That’s actually wrong. Short-term patterns will improve your chess short-term but you’ll hit the plateau pretty fast. They don’t teach you to play better at chess. It’s like copypasting some code from the Internet without fully understanding how it works.
In chess, it’s absolutely necessary to have a long-term plan: especially at lower levels, your ultimate goal should be attacking the king most of the time.
Use the engine smartly. Because it will tell you what move is objectively the best. It won’t tell you what move is best at your level. At different levels the best moves are different.
It’s important to train tactics but not with puzzles. Puzzles model situations that are out of context. It’s like training your swimming movements without hitting the water. It may be not harmful, but not particularly useful. Instead, you should get “puzzles” in your live games within the context, seeing the bigger picture.
Of course, you should play games. How can you expect to get better at playing chess without actually playing chess? Moreover, playing games and getting feedback on them should be the bulk of your training. Learning happens by trial and error.
Without dissecting this article any further, there is a battle-tested approach on how to get better at chess at any level and age:
- Be a kid. Don't overthink it. It's just a game after all. Feel it.
- Play as many games as possible. You can quickly get enough volume with bullet and blitz.
- Get feedback. You're able to give yourself decent feedback most of the time. Quickly analyze your games yourself or check them with someone, who can tell you what mistakes you have made at your level. Only afterward compare your thoughts to engine suggestions. Pay attention to the moves that drop the evaluation by more than (1, 2, 3, 4) points, depending on your level.
Of course, these are not hard and fast rules. But this is a solid foundation for the success.
P.S. My FIDE rating is 2423.