How do I justify that prior evidence matters?

Question

Suppose we lived in a world where regular interactions with the supernatural was common, so common to the point that it was independently confirmed. Imagine as if we could literally see gods and angels and everything. Let’s call this world A.

Let’s call our current world B.

Now, suppose someone is about to roll a million sided dice and asks God to help him roll the dice on the right side he predicts. Let’s say it lands on his predicted side and that the dice has been prior tested to be fair. Realistically, we’d probably get him to do more dice rolls, but for the purpose of this question, I’m mainly interested in how we should shape our beliefs after this one dice roll.

Now, imagine as if this scenario occurs in both world A and world B. Intuitively, despite the same scenario, this seems like more damning evidence of supernatural interference in world A than world B. It seems as if one should update their credence in God, but perhaps minimally so in world B compared to world A. This would be for the obvious reason that at the very least, no supernatural causes have ever been demonstrated. Or to be more charitable, supernatural evidence is disputed by many.

This seems to be a result of the difference in background knowledge that shapes my intuition here. But how do I justify this? How do I justify that prior background knowledge independent from a particular data piece that we’re concerned with (such as the dice roll) matters in the first place?

Answer 1

If one can literally see gods and angels and have detectable interactions with them, they would not be part of the supernatural. The effect on the dice would not be "supernatural interference" in World A.

You have just come upon the proposition that occurences and evidence that is consistent with existing theories does not require updates to those theories.

This kind of reasoning is also an element of Bayesian inference, where probability of a particular theory being true is informed by many prior observations, not simply the latest observation.

Answer 2

In your World A, where gods and angels are supposedly visible, I would first question the very premise. How do we know these entities are actually supernatural and not simply advanced aliens or some as-yet-unexplained natural phenomena? The history of human progress is littered with 'supernatural' events that were later explained by science.

As for the dice roll, even in this fantasy world of visible deities, one successful prediction is hardly convincing. What's more likely - that the fundamental laws of probability have been suspended at the behest of a cosmic overlord, or that we're witnessing a simple statistical anomaly? After all, someone has to win the lottery, but that doesn't mean divine intervention was involved.

In our actual world, World B, the situation is even clearer. We have centuries of scientific advancement demonstrating that the universe operates according to natural laws. Not once has a verifiable supernatural event been recorded or replicated under controlled conditions. Given this overwhelming body of evidence, it would be intellectually dishonest to suddenly abandon our understanding of reality based on a single fortuitous dice roll.

While extraordinary claims require extraordinary evidence. A god intervening in a dice game is about as extraordinary as claims get. One lucky guess doesn't even begin to meet the burden of proof required to overturn our entire understanding of the cosmos.

In both worlds, the rational response is the same: seek natural explanations first, demand repeatable, verifiable evidence, and maintain a healthy skepticism towards claims of the supernatural. Anything less is an abdication of our responsibility to think critically and seek the truth, however uncomfortable or mundane that truth might be