Live trading vs backtesting  which one do you trust more?

I've become a lot more careful with EA backtests lately.

An EA can look amazing in a backtest, but live trading can be completely different once spreads, execution, slippage, and changing market conditions come into play.

For me, the real test is how the EA behaves on a live account over time especially during difficult market conditions.

I'm currently testing one myself and so far it has been interesting, but I'm still collecting data before making any long-term conclusions.

What do you guys trust more when evaluating an EA: a strong backtest, verified live results, or a combination of both?

你好你好