Day 8/30 — I’m starting to understand why most trading bots fail
A week into this project and honestly…
I get it now 😭
From the outside, building a trading bot sounds simple:
“just define the strategy and automate it”
But once you actually start testing things in live conditions…
you realize markets are full of edge cases.
One candle behaves differently.
One session moves weirdly.
One fake breakout ruins the entire logic.
And suddenly your “clean strategy” becomes 50 different exceptions and conditions.
Today I spent a lot of time trying to simplify things again.
Because every time I add too many filters:
* the bot becomes slow
* logic becomes messy
* overfitting starts creeping in
But if I make it too simple…
it starts taking absolutely cursed trades 💀
So right now the biggest challenge is balance.
Trying to make the bot:
* structured enough to avoid nonsense
* but flexible enough to survive real market behavior
Honestly this whole process is teaching me more about trading psychology and execution than manual trading ever did.
The bot exposes every flaw instantly.
No excuses.
No “intuition”.
Just pure logic getting tested in real time.
Still learning every day.
Still debugging constantly.
But slowly starting to see where an actual edge *might* come from.