> Evidence implies something we can observe That's an arbitrary definition. We can't observe our local past events or even non-local events, yet we use them as

> Evidence implies something we can observe That's an arbitrary definition. We can't observe our local past events or even non-local events, yet we use them as theories/evidence depending on context (eg Red Shifted galaxies as evidence of expansion). Reasoning about past events would affect other future events is just as much evidence. That's how physics experiments are generally conducted.