Jeremy,
But fortunately we are talking about probabilities. Everything we observe seems to follow physical law, but there is a finite probability that we have made a mistake in our observations or that our understanding of physical law is incorrect, or even that we just missed something.
Or that our understanding of physics is incomplete. To take a fairly obvious example, pre-Einstein Newtonian physics was fine (and in many cases still is). When it broke down at the scales of the very small and the vary large though that didn’t imply supernaturalism.
As for probability, we’re talking here about the everyday use of language vs epistemological usage. For the most part “there’s no evidence for god/leprechauns” is ok because it reflects a commonplace reality of the speaker. It’s still though overstating when expressed as a fact rather than as an opinion, unless the speaker is omniscient.
There seems to be no interventionist god, but we might just have missed the interventions. The probability of that, is, I think, very small.
So do I, not least because you’d have to extend the same principle to leprechauns, to pixies, and to anything else that popped into anyone’s head. You might also point to the reasoning of asking why a god who wanted you to know he was there would cover his intervening tracks so thoroughly that the universe looks exactly as you’d expect it to look if he wasn’t there at all, and as Mr Occam tells us….
As Nearly will remind us though, probability is itself a naturalistic concept so applying it to claims about the supernatural is necessarily problematic.
