The great and awesome Day of the Lord

[Mithrae]

As y'all know despite being raised as one I'm no longer a Christian and haven't been for over a decade: I don't accept core Christian doctrines (eg. I'd only guess ~70-90% probability for the existence of some kind of 'God' and consider it significantly more probable than not that Jesus didn't rise from the dead), and more importantly I don't even adhere to the teachings of Jesus (to sell what you have and give to the poor, don't worry about tomorrow but trust in God's provision, stop working for money and start working for the kingdom of God).

But I try to keep an open mind, and over the past ten months or so I've been troubled by the topic of biblical prophecy. It actually began all the way back in 2011, when one prophecy enthusiast came to the forum with all kinds of fallacies and arbitrary conclusions, claiming that his pet theory had only 112 trillion to one odds of being false. So in response I did a more objective assessment, and came up with something like 100 to one instead; which is not mind-blowing or miraculous, but is still somewhat intriguing.

I've privately gnawed on that issue at times over the years, but never managed to falsify it to my own satisfaction and even found a couple of additional points worth considering too. So hopefully the rest of you can help prove me wrong.

As I see it, if a prediction has been made and parts of it had already come true, that's either coincidence or it's indicative of genuine foreknowledge. Hence the likelihood that the rest will come true is the inverse of the likelihood that the first part was coincidence. (Plus the likelihood that it would happen anyway, prophecy or no, but in the case of biblical prophecies that's basically zero and therefore irrelevant.) That is absolutely critical to my reasoning, but I can't find a fault with it: Either the fulfillment of the first part was coincidence or it was indeed foreknown, and if it was foreknown then the rest of it presumably is foreknown also; so the likelihood that the first part was not coincidence is roughly the same as the likelihood that the whole prophecy is genuine.

Thus we have -
Prediction: Prophecy and interpretation
Confirmation: Signs and complete fulfillment

The biggest problem I've found with many Christian prophecy enthusiasts is that they tend to include their interpretation as part of the 'sign,' like that fellow from 2011 (and with some particularly enthusiastic folk, simply make up what constitutes a sign from whole cloth!), and that's a key error I've tried to avoid in my reasoning. In each case I've tried to justify an interpretation of biblical prophecy as legitimate, not arbitrary, and only then begun to consider how likely it is that the 'sign' which came to pass is mere coincidence. I will try to be as brief as possible with the signs I've been interested in, but I'll still put them in another post because this is already getting on the long side for an OP. However I'll briefly comment on two of the most obvious objections first:

1 - Biblical prophecies are too vague
It's a fair point, but firstly, that is why I've tried to specifically quantify the likelihood of a fulfillment or 'sign' being mere coincidence, distinct from and after establishing a legitimate interpretation; and secondly, what would the alternative be? If a prophecy were very specific then anytime since 400 CE or so basically any 'fulfillment' would be subject to the criticism that it was engineered by Christians to match the existing prophecy. Some miraculous exceptions which could not possibly be engineered by humans might apply (though not for any of the ones below), but then there's the endless debate over whether there's good reasons why a deity would not openly and universally reveal himself in such a manner. Criticisms on those grounds are not particularly valid to my mind, since they simply assume certain things about what 'God' or prophecy should be like, rather than addressing the actual data available.

2 - Seemingly fulfilled prophecies, even remarkable ones, are still coincidental products of large numbers; many many prophecies and thousands of years of history
The charge that some biblical prophecies are obviously false prophecies (eg. those of Ezekiel or those that 'Matthew' put in Jesus' mouth about his return) falls more into this category than being a valid objection in its own right, I think; after all on its own, it amounts to nothing more than the absurd 'some prophecies are false therefore they all are.' However the more nuanced recognition of how large numbers interact with the notion of coincidence is important, and is potentially valid, if it can be shown that that the real probability of a 'fulfillment' is in fact other than what I have calculated. I have tried to be careful in considering other scenarios, other possible 'fulfillments' in my estimations of probability, so I don't consider it a valid objection to blithely state that it simply must have been more probable than I've concluded.



Questions for debate:
Is the reasoning above valid, particularly the section in blue?
And if so, are the assessments of probability for the prophecies/signs in post #2 correct (or at least reasonable)?

[Mithrae]

And yet all said was "the sun will turn to black".

No, you missed a bit:
"I will shew wonders in the heavens and in the earth, blood, and fire, and pillars of smoke. The sun shall be turned into darkness, and the moon into blood..."

The only thing I am missing is the part that you say is there but isn't. All he said about the Sun, is that it will turn to black (or into darkness, w/e).

Why should I work to make sense out of it? Why can't we allow for the possibility that it's just nonsensical rambling?

Because you're struggling mightily to make that view stick. Where would "blood, fire and pillars of smoke" occur, in the heavens or the earth? And where are the sun and moon? Trying to imagine that Joel was writing about a regular commonplace eclipse through either intention or ignorance requires a deliberate effort to avoid and deny what's clearly written there.

At any rate, I am hesitant to breech the topic too deeply not because I'm afraid of losing the argument, but because I'm afraid of starting it. It's an endless rabbit hole.

In other words I do not mean to just hand-waive your assertion about the strength of biblical prophecy, but neither do I mean to concede to it. I acknowledge that you have planted that flag, but I simply do not have [strike]anything[/strike] too much to say about it right now.

You're the one who is focusing on the broader 'set' of biblical prophecy, trying to suggest that Joel's prophecy is somehow cancelled out or invalidated by stuff from other authors who also happened to be included in the canon. The logic of that argument is questionable of course, but even more obvious is the fact that you haven't provided and probably don't have any basis for making that claim even if the logic were sound.

Of course it can be done, you just don't like the implication of a ~99% confidence that the prophecy was genuine. We've still got that ~1% possibility that it was coincidence - that's what coincidence means - and there's nothing stopping anyone from placing their bets on that long shot, but the rational conclusion if my reasoning is correct (as it seems to have been) is that it was very likely a result of genuine foreknowledge.

So I talk about how the observations are identical and you ignore it... Seems to me our disagreement is definitely about how induction fits into the picture.

I didn't ignore it; it just seems that again you don't like the answer You declared that because they're observationally identical it's impossible even in principle to assess the probability that it occurred as a matter of coincidence, and I pointed out that it certainly is possible. Drawing four aces looks observationally identical whether it's coincidence of a sleight of hand trick, so are you going to try telling me that assessing the probability of it happening by coincidence "cannot be done, even in principle"?

Definitely, it's an issue of induction. If induction didn't exist, you would be correct. Except you wouldn't, because you couldn't have generated any probabilities in the first place.

This isn't an argument, I'm afraid. I don't even understand what you're trying to assert, and you didn't answer my question.

Question: Does your assessment of the probability depend at all on whether you have have a possible trick in mind that can explain the 4 Aces? Or is it equal either way?

The odds of drawing four aces purely by coincidence are always the same. If that does happen when it's called in advance, we would very rightly conclude a near-certainty that it wasn't coincidence, that the fellow knew what he was doing... even though the result is observationally identical in each case. Similarly the odds of the prediction based on Joel's passage that there would be a solar eclipse of similar significance to the eighth feast tetrad being vindicated purely by coincidence are always the same. Since it was vindicated in spades, we should rightly conclude a very high probability that it wasn't coincidence, that it was indeed foreknown as Joel had said. Your objection about them being observationally identical simply doesn't make any sense.

[FarWanderer]

[Replying to post 101 by Mithrae]

Independent of the prediction, you had calculated the event had a 1-in-100 probability of occurring. Or in other words, you had calculated an epistemic possibility of the predicted event not occurring at 99%. If we take your simple equation about how the probability distributes after a successful prediction, 99% of the remaining 1% goes to success due to divine foreknowledge (0.99%), and the rest goes to success due to coincidence (0.01%). Prior to checking the outcome of the predicted event, the probability distribution would look something like this:

Epistemic probability distribution when a 1-in-100 prediction is made
99% chance the event does not occur.
0.99% chance the event occurs, and it was foreknown.
0.01% chance the event occurs, but only due to coincidence.

This implicitly awards any 1-in-100 prediction with being foreknowledge at a rate of 0.99% entirely a priori.

[Mithrae]

[Replying to post 101 by Mithrae]

Independent of the prediction, you had calculated the event had a 1-in-100 probability of occurring. Or in other words, you had calculated an epistemic possibility of the predicted event not occurring at 99%. If we take your simple equation about how the probability distributes after a successful prediction, 99% of the remaining 1% goes to success due to divine foreknowledge (0.99%), and the rest goes to success due to coincidence (0.01%). Prior to checking the outcome of the predicted event, the probability distribution would look something like this:

Epistemic probability distribution when a 1-in-100 prediction is made
99% chance the event does not occur.
0.99% chance the event occurs, and it was foreknown.
0.01% chance the event occurs, but only due to coincidence.

This implicitly awards any 1-in-100 prediction with being foreknowledge at a rate of 0.99% entirely a priori.

So you're starting with a calculation of a 1% chance that an event will coincidentally occur and turning that into a 0.01% chance that it will coincidentally occur, because... why? A further consequence if your reasoning were valid is that an even more remarkable/specific prediction (say, of a 1 in 1000 event) would have an even lower value as foreknowledge (0.099%), apparently regardless of whether or not it actually comes to pass. Our friend with the card tricks? Even after he has said he'd pull four aces and done so, you're trying to argue something like a 99.9999997% certainty that he didn't know he was going to do it. I'm afraid this is precisely the opposite of how any rational being would reason!

[FarWanderer]

[Replying to post 101 by Mithrae]

Independent of the prediction, you had calculated the event had a 1-in-100 probability of occurring. Or in other words, you had calculated an epistemic possibility of the predicted event not occurring at 99%. If we take your simple equation about how the probability distributes after a successful prediction, 99% of the remaining 1% goes to success due to divine foreknowledge (0.99%), and the rest goes to success due to coincidence (0.01%). Prior to checking the outcome of the predicted event, the probability distribution would look something like this:

Epistemic probability distribution when a 1-in-100 prediction is made
99% chance the event does not occur.
0.99% chance the event occurs, and it was foreknown.
0.01% chance the event occurs, but only due to coincidence.

This implicitly awards any 1-in-100 prediction with being foreknowledge at a rate of 0.99% entirely a priori.

So you're starting with a calculation of a 1% chance that an event will coincidentally occur and turning that into a 0.01% chance that it will coincidentally occur, because... why? A further consequence if your reasoning were valid is that an even more remarkable/specific prediction (say, of a 1 in 1000 event) would have an even lower value as foreknowledge (0.099%), apparently regardless of whether or not it actually comes to pass. Our friend with the card tricks? Even after he has said he'd pull four aces and done so, you're trying to argue something like a 99.9999997% certainty that he didn't know he was going to do it. I'm afraid this is precisely the opposite of how any rational being would reason!

Indeed, and that would be the point.

I believe the problem is with the assumption that the probability of a fulfilled prophecy being foreknowledge is inversely proportional with the probability it would occur.

If its not that, then I must have made a mistake in my reasoning. But where?

[Mithrae]

Indeed, and that would be the point.

I believe the problem is with the assumption that the probability of a fulfilled prophecy being foreknowledge is inversely proportional with the probability it would occur.

If its not that, then I must have made a mistake in my reasoning. But where?

I should have said complementary rather than inversely (1-p rather than 1/p) as a kind soul has pointed out via PM. But terminology aside, that part of the reasoning is sound as far as I've seen: Either the fulfillment was coincidence or it was indeed foreknown, so the likelihood that it was foreknown is the complement to the likelihood it was coincidence.

"Epistemic probability distribution when a 1-in-100 prediction is made
99% chance the event does not occur.
0.99% chance the event occurs, and it was foreknown.
0.01% chance the event occurs, but only due to coincidence. "

The problem here is that barring questions of choice and indeterminacy the 'real' probability of an event occurring is either zero or one. I'm not assessing the real probability of that eclipse occurring - which, since it did in fact occur, was always 100% - I'm assessing the probability of correctly guessing that it or a similarly significant eclipse would occur. That is, the possibility of coincidence lies in the guessing/predicting, not in the occurrence itself. I didn't make that clear, admittedly!


So if you wanted to include the occurrence or non-occurrence of the event in the distribution, it would be either
0% chance the event does not occur (since it did, and was always going to)
99% chance the event occurs, and the prediction was based on foreknowledge
1% chance the event occurs, and the prediction was mere coincidence

Or
100% chance the event does not occur (if it hadn't)
0% chance the prediction was based on foreknowledge, since it clearly wasn't known
0% chance the prediction was coincidentally accurate

[FarWanderer]

That is, the possibility of coincidence lies in the guessing/predicting, not in the occurrence itself. I didn't make that clear, admittedly!

I'm glad you figured it out. I realized we were both equivocating on this point several posts ago but was unsure how to best go about explaining it. My intuition has always exceeded my eloquence.

Indeed, and that would be the point.

I believe the problem is with the assumption that the probability of a fulfilled prophecy being foreknowledge is inversely proportional with the probability it would occur.

If its not that, then I must have made a mistake in my reasoning. But where?

I should have said complementary rather than inversely (1-p rather than 1/p) as a kind soul has pointed out via PM. But terminology aside, that part of the reasoning is sound as far as I've seen: Either the fulfillment was coincidence or it was indeed foreknown, so the likelihood that it was foreknown is the complement to the likelihood it was coincidence.

Your response here is not talking about the same "coincidence" as I was. You are saying the probability of a fulfilled prophecy being due to coincidence and the probability of that same fulfilled prophecy being due to foreknowledge must compliment. That's true. I mean, of course it's true. The definition of coincidence in this context may as well be "not foreknowledge". However, the probability of a fulfilled prophecy being due to coincidence is not necessarily the same as the probability of the prophesied event occurring due to coincidence. You have calculated the value of the latter to be 1%, but that doesn't mean the value of the former is also 1%.

Of course that leaves us with the question of what is the probability of a fulfilled prophecy being due to coincidence, if not 1%?

The answer is, unfortunately, highly indeterminate. This is because the missing factor is the probabilistic expectation of the prediction's result prior to confirmation. In other words, before seeing the result what did you think was the probability that the prediction was foreknowledge? In the context of this particular problem it would be a presupposition (in probabilistic form), but in the context of experience in general it may well be the result of valid inductive inference.

Needless to say, such expectations can vary widely. Mine are non-zero, but still trivially small, like 0.00001%. Yours are likely considerably more generous (I won't guess exactly how much). And a presuppositionalist Christian's expectations would be at or near 100%.

The final piece of the puzzle, though, that I couldn't figure out on my own, was how the prior expectation fit into the problem mathematically. I did some research, and found that Bayes' Theorem does the job.

https://en.wikipedia.org/wiki/Bayes%27_theorem

I plugged in some values (correctly I hope) and got some very satisfying results.

"Epistemic probability distribution when a 1-in-100 prediction is made
99% chance the event does not occur.
0.99% chance the event occurs, and it was foreknown.
0.01% chance the event occurs, but only due to coincidence. "

The problem here is that barring questions of choice and indeterminacy the 'real' probability of an event occurring is either zero or one. I'm not assessing the real probability of that eclipse occurring - which, since it did in fact occur, was always 100% - I'm assessing the probability of correctly guessing that it or a similarly significant eclipse would occur.


So if you wanted to include the occurrence or non-occurrence of the event in the distribution, it would be either
0% chance the event does not occur (since it did, and was always going to)
99% chance the event occurs, and the prediction was based on foreknowledge
1% chance the event occurs, and the prediction was mere coincidence

Or
100% chance the event does not occur (if it hadn't)
0% chance the prediction was based on foreknowledge, since it clearly wasn't known
0% chance the prediction was coincidentally accurate

This may be moot at this point but I will address it anyway.

The "it already happened therefore it's 100%" doesn't apply to hypothetical situations where it hasn't happened yet and you don't know it will happen (hence why I specified epistemic probability). And a hypothetical situation is entire valid since this is only a logic check.

[PinSeeker]

I've been unsure how to respond to your comments, so I hope you'll forgive my tardiness

Dude. No apology necessary!

From your comments above and from Wikipedia it seems that your general position is that there will be a literal return of Christ, but most if not all of the stuff the bible describes before that are more along the lines of symbolic representations of timeless patterns of rebellion, unrest, persecution and so on, is that right?

• "The amillennial view regards the "thousand years" mentioned in Revelation 20 as a symbolic number, not as a literal description; amillennialists hold that the millennium has already begun and is identical with the current church age. Amillennialism holds that while Christ's reign during the millennium is spiritual in nature, at the end of the church age, Christ will return in final judgment and establish a permanent reign in the new heaven and new earth."

I don't usually use Wikipedia as a source, though; it doesn't seem very reliable to me. But this much is generally correct, yes.

I think that might be a legitimate interpretation of Revelation. It was written probably in the 90s CE, at a time of (possibly local) persecution during the reign of Domitian.

Well, there is disagreement on this. I would say that the events described in Revelation place its writing before the destruction of Jerusalem (70 A.D.) near the end of Nero's reign in Rome, which would put it in the 67-68 A.D. range. Many, if not most, Bible commentators would agree with me here.

Its seven hills are an obvious reference to Rome...

Well, yes, but Rome, at the time was the principal embodiment of the opulent, "worldly" city of Babylon. So it's really symbolic and not just a reference to Rome in particular, which calls us back to not getting caught up with trying to make 1:1 correlations with concrete things, past, present, or future.

...666 a possible reference to Nero the first great persecutor of the church, and perhaps reference to the Nero redivivus legend in the suggestion of Rev. 17:9-11 that one of the five fallen kings associated with those seven hills (ie, one of the first five emperors, Nero having been the fifth) would return as the eighth (ie, Domitian, supposedly the second great persecutor of the church).

Ugh. No, God named the beast 666 as a symbol of its defining attributes. The number 666 falls short of the divine completeness of seven. When we expect 777, we get a consistent falling short in 666. Thus 666 has an obvious symbolic value. The Beast is the Antichrist, who counterfeits Christ but falls short.

However even if all of that is true it would still have been over sixty years since Jesus' death, over twenty years since the shocking events of the Jewish revolt and destruction of the temple.

No, thirty. Im not sure of your sources in saying Revelation was written in the 90s. I would argue (as would many experts/scholars) that it was written around 67-68 A.D., as I said above. But either way, time really shouldnt be an issue. It may be for you, but its really not.

not to say that everythingwritten by the prophets of the Tanakh was intended literally, but they ought to be understood on a case-by-case basis rather than assuming an overall eschatological framework before even reading them.

I generally agree here, but in taking them on a case by case basis, they are not to be considered independently of and/or in isolation from other parts of the Bible. For example, in reading Joels prophecy, it is not to be considered solely on own, but in relation to what is written in in the Gospels (especially Matthew), Romans (because Paul references Joel 2 in his writings) and what we see in Revelation.

not to say that everything For one example of a possibly allegorical passage, Isaiah 11 is one of the strongest 'messianic' passages in the Tanakh, declaring things like "He will strike the earth with the rod of his mouth, And with the breath of his lips he will slay the wicked" and "the iron shall lie down with the lamp." But read in context of its surrounding chapters, it's possible that this was all merely a hyperbolic expression of exuberance that the young prince/king Hezekiah was/might be more receptive to Isaiah's monotheism than his father Ahaz had been. Possible, though hardly certain even in that case.

This is clearly a prophecy of the coming Messiah. Its not symbolic, and its not allegorical. Isaiah was writing of Jesus; Jesus Himself verified this in several places in the New Testament, notably John 5:46 and in Luke 24. He said Moses and all the prophets wrote of Him. Much of Isaiah is poetic in style, but it is not apocalyptic in genre (as Revelation is).

By contrast, while the predictions of Daniel certainly use a lot of symbolism it's virtually impossible to argue that they were expected to have anything other than a literal, 1 to 1 fulfillment in history: You can literally go through Daniel 11 and match the events to history almost verse by verse, for example.

Agreed on Daniel 11, generally speaking, but this is not apocalyptic in genre. It's not an apples to apples comparison to what we see in Joel 2 or in Revelation. Daniel is interesting, because parts of his prophecy are apocalyptic and parts are not. But this is one part that is not. Yeah, generally speaking, Daniel 11 contains things for which there are fairly immediate fulfilments.

In the case of Joel, there's really no basis for taking it at anything other than face value, as far as I can see: Those various signs would precede the coming of the great and terrible day of the Lord in which there'd be judgement of the nations but escape for those who call on the name of the Lord.

I agree with you here, but apparently don't see it as literalistically (as opposed to literally; hopefully you understand what I mean here) as you apparently do. "Face value" is fine, but I don't believe for a second the moon is literally going to change to blood... or even turn dark red. Do you? So no, the sun being turned to darkness and the moon into blood is a metaphor for events of great disorder and "terrible darkness" and probably great bloodshed to come in history, some of which have occurred, obviously, some of which are occurring now, and some of which are yet to come. And again, Jesus actually refers to this very passage in Matthew 24-25; the signs of the end of the age will not just happen at the very end of history, but throughout. Now, that being said, I do believe that there will be a ramping up of this disorder and terrible darkness near the very end, and possibly a very great ramping up. Is this happening now? Possibly so, but no matter; our place is to be ready at any time.

[Mithrae]

The answer is, unfortunately, highly indeterminate. This is because the missing factor is the probabilistic expectation of the prediction's result prior to confirmation. In other words, before seeing the result what did you think was the probability that the prediction was foreknowledge? In the context of this particular problem it would be a presupposition (in probabilistic form), but in the context of experience in general it may well be the result of valid inductive inference.

Needless to say, such expectations can vary widely. Mine are non-zero, but still trivially small, like 0.00001%. Yours are likely considerably more generous (I won't guess exactly how much). And a presuppositionalist Christian's expectations would be at or near 100%.

Sure, if someone holds a dogmatic presupposition that predictive prophecy cannot occur, then they will not be persuaded of its legitimacy no matter how powerful an example may be. If someone is only very marginally more open-minded then pretty much the same is true. I don't think there's any question about that.

So the question is what is a reasonable expectation for the legitimacy of prophecy?

Going back to that passage in Daniel 11 which I mentioned earlier, I think it's safe to say that if you were certain that it was written before the events, even with a 0.00001% anti-prophetic starting point you'd still be forced to unequivocally conclude that it's genuine predictive prophecy: The odds of it being chance are something like one quadrillion to one even by a fairly conservative estimate (simply with the generous assumption that each of the fifteen main 'events' highlighted had a 10% likelihood). So in that case the only question is whether it actually was written before the events predicted; and short of the circular assumption that it must have been written later because of the prophecy, there really are good arguments both ways. Absent that presupposition, it would be fair to assume only a 50 or 40 or even 30% likelihood that it was written before the events.

Of course a 30 or 40 or 50% confidence that Daniel wrote genuine predictive prophecy doesn't translate directly into a 30 or 40 or 50% prior plausibility for prophecy by other authors. But then we go back to the point that there are actually more than just one or two remarkable predictions by Hebrew prophets in the biblical canon. Many vague predictions have been 'fulfilled' and there remain some really quite remarkable ones:

• > Daniel's rough time-frame that a great messiah would be 'cut off' and prediction that Jerusalem and the second temple would be destroyed by the people of a future kingdom shortly afterwards;
  > the seemingly absurd boast of that soon-to-be-executed backwater peasant that his words would outlast heaven and earth, now by far the most-printed, most-translated and most-widely known words of human history;
  > the fall and eternal desolation of what at the time were essentially the two greatest cities in the world, Nineveh and Babylon, as predicted by Isaiah, Nahum and Zephaniah;
  > the survival of the Jewish people when so many mighty oppressors fell by the wayside, and even more remarkably the return to their homeland as prophesied and eventual flourishing in spite of their much larger neighbours' extermination efforts...

So given this information what is a reasonable prior expectation regarding canonical Hebrew prophecy? 10%? 5%? Qualitatively that would be saying that a prediction is very, very unlikely to be genuine foreknowledge - almost certainly not! What about 1%? Honestly I'd have to say that if you're assuming it to be that low without comprehensively refuting the information suggesting its plausibility, we'd be starting to tread dangerously close to the realm of dogmatism.

The final piece of the puzzle, though, that I couldn't figure out on my own, was how the prior expectation fit into the problem mathematically. I did some research, and found that Bayes' Theorem does the job.

https://en.wikipedia.org/wiki/Bayes%27_theorem

I plugged in some values (correctly I hope) and got some very satisfying results.

I'm not sure what you've done there. Using Bayes' theorem as stated in the link -
P(F|O) = P(O|F)*P(F) / P(O)
- where P(F) is probability of foreknowledge, P(O) is probability event occurs, P(O|F) is probability it occurs given foreknowledge (99-100% since if it's foreknown it must occur) and P(F|O) is probability it's foreknowledge given occurrence. For a 1% prior expectation of foreknowledge P(F) and 1% probability the event occurs P(O) I get a 99-100% probability that it's foreknowledge given occurrence P(F|O):
P(F|O) = 0.99*0.01 / 0.01 = 0.99

Your result for the same expectation of foreknowledge and likelihood of event is only 50%, so it seems one of us is doing something wrong.

For my own convenience I've put Bayes' formula and your formula together in a spreadsheet, with the blue cells comparing/contrasting the two formulas while (if I've set the permissions right) anyone should be able to edit the grey cells.



Edit: It seems you might be the formula further down the page (https://en.wikipedia.org/wiki/Bayes%27_ ... ative_form), but I can't see how that would be legitimate if so:
P(A) is the prior probability, is the initial degree of belief in A.
P(A|B) is the posterior probability, is the probability for A after taking into account B for and against A.

Obviously what we'd be trying to find here is P(A|B) when A is the genuineness of prophecy. But the problem is here:
P(B|A)is the conditional probability or likelihood, is the degree of belief in B, given that the proposition A is true.
P(B|~A) is the conditional probability or likelihood, is the degree of belief in B, given that the proposition ~A is true.

You're using the "probability of event NOT occurring" as P(B|A), but the probability of the event not occurring doesn't change regardless of whether A is true or false (whether prophecy is genuine or not). P(B|A) should always be the same as P(B|~A); the event's likelihood is the same either way. It seems that when P(B|A) and P(B|~A) are the same, the end result P(A|B) is always the same as P(A), meaning that we haven't learned anything at all (ie, it's not a legitimate use of the formula).

That's if B is an independent variable, remaining the same regardless of A. On the other hand if it's a variable which is somehow dependent on the truth or falsehood of A (plausibility of prophecy), then when prophecy is ruled out completely in P(B|~A), the value will always be zero. That in turn would mean that the final result P(A|B) will always be 100%, since it'll be
P(B|A)*P(A) / ( P(B|A)*P(A) + P(B|~A)*P(~A) )
= P(B|A)*P(A) / ( P(B|A)*P(A) + zero*P(~A) )
= P(B|A)*P(A) / P(B|A)*P(A) = N/N = 1

Either way it doesn't seem legitimate applying that formula to this question.

[FarWanderer]

The final piece of the puzzle, though, that I couldn't figure out on my own, was how the prior expectation fit into the problem mathematically. I did some research, and found that Bayes' Theorem does the job.

https://en.wikipedia.org/wiki/Bayes%27_theorem

I plugged in some values (correctly I hope) and got some very satisfying results.

I'm not sure what you've done there.
[...]
Either way it doesn't seem legitimate applying that formula to this question.

Yes I was wrong. But fairly close. I've figured it out. It's rather simple actually.

You take the belief prior, say 40%, and add to it the chance of coincidence (1%) times the prior's compliment (60%) for 40+0.6=40.6%. Then after the prediction is fulfilled 40.6% "becomes" 100%, and you stretch out each portion (40%,0.6%) accordingly.

40/40.6 = ~98.5% Chance it was foreknown
0.6/40.6 = ~1.5% Chance it was coincidence

The answer is, unfortunately, highly indeterminate. This is because the missing factor is the probabilistic expectation of the prediction's result prior to confirmation. In other words, before seeing the result what did you think was the probability that the prediction was foreknowledge? In the context of this particular problem it would be a presupposition (in probabilistic form), but in the context of experience in general it may well be the result of valid inductive inference.

Needless to say, such expectations can vary widely. Mine are non-zero, but still trivially small, like 0.00001%. Yours are likely considerably more generous (I won't guess exactly how much). And a presuppositionalist Christian's expectations would be at or near 100%.

Sure, if someone holds a dogmatic presupposition that predictive prophecy cannot occur, then they will not be persuaded of its legitimacy no matter how powerful an example may be. If someone is only very marginally more open-minded then pretty much the same is true. I don't think there's any question about that.

So the question is what is a reasonable expectation for the legitimacy of prophecy?

Going back to that passage in Daniel 11 which I mentioned earlier, I think it's safe to say that if you were certain that it was written before the events, even with a 0.00001% anti-prophetic starting point you'd still be forced to unequivocally conclude that it's genuine predictive prophecy: The odds of it being chance are something like one quadrillion to one even by a fairly conservative estimate (simply with the generous assumption that each of the fifteen main 'events' highlighted had a 10% likelihood). So in that case the only question is whether it actually was written before the events predicted; and short of the circular assumption that it must have been written later because of the prophecy, there really are good arguments both ways. Absent that presupposition, it would be fair to assume only a 50 or 40 or even 30% likelihood that it was written before the events.

It's only pseudo-circular. It's perfectly normal to make inductive inferences about a thing like prophecy. The question is what's more plausible? That everything you think you know about prophecy is false, or that the 60% likelihood that he wrote after the events is actually what happened?

You can apply the same formula to it as above, just replacing the coincidence probability with the post-event-writing probability. Against 60%, a 1% prior only rises to about 1.6%.

Of course a 30 or 40 or 50% confidence that Daniel wrote genuine predictive prophecy doesn't translate directly into a 30 or 40 or 50% prior plausibility for prophecy by other authors.

Heh, it would boost immeasurably for someone like me, if I accepted such a thing. But setting that aside...

What you are touching on here is probably the trickiest part of these induction problems. What things are "like" each other to rightfully serve as samples from much to make predictions about other things in the same category, and what things are not "like" and should not be used as such samples? There seems no straightforward way to settle the question, as far as I can tell.

But then we go back to the point that there are actually more than just one or two remarkable predictions by Hebrew prophets in the biblical canon. Many vague predictions have been 'fulfilled' and there remain some really quite remarkable ones:

Honestly, I can't examine all of this stuff. I just don't have the energy. Or the time.

So given this information what is a reasonable prior expectation regarding canonical Hebrew prophecy? 10%? 5%? Qualitatively that would be saying that a prediction is very, very unlikely to be genuine foreknowledge - almost certainly not! What about 1%? Honestly I'd have to say that if you're assuming it to be that low without comprehensively refuting the information suggesting its plausibility, we'd be starting to tread dangerously close to the realm of dogmatism.

It's way lower than 1%. But I wouldn't confuse certainty with dogmatism.

[Mithrae]

Sure, if someone holds a dogmatic presupposition that predictive prophecy cannot occur, then they will not be persuaded of its legitimacy no matter how powerful an example may be. If someone is only very marginally more open-minded then pretty much the same is true. I don't think there's any question about that.

So the question is what is a reasonable expectation for the legitimacy of prophecy?

Going back to that passage in Daniel 11 which I mentioned earlier, I think it's safe to say that if you were certain that it was written before the events, even with a 0.00001% anti-prophetic starting point you'd still be forced to unequivocally conclude that it's genuine predictive prophecy: The odds of it being chance are something like one quadrillion to one even by a fairly conservative estimate (simply with the generous assumption that each of the fifteen main 'events' highlighted had a 10% likelihood). So in that case the only question is whether it actually was written before the events predicted; and short of the circular assumption that it must have been written later because of the prophecy, there really are good arguments both ways. Absent that presupposition, it would be fair to assume only a 50 or 40 or even 30% likelihood that it was written before the events.

It's only pseudo-circular. It's perfectly normal to make inductive inferences about a thing like prophecy. The question is what's more plausible? That everything you think you know about prophecy is false, or that the 60% likelihood that he wrote after the events is actually what happened?

On what basis do you think you know anything about prophecy one way or the other? Reasonable scepticism is one thing, but a claim of possessing certain knowledge to the contrary seems singularly unfounded as far as I can tell. It's not even as if we're talking about something which violates the supposed 'laws of nature'; certainly not in the biblical sense of a powerful being merely revealing certain events he's going to bring to fruition in the future. If you had some kind of overwhelming proof that predictive prophecy is impossible I'm sure you would have shared it long, long ago in the thread.

If the question under consideration were merely "When was this written?" then reasonable scepticism of predictive prophecy might have a legitimate bearing on how we answer the question. But for the question "Is predictive prophecy ever genuine?" then it is clearly and unequivocally circular to incorporate the view that it is never genuine into your consideration of the evidence. Without that circularity there's a good case to be made either way, suggesting perhaps a 30-70% possibility that it was written before the events described. I figured you'd consider only the lower end of that range, but not that you would insist on using the circular reasoning!

The final piece of the puzzle, though, that I couldn't figure out on my own, was how the prior expectation fit into the problem mathematically. I did some research, and found that Bayes' Theorem does the job.

https://en.wikipedia.org/wiki/Bayes%27_theorem

I plugged in some values (correctly I hope) and got some very satisfying results.

I'm not sure what you've done there.
[...]
Either way it doesn't seem legitimate applying that formula to this question.

Yes I was wrong. But fairly close. I've figured it out. It's rather simple actually.

You take the belief prior, say 40%, and add to it the chance of coincidence (1%) times the prior's compliment (60%) for 40+0.6=40.6%. Then after the prediction is fulfilled 40.6% "becomes" 100%, and you stretch out each portion (40%,0.6%) accordingly.

40/40.6 = ~98.5% Chance it was foreknown
0.6/40.6 = ~1.5% Chance it was coincidence

You linked to Bayes' theorem, but that's not what you're using. Why not? And what's the rationale behind whatever it is you are using?

As I showed, with a borderline unreasonable level of scepticism against canonical Hebrew prophecy (presupposing a 99% certainty that it's bogus), in the case of a 1 in 100 'coincidental' fulfillment applying Bayes' theorem would give us a ~99% confidence that it is genuine. I'm not sure how legitimate that application is, but it's the one you suggested and instead you now seem to be opting for something else which you haven't really explained.

But then we go back to the point that there are actually more than just one or two remarkable predictions by Hebrew prophets in the biblical canon. Many vague predictions have been 'fulfilled' and there remain some really quite remarkable ones:

Honestly, I can't examine all of this stuff. I just don't have the energy. Or the time.

So given this information what is a reasonable prior expectation regarding canonical Hebrew prophecy? 10%? 5%? Qualitatively that would be saying that a prediction is very, very unlikely to be genuine foreknowledge - almost certainly not! What about 1%? Honestly I'd have to say that if you're assuming it to be that low without comprehensively refuting the information suggesting its plausibility, we'd be starting to tread dangerously close to the realm of dogmatism.

It's way lower than 1%. But I wouldn't confuse certainty with dogmatism.

You've provided no proof or compelling evidence to suggest that predictive prophecy is impossible - and I'm sure you would have provided it long ago if you'd had any!

You've also been unwilling - at every turn at which you have broadened the question to the prior plausibility of prophecy - to address any of the quite remarkable examples I've raised (besides a somewhat arbitrary declaration that the survival and return of the Jews was a 'self-fulfilling' prophecy).

For my part I've been happy to acknowledge that reasonable scepticism is entirely legitimate, but you're insisting that a 10% or even 1% possibility of prophecy is far too high, without a shred of justification for your claim. I wouldn't say that you are dogmatic; it's a big subject to broach after all, though you're the one who's broached it. All I can say is insofar as this discussion goes, dogmatic insistence seems to be all that you've offered.