Cultists will say they've met jesus. When you ask "What does he look like?" they'll come up with all sorts of abstract answers until your eyes glaze over.
Back testing is not the holy grail, it can be and often is curve fitted to justify a bias in thinking.
If the definition of "TA" is using past price movements to future predict price movements, then yes, TA works. Price information has predictive power in certain contexts. In my experience you need a bunch of computer power to extract this information. If the definition is drawing a bunch of lines on a graph, I can't help you there, but I suspect that way lies madness. Either way, anyone with a real edge isn't going to tell you what it is, especially not for free on an internet forum. That's not how this business works.
True - no one who has developed a REAL edge that is consistent is going to tell others what it is. The best way to make money with TA? Program and code your own “indicator” pkg that generates up and down arrows for entries and then do some backtesting to provide “proof” to the lazy and uninformed that it works. Then sell the indicator to lazy traders (a/k/a Robinhood, stocktwits, Twitter and YouTube) who want a piece of software to make their life easy by pointing out when to buy and sell based on the up and down arrows the indicator generates and be sure to include all the usual financial disclaimers that past performance is no indication or guarantee of future results and all the other usual disclaimers all other indicators are sold with
I use TA exclusively but I think the TA that most people use isn't that valuable. People love to argue for or against TA. I generally don't care but when that TastyTrade guy says "nothing works" just because he tests some individual candlestick pattern in isolation, he's just confirming his own bias - a lot of people say TA is useless simply because they haven't figured out a way to make it work for them. Again, it's just a set of tools. Either you're able to use those tools in a useful manner or you can't. And if you can't, that's fine. Move along.
Since we do not know what the price will be tomorrow, so we try to predict it was past data. The saying goes... the market always repeat itself. So we employ various strategies using past data to predict the future price which is what back test essentially is. With the power of modern PC, it's not difficult to test various strategy from simple to very complex ones. I look at daily charts every day and all seem random. Occasionally, I see one or two that trend... ah ha... I jump it. Sometimes, after I jumped in and the trend broke the very next day. It is just a bet after all. You win some and you lose some. Hopefully, you win more than lose. We just have to accept that the market will be the final judge.