Biggest challenge of markets is emotional, even once you have an edge. And emotions can interfere even with finding an edge as biases get in the way.
Intraday trading with discretion probably has the biggest impact on emotions. Markets are running live, and money matters messes up the brain.
Any trade that you did not take was not a trade. If a trade failed you cannot go back and feel good that you missed it and if it did well you cannot start regretting. There is no end to this and FOMO is not an edge.
I think pure discretion is incredibly hard to do, and do not consider it to be worth it. Obv someone may disagree.
But discretion does matter - whether in system design or as part of the system.
This is what helped me -
Test the market data. I dont mean just take an algo and optimize a b c and get a system. I mean to test it to understand how it actually works. How it responds to large moves, momentum or mean reversion ? Impact of Volatility etc. From a basic understanding of a market, you can get a feeling and a source of ideas to derive a system.
You can also test your discretionary ideas and check if they have any basis in reality. Its hard to map discretion with logic but an approximation may be enough.
For me, it was a major change once i went from pure discretion reading only charts to testing the data to understand how it works. For ex, Pullbacks work but when do they work ? How should i manage trades etc etc.
Testing this manually is incredibly and deceivingly hard to do. Emotions and biases get in the way along with the need to test a large sample of data. I was a loop of making changes to plan and testing them and again changing them based on recent result which is pointless.
We need to appreciate how random the market is, how you can get a streak of bad results for a prolonged period of time even with a very good system, even for very good experienced traders. Drawdown is part of trading - both price and time.
All of this takes time, but once you have a probabilistic mindset, you will realize that individual trades do not matter at all. There is no point in banging you head on the wall in response to what happened recently. Its easy to say and everyone will feel something, but it gets easier with time. Moving to completely mechanical trading in intraday has almost freed me from such problems. But i still feel the ups and downs a bit but its easy to ignore.
A single trade can be source of ideas instead, if you missed a trade - do such trades have good edge? Define it and test. Once you have this kind of process, it will be very hard to break your rules - you have be afraid to do that and lose money in the long run. Instead iteratively improve your system based on emotional feedback like this.