Nothing to do with randomness. Traders look for trades, but they need not do so randomly. They can just as well do so methodically. Sellers will raise the ask until buyers are no longer willing to pay. Buyers will withhold their bids until sellers are willing to lower their prices no further. Either way, the extremes will offer the fewest number of trades. The greatest number of trades and trading opportunities will be found toward the middle. Thus the doji, which of course will exist only within a particular bar interval.
There's a big (enormous) difference within Japanese Candlestick Analysis between single candlestick lines versus candlestick patterns...as in one interval versus multiple intervals to provide price action context. Simply, a doji line all by itself provides LESS information versus a doji pattern. This thread mentions nothing about doji patterns. Thus, I'm going to assume the thread is started about a single doji interval. Further, dojis (lines or patterns) provide less information about price action context in comparison to other candlestick lines or patterns. For example, the Hammer line/pattern provides more information about key changes in supply/demand in comparison to doji lines/patterns. Without getting into any statistical details, overall Japanese Candlestick lines/patterns are not profitable all by themselves and the results varies dramatically from one trading instrument to the next and the same from one trade management plan to another. In contrast, traders that are profitably using Japanese Candlestick patterns are not using them alone. In fact, they are just using them to provide information about key changes in supply/demand instead of as entry signals. Yet, if you want some kind'uv of statistics...take a look at the stuff by Thomas Bulkowski @ thepatternsite.com I think he only monitors five types of dojis and one of them ranks in the top 5 out of 103 candlestick lines that he follows via whatever methodology he uses. I'll let you do your own research to determine what candlestick lines are ranked higher than dojis via whatever it is that you trade because as stated above...the performance rankings will dramatically change depending on what you trade and how you trade it (trade management after entry) if you're going to use candlestick lines/patterns as entry signals. My own testing (will not disclose) involves Emini ES, NQ, TF, EMD, Eurex DAX, Eurex DJstoxx50, Euronext CAC40, Euronext FTSE100 since the birth of those trading instruments. Also, 15 years of info in other futures instruments (e.g. 6C, 6E, ZB, ZN and others). Hammer lines/patterns in my stats rank higher than doji lines/patterns. Regardless, Japanese Candlesticks aren't my primary strategy, I don't use them as entry signals and I mainly use them for information about changes in supply/demand and profit targets. P.S. Don't trade them via codes or automation...they were designed for such.
Correction in my prior message P.S. statement because I forgot to write the "not" word in that statement. Thus, it should have said the following... P.S. Don't trade them via codes or automation...they were not designed for such.