Sometimes I run the Nisons Candlestick Unleashed (NCU) Explorations, and it seems to not see some of the candle patterns it purports to be able to find. For example on my Daily DOW Transports chart, on 07/09/2003, an apparent Northern Star formation was missed. Why does this happen, is NCU not reliable?

For the most part, NCU is dependant on the MetaStock Candlestick Functions to identify the candlesticks. These functions are strictly defined (having been created following Mr. Nisons direct instructions). For instance, the DOJI, HAMMER, INVERTED HAMMER or HANGING MAN, have a "Open and Close prices must be the same or very close" requirement that MUST be met. If the distance is even a small amount larger than the "very close" defined in the function(s), the candle(s) will not be identified by MetaStock, as what you might expect it to be.

If you bring up the DOW Transports index and view the chart zoomed-out to a normal viewing level, the candle on 07/09/2003 looks like a DOJI, and since it is preceded by a Big White Candle, the two together look like a "Northern Doji" formation. However, if you zoom-in tight on that area of the chart, you will see the body of the 07/09/2003 candle is just a bit too big. In fact, MetaStock would identify it as a SPINNING TOP candle, not a DOJI.

NCU is as dependable as it can be. It can only look at the candles based on the mathematical relationships of the prices as defined in the MetaStock Candlestick functions. What you may see as a particular formation, may not meet its strict definition of that formation.

What do you do when you see what looks like a particular tradable candle formation, but MetaStock/NCU does not see it? I believe I saw Mr. Nison say on his web site once, that sometimes you have to go with what looks good, since sometimes "if it looks like a DOJI, and acts like a DOJI, then it must be a DOJI."