DislikedThe flaw seems to sit on the fact that there is no way to identify if that luck attributed to that success, and therefore it is hard to be confident that it will continue to work in the future.
I am not trying to say that some technical systems can not have some sort of logic rooted in them, however, some do not but can work for a given period of time.
Breaking down the systems and assigning some sort of logic to them is a way to differentiate the two.Ignored
If you recognized a pattern of continuation, for example, in the past, you measure its attributes using technical indicators, and conclude that it has a statistically-significant probability of success, that's enough logic for me. In fact, statistical significance is exactly what allows me to rule out luck. When you say "there is no way to identify", you are ignoring the whole science of statistics.
By the way, did you backtest your own system? Did you forward-test it, either live or in demo/paper mode? Let's assume that you did, and that it was successful. I am now going to turn back your own question back to you:
how can you be confident that it will continue to work in the future?
I appreciate that "logic" may be one part of your answer, but since markets are not logical, is it "logic" that ultimately gives you that confidence - or is it those back-/forward-tests?