Paper trial

A test run of a new version of a trading rule, using simulated money, alongside the version you already trust. Both see the same prices on the same days, so any difference between them is the rule, not the market. A backtest can’t promise that, because it compares against history.

Say your current rule made $18 on five trades over two weeks, and the new one made $42 on six trades over the same two weeks on paper. That’s a reason to look closer. It isn’t proof: eleven trades is a small sample. The point of the trial is that the evidence comes from live prices, not from the history the new rule was tuned on.