Hi all, I'm new to this forum and this thread, so thanks for having me here! I haven't read (nor will I read) all 108 pages of this thread, but would like to weigh in.
rcanfiel, first of all, I would ask why you started this debate in the first place? I'm amazed at why so many people consistently feel the need to disprove things that they themselves have nothing to do with. If you were a TA user, then I would understand. But why point fingers when it doesn't affect you? In my experience, this is what people do when they feel insecure about their own methods or themselves--after all, if you're content with your own methods, by the definition of contentment, you wouldn't even consider attacking other methods. But maybe that's not true--just an observation.
Secondly, when you talk about large sample sizes and statisticians, it makes me think you're talking about a rigid science. Perhaps you don't understand what TA really is? Nowhere should TA be defined as mechanical--to the contrary, it is often quite discretionary, and therefore no "TA logic" exists which could be tested. Sure, mechanical systems exist, and of course they will fail over a large sample size, as times and markets change. Duh. Good TA is merely an instrument that a good trader will use to his or her advantage. A good technical analyst should be able to adapt methods to fit different markets and times.
Could you take a good fundamental trader's response to fundamentals, and code some "logic" on how to trade based on these responses, and give these to some statistician? If so, you'd have utter failure as well. I suggest you try real TA before you trash it, if you have not already--if you had, however, I don't think you would make the statement you did.
