What I've found is that most people I've spoken with who believe they have a trading plan based on research have, in reality, a trading idea based on observation which is notoriously deceptive.
A few years ago I was caught in some losing trades based on a similar pattern and I noticed something about the context that led me deduce this setup (pattern + context) had extremely negative expectancy.
Then later I was putting together some trading rules in a way that a programmer could automate them. I was trying to figure out how to filter out this setup and I wanted to be able to compare this same pattern in different context to the one that had caused my losses, so I did an objective statistical analysis.
The way I help ensure that my analyses are as objective as possible is I scroll the charts all the way to the hard right edge and reveal a single bar at a time. This way when the pattern appears I can only see the context (to the left) and the pattern itself; I have no idea what's going to happen next until I reveal the next bars.
Surprisingly, I found that this setup had a 55% win rate when I applied my risk:reward parameters to it.
Because I'd felt pain when taking the losses, I instinctively came to the conclusion that the pain surrounding the pattern+context I'd identified (observationally) indicated a poor trade setup.
The problem is our observations are colored by our beliefs and experiences. If we notice price does something similar a few times , we then begin to only see the successful occurrences down the road. It's human nature: we stereotype and judge based on a small sample size of consequences. It's a protective mechanism in life, but dangerous in trading.
If you have a trading idea, test it objectively. Then have another person apply the rules and see if they come up with the same results.
The reason my trading plan evolved so nicely is that I was talking about it with other traders several hours a day in real time and after hours as well. I had to become more and more objective and precise in my rules and definitions.
