/
Support
Press / to searchBest practices
01
Always test on demo first
No , however well built, can guarantee profits in live without sustained demo testing. Validate behaviour over at least 2–4 weeks of historical data.
02
Start with one EA per symbol
Limit early tests to a single and . Adding complexity before understanding base behaviour hides which variable caused the change.
03
Be specific, not generic
Concrete details such as exact values and precise conditions give the generation process more context to reflect your intent. Vague descriptions can produce arbitrary parameters.
Heads up
Algorithmic trading carries significant financial risk. Never run an EA live without first testing it on a demo account across different market conditions.
Can't find an answer?