1. Can you verify the record yourself?
Screenshots and spreadsheets are only hints. A read-only account you can open, or statistics synced automatically from the account, is evidence.
2. Look at the maximum floating loss, not just the balance curve
Martingale and grid EAs can show a smooth balance curve while the risk sits in open trades. Look for the maximum floating loss in the statistics.
3. Is the sample big enough?
Under 3 months or under 100 trades, treat the returns as a preview. A gold record means more once it has been through NFP releases and rate decisions.
4. Does the style match what you can stomach?
Decide which one you can live with before you pick.
- A stop loss on every trade: shallow drawdowns, but losing streaks.
- Adding to positions: profitable most of the time, with an occasional deep drawdown.
5. Can you try it first?
Ask for a 24-hour trial code and run it on your own account for a day to see how often it trades and what the panel shows.
FAQ
- Can I trust a good backtest?
- A backtest only shows that the logic worked on past prices. Ask how spreads and commissions were modelled; our own backtests use ECN-style costs (low spread plus a commission per lot).
- How many accounts can one EA run on?
- It depends on the licence; usually one code covers one MT5 account.
Read next
How to get MT5 account statistics without installing a plugin →How to use an EA trial code: try it for 24 hours first →Open NB Stats →All guides →
Updated 2026-11-13