You know what? I'm going to give this a crack just to get some Bayesian juices flowing!
Let's discuss the probability that a God that intervenes in our affairs and occasionally heals people in unambiguously supernatural ways exists. We'll call it P(G). Being reasonable, we should assign a very low initial probability value to P(G). This should be reasonable. If you believe it should be higher, then that means we have to go backwards a bit and do a different set of Bayesian calculations to see if it should. That said, let's be generous and make it a probability of 1 millionth (this is a very generous number):
P(G) = 0.000001
Now let's include the probability of at least one amputee's limb being fully and spontaneously restored to adequate length and functionality, given that such a God exists. We'll call this R. We'll give it a good probability given this particular God's existence, and rather low given such a being doesn't exist.
So P(R|G) = 0.60 and P(R|~G) = 0.000000001 (again, being generous here)
I'll use this link to avoid any errors with my calculations (hopefully this is a good calculator).
https://ludios.org/bayes/
So putting in the initial values, and knowing that we live in a world where R has not happened, we get that:
P(G|~R) = 0.0000004000002404001441 (Which is a magnitude lower than the initial P(G)).
So probability of God goes down not up.
Now if we were to have R occur, then P(G|R) would be quite high (virtually 100%):
P(G|R): 0.9983361081503096641436
And this is of course all based on my initial values (I was being generous). Perhaps it needs some major/minor tweaking here or there, and agreement by others, or perhaps I got it all wrong (correct me if so, Alex), but that's the gist of how it should be applied.
Let's discuss the probability that a God that intervenes in our affairs and occasionally heals people in unambiguously supernatural ways exists. We'll call it P(G). Being reasonable, we should assign a very low initial probability value to P(G). This should be reasonable. If you believe it should be higher, then that means we have to go backwards a bit and do a different set of Bayesian calculations to see if it should. That said, let's be generous and make it a probability of 1 millionth (this is a very generous number):
P(G) = 0.000001
Now let's include the probability of at least one amputee's limb being fully and spontaneously restored to adequate length and functionality, given that such a God exists. We'll call this R. We'll give it a good probability given this particular God's existence, and rather low given such a being doesn't exist.
So P(R|G) = 0.60 and P(R|~G) = 0.000000001 (again, being generous here)
I'll use this link to avoid any errors with my calculations (hopefully this is a good calculator).
https://ludios.org/bayes/
So putting in the initial values, and knowing that we live in a world where R has not happened, we get that:
P(G|~R) = 0.0000004000002404001441 (Which is a magnitude lower than the initial P(G)).
So probability of God goes down not up.
Now if we were to have R occur, then P(G|R) would be quite high (virtually 100%):
P(G|R): 0.9983361081503096641436
And this is of course all based on my initial values (I was being generous). Perhaps it needs some major/minor tweaking here or there, and agreement by others, or perhaps I got it all wrong (correct me if so, Alex), but that's the gist of how it should be applied.
