Yes, I read the other bit about self-fulfilling prophecies, but Jesus did have the opportunity to fulfill at least a few of them deliberately. I’m not talking about all of them here. The ones which are “astonishing in their specificity” and out of the main man’s control are better candidates for Made to Order.

In the first bit, about Stoner’s calculation, you rightly say that you could well have repeated stories in many newspapers. I agree – it would not be surprising to find someone doing one isolated thing the same.

But then you say “Just one good, evidenced, fulfilled prophecy would be enough”. Haven’t you just refuted this? I’m sure many people could prove they’ve fulfilled one of those prophecies… which is why the whole list is needed.
It’s like a menu, so that when *all* these things are combined, we’d recognise him.

And I dealt with the self-fulfilling in my post on prophecy here. Specifically this bit and following:
“Some are deliberately fulfilled by Jesus, so you could argue there’s nothing supernatural in that (eg Isaiah 53:7, fulfilled Mark 15:4-5), though then you’d have to accept that Jesus was conscious of his massive claim.

But then there are those which Jesus had no control over and are astonishing in their specificity. Eg: …”

Stoner’s calculations use the hypothetical premise that it all actually happened, and without any knowledge of the existing prophecies. The probability that I will repeat a specific action described in a newspaper in Oslo in 1926 is in the same ballpark, but it’s the same for any paper in any country on any date, and I’d bet I’ve mirrored stories from throughout the century.

Within the Made to Order scenario, the sheer number of prophecies fulfilled might actually decrease credibility. Just one good, evidenced, fulfilled prophecy would be enough for a godly person to establish his cred. Is all the extra credit necessary?

Self-Fulfilling would work along the same lines. Jesus knew of some prophecies himself; could he have seen them as a checklist?

Are you familiar with the work of Peter Stoner (Professor Emeritus of Science at Westmont College)? I’m not, but a google search turned up his name
He calculated the odds of a person filling 8 messianic prophecies to be 1 in 10^17. Of 48: 1 in 10^157. To my knowledge he never calculated the probability of someone fulfilling all 456 prophecies identified by Edersheim.

Now, his research might be totally bogus, I don’t know. But on the face of it, that seems to take you into the realm of significant probability of miracle, no?

Labeling all supernatural events as miracles, an event can’t be both natural and a miracle, but it must be one of them. Therefore the probability that it’s natural OR a miracle is 1, or
P(natural) + P(miracle) = 1

Shift P(natural) across, and you’re there.

Your source of confusion is this: it does NOT follow that if an event has a natural explanation, then it is definitely not a miracle. That an event has a natural explanation doesn’t mean P(natural) = 1, it just means P(natural) is not zero. That leaves room for a miracle. How much room depends on the natural explanation.

How do you show something is definitely natural? Surely now we’ve stopped talking about explanations…

P(definitely a miracle) = 1 – P(possibly natural)

P(possibly a miracle) = 1 – P(definitely natural)

Which one you look at depends on how strongly you want to argue for an event as a miracle.

P(definitely natural) cannot possibly be 1, because perhaps the supernatural is emulating natural happenstance exactly. It’s the unprovable negative again. P(definitely a miracle) of 1, on the other hand, might be achieveable in tightly controlled conditions where every natural mechanism can be ruled out.

Hrmmm… I’m not sure where you get this. Does it follow that if a result can be explained with a total probability of 1, that in this situation a miracle is impossible?
That seems totally absurd.

I just wonder where this idea comes from?

(Of course in that situation we’d have no basis for thinking it was a miracle, but that does not make it impossible that it was one)

That doesn’t mean it rules them out. It simply considers the likelihood or unlikelihood of only known natural effects producing the same results. Next is the simplest part of the calculation: one minus the probability of all natural options (which individually are much easier to estimate, and in some cases even calculate) is the probability of supernatural options.

The next job is narrowing those down to the right deity.

I did a bit of statistics in Psychology, and that all rings familiar bells. I don’t remember statistically significant results having a set probability, but rather a certain test had a certain threshold… but maybe my memory fails (or we didn’t go deep enough).

It’s difficult to give probabilities to these things, though, isn’t it. Especially over against the probability of the explanation (because, of course, the probability of an alternative explanation is never 100%). What of the case where the ‘God explanation’ is more likely than the alternative on balance of probabilities?

And don’t those calculations depend on your starting assumptions anyway? The committed materialist will maintain the probability of a miracle is 0 (or minute), while a theist will say it is good (and 1, if the theist is convinced God wills it).
C. S. Lewis has some great things to say about this in Miracles (a book I’ve only flicked through)

You’re right about the disciples, but you’ve got me thinking now. If they were in it for the money, that doesn’t mean they were successful.

I have no positive reason to disbelieve in God. It’s an unprovable negative, if indeed it’s a negative. I became an atheist when I realised that my reasons for believing weren’t good or even real reasons.

The lack of a positive reason to disbelieve is not in itself a reason to believe, as is obviously the case with, say, fairies. Otherwise we would be compelled to believe in multiple, contradictory things, and have to choose arbitrarily.

The hypothesis of the absence of gods is what’s known in statistics as a null hypothesis. It’s an uncontroversial hypothesis I adopt by default in an apparently natural world, one which requires evidence to contradict rather than confirm. In order to reject the null hypothesis and accept an alternative hypothesis such as the existence of a given god, a hypothesis which must be confirmed rather than contradicted, the evidence for it must be significant.

That’s not a vague statement, it’s very specific: “significant” means there’s a set probability (ranging from 90% to 99.5% for most purposes) that the evidence is not simply happenstance, coincidence or misleading random variation. Once one reaches that point, one rejects the null hypothesis and adopts the particular alternative, accepting the remaining 10% to 0.5% probability that one has made a mistake. It’s a justifiable point at which to change one’s mind without having any conclusive proof one way or another.

To apply this to the topic at hand: would I allow a minimum 90% probability that this dream, combined with all other claimed fulfilled prophecies I know of, are too many and too detailed to have occurred as part of normal world workings? No. Not when I consider the set of alternative explanations which apply to most or all of them, the fulfilled prophecies (about Muhammed, for example) which support contradictory conclusions and the silent multitude of predictions and premonitions which never come true, and thus are never reported. (Occasionally you do see public prophecies go pear-shaped; witness the works of Ronald Weinland.) For an atheist, Hayesy, the potential problem isn’t too many explanations but too few. Right now there are tons.

Hope that makes sense, maths/comp-sci grad to law student.

When, in essence, is there enough evidence that one stops justifing a multitude of explaining-aways to account for a hypothesis one has no independent reason to hold?