I'll play along and track this with you. However, I am a strong believer that the past does not predict the future. The core premise of technical analysis is that the past predicts the future.
No, that is not the core premise. This is your first misunderstanding of what TA actually is. Past trades most certainly do correlate with the current positions of the various market participants. Support and resistance levels are areas where lots of trades have happened in the past, so they mark important price levels for many people.
In reality, there are always many different variables in the equation when comparing different time periods and chart action. In the stock market, the only reason technical analyis works is because enough people make purchase and sell decisions using it. With bitcoin, I am hoping that the psuedo-science of "technical" analysis will not catch on.
I agree there is a lot that passes for TA that is not rigorous enough to be worth anything, but I believe that when rigor is applied and the number are crunched, you can do better than random. Once you have that, it's all about letting the winners run and cutting losses quickly.