Using TA, we can better determine when a stock is in the process of becoming more valuable aka trending up. We also can observe when it's likely that a stock will not become more valuable (the top) and even when a stock is likely to become less valuable (trending down).
Things like moving average crossovers might work for some people. I'm sure there are people out there that swear by the MESA Sine Wave or some obscure study like that. Personally I've had more success just keeping it as simple as possible, and that's what you might find in Stan's book (which is not a referral link, and I'm sure you can find the information elsewhere).
This works because of Efficient Market Theory, which I do subscribe to. It doesn't invalidate other means of choosing investments, but for certain minds, it provides a good framework for making decisions. That's really why I'm confused about all the hate for TA in here: I was under the impression that hackers like to make decisions based on data. Maybe WallStreetBets has given it a bad name or something.