of the book, in ATR (H1):
- Standard doji: body of at most 10% of the range and range of at least 0.7 ATR, after a trend (change of the closes of at least 1.5 ATR in the previous 10 candles). The direction the theory expects is reversal: after a rise, down; after a fall, up.
- Dragonfly: small body and lower wick of at least 60% of the range, upper wick at most 10%. Expected direction: up.
- Gravestone: upper wick of at least 60% of the range, lower at most 10%. Expected direction: down.
- Spinning top: body between 10% and 25% of the range, with both wicks of at least 30%, after a trend. Expected direction: reversal.
The trade is that of the book: entry at the close of the doji, stop beyond its extreme plus 0.25 ATR (risk 0.5-4 ATR), target equal to the risk, 24 candles, cost 0.30 $. A variant with confirmation enters only if the next candle closes beyond the extreme of the doji in the expected direction, at that close.
How many there are and what follows
| Formation | Found (chance, 5th-95th) | In expected direction | Trades | Mean R | Random (median) |
|---|---|---|---|---|---|
| Standard doji | 151 (158, 147-173) | 34% | 132 | −0.19 ± 0.17 | −0.10 |
| Dragonfly | 29 (24, 20-29) | 38% | 28 | −0.19 ± 0.37 | −0.07 |
| Gravestone | 26 (21, 17-24) | 54% | 26 | +0.21 ± 0.38 | +0.09 |
| Spinning top | 94 (101, 90-114) | 37% | 93 | −0.08 ± 0.20 | −0.06 |
| All dojis | 206 (204, 190-220) | 37% | 186 | −0.14 ± 0.14 | −0.07 |
The control is 37% up, 36% down, 27% no threshold at ±1.5 ATR in 10 candles. The standard doji appears as often as in chance. After a standard doji price reverses (expected direction) in 34% of cases (±8 points), continues the trend in 50% and touches no threshold in 16%. The 132 trades give −0.19 ± 0.17 R (55 targets, 76 stops, 1 time exit): worse than 86% of the random series. The gravestone is the only one with a positive R (+0.21 ± 0.38, 26 trades), but it sits inside the range of chance (28% of the random series do at least as well). The spinning top reverses in 37% and continues the trend in 52%.
The shape of the doji (body ≤ 10%, range ≥ 0.7 ATR) appears 387 times: 177 after a trend and 210 without. With a trend, price reverses in 34% and continues in 50%; without a trend it goes up first in 40%, down in 35% (control 37%/36%). After the doji the range of the next three candles is a median 1.07 ATR against 0.92 ATR for any candle: the doji precedes slightly more movement, not a direction.
The confirmation
Theory says to wait for the next candle: if it closes beyond the extreme of the doji in the expected direction, the signal is valid. For the standard doji the confirmation arrives in 39 cases out of 151 and the 38 valid trades give 0.00 ± 0.31 R (17 targets, 18 stops, 3 time exits); the random series with the same rule give −0.07 R (from −0.29 to +0.22, 31 trades). For all dojis: 67 confirmations out of 206, 65 trades, 0.00 ± 0.24 R (random −0.03). The confirmation cuts the trades to a third and brings the result to zero, inside the range of chance. For the spinning top, 18 trades at −0.32 ± 0.40 R.
Four examples
Figure 1 shows a winning standard doji: on 3 January 2026 at 09:00, after a rise of 3.6 ATR, a candle with open 4,352.20 $, close 4,352.92 $, high 4,363.66 $ and low 4,347.89 $. Entry at 4,352.92 $ (downward), stop at 4,365.45 $ (12.53 $, 1.7 ATR), target at 4,340.39 $ reached at 13:00: +0.98 R. Figure 2 shows a doji that fails: on 29 December 2025 at 07:00, after a fall of 3.3 ATR (4,480.83 → 4,482.29 $, high 4,488.82 $, low 4,462.85 $); entry at 4,482.29 $ upward, stop at 4,458.74 $, target at 4,505.84 $; stop hit at 11:00: −1.01 R.
Figure 3 shows a dragonfly: on 16 July 2026 at 07:00 (open 4,033.16 $, close 4,033.62 $, low 4,026.13 $), after a fall of 2.5 ATR; entry at 4,033.62 $, stop at 4,023.34 $, target at 4,043.90 $ reached at 11:00: +0.97 R. Figure 4 shows a gravestone: on 22 July 2026 at 03:00 (open 4,134.48 $, close 4,134.24 $, high 4,145.34 $), after a rise of 4.5 ATR; entry at 4,134.24 $ downward, stop at 4,148.48 $, target at 4,120.00 $ reached at 06:00: +0.98 R. Four examples, two win and two lose, and they say nothing on their own.
Read the chart in words
H1 gold chart of 3 January 2026 with the numbered doji and the lines of entry, stop and target.
| When | What | Price |
|---|---|---|
| 3 Jan 2026 09:00 UTC | Point 1 | 4,363.66 |
| 3 Jan 2026 09:00 UTC | Entry | 4,352.92 |
| 3 Jan 2026 09:00 UTC | Stop | 4,365.45 |
| 3 Jan 2026 09:00 UTC | Target | 4,340.39 |
| 3 Jan 2026 13:00 UTC | target reached | 4,334.40 |
Times in UTC.
Read the chart in words
H1 gold chart of 28-29 December 2025 with the numbered doji and the lines of entry, stop and target.
| When | What | Price |
|---|---|---|
| 29 Dec 2025 07:00 UTC | Point 1 | 4,462.85 |
| 29 Dec 2025 07:00 UTC | Entry | 4,482.29 |
| 29 Dec 2025 07:00 UTC | Stop | 4,458.74 |
| 29 Dec 2025 07:00 UTC | Target | 4,505.84 |
| 29 Dec 2025 11:00 UTC | stop hit | 4,451.10 |
Times in UTC.
Read the chart in words
H1 gold chart of 15-16 July 2026 with the numbered dragonfly and the lines of entry, stop and target.
| When | What | Price |
|---|---|---|
| 16 Jul 2026 07:00 UTC | Point 1 | 4,026.13 |
| 16 Jul 2026 07:00 UTC | Entry | 4,033.62 |
| 16 Jul 2026 07:00 UTC | Stop | 4,023.34 |
| 16 Jul 2026 07:00 UTC | Target | 4,043.90 |
| 16 Jul 2026 11:00 UTC | target reached | 4,046.22 |
Times in UTC.
Read the chart in words
H1 gold chart of 21-22 July 2026 with the numbered gravestone and the lines of entry, stop and target.
| When | What | Price |
|---|---|---|
| 22 Jul 2026 03:00 UTC | Point 1 | 4,145.34 |
| 22 Jul 2026 03:00 UTC | Entry | 4,134.24 |
| 22 Jul 2026 03:00 UTC | Stop | 4,148.48 |
| 22 Jul 2026 03:00 UTC | Target | 4,120.00 |
| 22 Jul 2026 06:00 UTC | target reached | 4,116.46 |
Times in UTC.
Procedure in four steps
- Write the rule of the doji in ATR and the definition of trend before looking for it.
- Count how many you find and compare with a random series.
- Measure how many reverse and how many continue the trend, not only those that worked.
- If you add a confirmation, count how many trades it leaves and what the result is worth with the error.
Exercises
Chapter exercises
1. On the chart the letters A-D are on four candles of 16 July 2026. Which letter is the dragonfly (small body and lower wick of at least 60% of the range)?
Read the chart in words
| When | What | Price |
|---|---|---|
| 16 Jul 2026 03:00 UTC | Point A | 4,038.95 |
| 16 Jul 2026 05:00 UTC | Point B | 4,036.46 |
| 16 Jul 2026 07:00 UTC | Point C | 4,033.96 |
| 16 Jul 2026 10:00 UTC | Point D | 4,038.18 |
Times in UTC.
2. After a trend the standard doji reverses in 34% of cases and continues the trend in 50%. What does it mean?
3. The standard doji with confirmation gives 0.00 ± 0.31 R over 38 trades, without confirmation −0.19 ± 0.17 R over 132. What do you conclude?
4. A doji appears 151 times on H1 gold and a random series has 158. What does it mean?
Have a question or want to share your exercise?
Post in the community, or join the free signals room on Telegram.
Educational content, not financial advice. Trading involves risk.