Hey I would disagree here. The reason why most the standard indicators are not that useful is because they simply re-represent price. Whereas the change in price in one period is not a good predictor of future changes in price i.e. the auto-correlations of price price movement is low.
Price movement is generally random especially in the lower time frames M1 M5 etc… i.e. price movement is Markovian:
That it is why the best way to model/estimate future price movement is to use Geometric Brownian motion (Wiener Process).
To have a level of predictability on where the price might go in the future based on the past prices you should use a simple moving average to establish the trend; this is the drift:
But also allow for the fact that interim price movement will be quasi random about the normal trend. To find your edge you need to quantify the drift (general average direction) and also allow for the randomness.
If you imagine a situation where the price is trending up with low volatility in randomness (low standard deviation) these are the conditions where an indicator will be useful by providing a clearer picture on price action for entry and exit signals.
If the volatility in randomness is very high on the other hand, then it’s anyone guess where the market will go. The pair might be going in the direction of the trend but who can say it won’t knock you out on your stop loss before doing so.
The best way to describe and quantify the risks of this happening i.e. find your edge is to use those variables:
Future price (@ time T) = Drift + (Random Price Movement)
and model the probability of a successful trade by doing numerous trial runs: Monte Carlo Simulation.