Your perception on this is generally correct. You can't just find some pattern or anomaly in markets and just consistently exploit it, as you need to assume everyone else can and will. You need to accept this premise and then think about situations and techniques that might work in spite of this.
Anything you can program into an algorithym, any technicsl indicators calculated by comuputer are unlikely to provide a competitive edge. Any single pattern or price assessment unlikely. Daily and other small timeframes more exposed to noise or direct manipulations of price.
What can work is discretionary chart reading not automated as it relies on individual skill and training. Using combinations of TA to interpret direction, e.g. combined price, pattern, time techniques to read the chart and seek confirmation. Use longer timeframes and weekly and monthly charts avoids the noise and gives more reliable signals.
