Why do many EAs fail while the same strategy works perfectly in manual trading? - page 2

New comment
 

You need to add your entire strategy to your bot so it can act with greater precision, such as using breakeven, trailing, and everything else you think is correct to do; you must pass it to the bot in the code.

 
RENATA PALIOKIENE:
Why do many EAs fail while the same strategy works perfectly in manual trading?
Manual trading lets us adapt, filter trades, and read the market in real time. EAs, on the other hand, follow strict rules — they can’t “think” like a human. Automation often reveals the gaps in a strategy rather than fixing it.

Have you ever automated a strategy and been surprised by the results?
Cuz most people discover the ugly truth about their strategy the moment they automate it ⚠️

Manual trading gives you something an EA doesn't have:
"JUDGMENT"

You skip bad setups
You adapt to volatility
You feel when market conditions change
You avoid trading during garbage price action

But the moment you turn that same strategy into an EA… the machine exposes every weakness in the system 💀

The EA doesn't "believe" in the setup.
It executes exactly what you told it to do.
No emotions. No hesitation. No intuition.

That's why many manual strategies collapse under automation.
Not because automation is bad…
But because the strategy itself was relying too much on human filtering instead of having a strong core structure.

A strong EA strategy should survive even without babysitting.
If the system only works when a human constantly interferes with it… then the edge probably isn't as strong as people think 
 
Alex Holloway #:
You skip bad setups
You adapt to volatility
You feel when market conditions change
You avoid trading during garbage price action
Exactly. All of that must be quantified in order to reproduce the actual manual strategy as a bot. Again, it initially sounds absurd to suggest that a person can't define what it is that they, themself, are doing. At the same time, sufficient knowledge of MQL5 is required because if you don't know the code elements required to capture your thoughts, then you can't even frame the conversation with yourself─so to speak.
 

I recently experienced exactly this while converting discretionary ideas into EAs.

Once realistic spread, slippage and multi-year backtesting were added, many “good” strategies became flat or slightly negative. The EA wasn’t necessarily bad — it just exposed how much human discretion was filtering trades manually.

 
Victor Eduardo Velazco Delgado #:
You need to add your entire strategy to your bot so it can act with greater precision, such as using breakeven, trailing, and everything else you think is correct to do; you must pass it to the bot in the code.

Exactly. I also realized that many discretionary decisions are actually trade management logic rather than entry logic.

Things like breakeven, trailing, partial exits, volatility filters and even “not trading” during certain conditions all need to be quantified properly, otherwise the EA behaves very differently from the human trader.

 

One underrated reason: most EAs use 1-2 conditions to enter (EMA cross + RSI), but manual traders unconsciously check many more things — trend strength, structure, volume, candle quality, higher timeframe context.

That's why I moved from building EAs to building scoring-based indicators. Instead of a binary "buy/sell," each trade gets scored across multiple independent dimensions. A score of 92/100 means almost everything aligns. A score of 65 means proceed with caution.

The manual trader's edge isn't the strategy — it's the multi-factor filtering they do instinctively that's hard to code into a simple if/then EA.

12
New comment