I suggest you go back over a trend at look at how many doji's and shooting stars there were on the way up before the one at what proved to be the top that caught your attention
Good point. Lots of false signals. Of course, other factors should be taken into consideration when using these candles. The marubozu can be very helpful in the right circumstances.
Hammers seem to be most reliable - think of the logic - participants are so sure they will bid aggressively to take a countertrend* position, whereas with shooting stars someone took partial profit, stops got hit, all the way up. Ultimately what they provide is quick visual recognition of OHLC and where supply/demand were out of balance for the period. Most (all?) multi bar candle patterns blended together seem to really be a hammer, shooting star or doji. *Countertrend as in against a correction in the main trend
I think Japanese Candlesticks are fatter than bars. So it is really easier to see them. Except for the doji's. They are just as skinny as the bars. Because they are just like the bars with a skinny little bar going across in the middle or near the top or sometimes near the bottom... I hope this has been a wonderful explanation of Japanese Candlesticks... Now you all can send me $19.95 for this secret information you just learned...
There's some reliable stuff out there. However, just like anything else...it takes a lot of time and experience of finding out what doesn't work and what works. Yet...a great place to start learning or re-visiting candlesticks is the Long Shadow family. NihabaAshi
Some candlestick patterns are of higher reliability than others but I consider them all just part of the stew. I don't think you can build a good trading system off of just candlesticks any more than you can off just any single indicator, or average, etc.