Learn
Candlestick patterns, defined plainly
Every pattern in Nison's book: what it looks like, what it actually requires — including the confirmation rules most cheat-sheets drop — and what each one measured when we tested it across 2,022 trades. Including the ones that measured nothing.
First, how to read one
A candle encodes four numbers — open, high, low, close — in a shape you can read at a glance. That is its real value, and it is not in dispute. What is in dispute is whether the named combinations predict anything.
- The real body
- Open to close. Nison calls it "the essence of the price movement" — colour and height tell you who won the session and by how much.
- The shadows
- The wicks out to the high and low. A long shadow means price went there and was rejected.
- Why traders like them
- A long lower shadow at a level you already care about tells you sellers were absorbed there. That is real information — about that level.
Read the measured figures below with care. "Edge" is the pattern's average return per trade minus a random entry using identical exits on the same symbol. It is the only number that separates signal from market drift.
Not one of these edges is statistically significant. They are shown so you can see how small and how noisy the effects are — not as a ranking to trade off. Full method →
What candles are actually for
Nison is clearer about this than the people selling his patterns. In his own words, candles are "just one weapon" and "do not provide price targets." He devotes the entire second half of the book to combining them with Western tools, precisely because the patterns alone are not a system.
That reframes the honest use. A hammer does not tell you to buy. A hammer at a level you had already marked tells you that sellers were absorbed at that level — which is a reason to trust the level, not the candle. Every pattern in the book is conditioned on prior trend, and Nison repeatedly uses patterns to confirm a level identified some other way rather than to generate one.
Used as confirmation and execution timing on a thesis you already hold, candles are defensible and our test cannot refute them, because that is not a mechanical rule. Used as a signal generator, the data is not on their side.
For educational purposes only. Measured figures are backtested on daily bars, Aug 2016 – Aug 2026, across ten symbols, and are not indicative of future performance. Nothing here is investment advice.