In the Craig vs. Hitchens thread, NotReady linked to some debates that William Craig had been in. I read one of the transcripts through and we had a brief discussion about it in that thread. Here is that debate if you're interested.

What I was most interested was the part where Craig brought in Bayes's Theorem as one of his arguments for the resurrection occurring. For those of you who don't know, Bayes's Theorem is useful when trying to determine the cause of a certain event. In other words, given an observed result, Bayes's Theorem can be used to determine the probability that the result was caused by a given event.

In the debate, Craig gives a talk about Bayes's Theorem, but never once mentions a specific number. At the end of the debate there is a Q&A, and one of the audience members asks Craig to give some numbers for his calculation. Craig only references Richard Swineburne of Oxford, saying that he calculated a 97% chance that Jesus was resurrected. Being naturally curious about this calculation I decided to look it up and I was able to find a good review on the book and the calculation. Here it is. His assumptions are quite humorous, but they are not the point of this thread.

Basically what Craig says during the debate is completely different than what Swineburne did. Doesn't say a lot for his honesty but again, its not the point of this thread. But because math is awesome, I thought we here at RGT could give Dr. Craig a hand and fill in some numbers in his formula, with the help of Dr. Swineburne of course. Below is the formula and the key:



Let's be very kind to Dr. Craig and make the probability that we'd have the evidence we do given that the resurrection occurred = 1. So Pr (E|B & R) = 1. First we need to find Pr (E|B & not-R). Dr. Swineburne has been kind enough to help us with this one, and he views the probability of having the kind evidence we do given no resurrection equal to 1/1000.

So now all we're missing is Pr(R|B), since finding that gives us Pr(not-R|B), as they are complements. When using Bayes's Theorem, the probability of an event given just the background is the prior probability of that event, absent our evidence. So in our case, the prior probability of the resurrection is what we think the chances are that Jesus was resurrected, while knowing nothing about him. Our mileage may vary when assigning this probability, but I think it’s safe to say that the probability of a random person coming back from the dead is very very low. Just for ****s and giggles, let’s put the probability of an unknown person being resurrected at 1/1,000,000. That would mean that roughly 30 people alive in the US today will come back to life after they die! But I’m feeling generous so let’s do this and see what kind of number we come up with.
Pr (R|B&E) = [(.000001)*(1)] / [(1/1,000,000)*(1) + (.999999)(.001)]
= ~.0001

The point of this thread is to show why the evidence of the resurrection should not cause an objective observer to believe that it happened. If theists want to convince me that Jesus rose from the dead, they’re going to have to do a lot better than 2,000 year old second-hand sources. A LOT better.

Justin -

I think you are making a massive error by assuming that the chance of "The Son of God" (who's purpose in the universe is to go to Earth, get crucified, go to the gates of hell, then get resurrected....or something) getting resurrected is the same as a normal human.

Josh

Justin,

With all due respect, this is a ridiculous thread and you know why.

Basically, what you've done is that by assuming that Jesus is just like an ordinary person, you reach the conclusion that Jesus was probably just an ordinary person. Congratulations, you affirmed your hypothesis.

By the way, God makes it clear that we will not see any other signs than the Resurrection, "the sign of Jonah."

So, if you don't believe in the Resurrection, then perhaps you should exhaust your entire search to deciding whether the event is true. It is not up to the theists on this board to "prove" it to you. You're the one who has to decide and you've decided that you need to look through the evidence.

Since you're facing the potential of eternal bliss/damnation, maybe you should delve through the evidence and research the sources and accounts yourself?

I haven't read the debate so I don't know if you misrepresented Craig or not, but for a better explanation by him of probability re the resurrection, go here:

http://www.reasonablefaith.org/site/...odcasting_main

Under the Defenders Podcast list is a series on the Doctrine of Christ. He deals with Bayes in #17 and #18. BTW, about the last 5 or so are on the resurrection and if you want to hear a powerful argument for it's factuality, listen to all of them.

At any rate, in the last one he states that he doesn't like probability arguments for the resurrection. The reason is that Bayes is useful for (random?) events like the big bang causing the fine tuning necessary for life and would apply if the resurrection was a random event caused by natural law, in which case he agrees the probability would be virtually nil. But we don't claim that in the case of Christ a dead person came back to life through natural means - rather, God resurrected Him, i.e., it's a miracle, a supernatural event. As Craig says, there's no realistic way to apply probability to the supernatural.

A long time ago, when these religious discussions first began, DS tried to make a case that the events of the Bible and the existence of the Christian God was almost zero due to probability. I basically disagreed with him for the same reasons Craig asserts - you can't apply probability to a non-random event, especially if it's an act by God. There's no sample space, no prior probability, and the idea only applies to naturalistic events.

Perhaps in the debate Craig wasn't clear about it - I don't think you can call him dishonest though because there is other material by him that makes his position clear. You have to give people some slack when they say things in the heat of a live debate and what they say has to be taken in the context of all their published work. I would give the same slack to an atheist.

Edit: BTW, in #18 he mentions Swinburne's calculations and states they are fun to play with but are not to be taken seriously.

Assuming Jesus is the Son of God would be a terrible error. Christians believe Jesus was the Son of God mainly because of the resurrection. The resurrection and the miracles (with which the same logic applies) are the evidence for his divinity. Assuming the divinity and then figuring out probability is just begging the question in a big way.

I basically said the same thing to Josh, but let me restate. Of course I start by assuming Jesus was an ordinary person. Isn't it fair to say that without the evidence presented in the Bible you would agree? You need some evidence as a starting point to shift your view of a person from ordinary to divine. All I have done is shown that there is no evidence strong enough to do so. Do you see the circularity needed to show that Jesus was the Son? You need to start by assuming he is divine!

So this affirms the point I made to Josh that the resurrection is the evidence for Jesus being divine, so you can't assume he is divine when discussing the validity of the evidence.

Why would I keep searching? I've done that extensively in the past and decided it was an old myth. The point of my post is to show you exactly what the first few posts have said, that you can't come to believe in the resurrection without first assuming divinity.

NR I will get to your post later I have class for the next five hours.

Well, if you've decided, you've decided but, keep in mind this caveat: if the resurrection is true, then the rest of the Bible is also likely to be true...so you're going to have to deal with the implications of an entity that is very malicious to the truth. This can be a daunting proposition.

Circular, an appeal to fear, and a false dichotomy. Impressive.

So what? It's either true or it's not.

If you're right about the resurrection, then your deep study and exploration of the subject contributes to your personal knowledge, possibly the knowledge of society, and deepens your understanding of the world as we know it (a world which cannot be understood without understanding religion). So, you really haven't wasted your time, have you?

You write a highly acclaimed book and go on lecture tours with Hitchens and the like (you obviously like doing this anyway).

If you're wrong about the resurrection, then you've just found out the single most important piece of knowledge that you can ever possess. You write a book that is highly-touted in the Christian world and you go on the Christian lecture circuit.

Consider it Pascal's Wager with a monetary incentive. I just gave you the path to being a millionaire intellectual either way.

Your hypothesis, please?

No.

How many people from that era started religions that lasted 2000 years? How many of them had followers claiming that they were coming back from the dead? How many of them have several thousand copies of documents testifying to their actions? How many of them were important enough that basically every major religion has some sort of position on who he is?

An objective person does not reach the conclusion that Jesus is an ordinary person. Whether he is divine is another matter, but he was certainly not ordinary.

Besides, your assumption isthat people coming back from the dead is a natural event. It isn't. You cannot use your model in a reasonable way here.

This is all evidence for the resurrection that can be taken into account, but it does not factor into prior probabilities. In Craig's formula, this is the Pr (E|not-R & B). If you think 1/1000 is too high then that's a different story, but all the stuff about how Jesus started a huge religion is all just evidence we have about Jesus, and it's not background evidence.

Also I should have chosen my words better. I meant isn't it fair to say that without all of the evidence you have about Jesus that we should assume he's ordinary.

I meant ordinary as opposed to divine. Besides, if we didn't have any evidence about Jesus we'd have to start out assuming he was ordinary given either definition of the word.

This is not true. If I were to assume that then the prior probability of resurrection would be effectively zero. I'm assuming probability of a resurrection by any means. Also it's not my model, it's a Christian apologist's model. I'm just plugging in some numbers based on what I think are generous assumptions.

The resurrection is being presented as evidence that Jesus is the son of God. Thus, the event is not "the son of God being resurrected," but "some guy being resurrected." The details such as being born in Bethlehem and having a book written about him are covered in B, and details such as the failure to positively identify a "body of Jesus" in any tomb are covered in E. The claim that Jesus was the son of God is not included in either, it extends from R and thus can't be used in a calculation of the probability of R given B and E.

Compare the bolded sentence to

You're assuming that the same means of Jesus' resurrection would be equally applicable to people today. In other words, you assert a sort of 'naturalistic' model in your claim despite the fact that the claim about Jesus is that his resurrection was not natural.

FWIW - I'm not supporting the model, either. I just think the whole thing is bad math. It reminds me of the Bible code people. I think they're full of some pretty bad math, too.

Can I find transcripts anywhere? I get extremely bored listening to someone talk and I probably won't be able to make it through those.

If he really thinks that then it's really weird that he brought up this argument in the debate. This was not a heat of the moment thing, he had slides prepared to illustrate it with titles like "Ehrman's Egregious Error" (I thought that was clever).

But you can apply probability to a non-random event. We do it because of our imperfect knowledge about a situation. For instance in live poker, once the deck has been shuffled the flop turn and river cards are no longer random. They are based on the exact ordering of the cards. We can still apply probabilistic arguments to them because of our lack of information about that ordering.

The same applies to supernatural event. Our lack of information about reality means we need to assign some sort of probability based on our limited knowledge. We can do this based on the evidence we have about the event in question along with a rough idea of how often supernatural events occur (of course this estimate will vary greatly from theist to atheist, but in Craig's model I tried to use numbers that were well on the theist side of things).

To put it another way, this whole thing is just a fancy way of saying extraordinary claims demand extraordinary evidence. You believe that there is extraordinary evidence. Because of your personal experiences along with the evidence you've seen, you've come up with a probability for the resurrection that is somewhere close to 1. All I'm trying to do is show that just from an objective base of looking at the evidence presented in the writings from the time, there is not nearly enough evidence to believe the resurrection occurred.

He definitely planned before the debate to bring this stuff up. If he really thinks probabilistic arguments cannot be used for the resurrection then this was certainly dishonest. I do give him some slack for the Swineburne reference since that was during the Q&A session.

Fair enough, although I wonder what Swineburne would say if he heard that.

Looks like they're all audio. You really should listen to the last 2 as he addresses many of the issues you're raising.

I'm sure he planned to bring it up to counter something he expected from Ehrman - he's stated before he prepares extensively for his debates.

But it's random for us which is what is meant. Laplace - if we knew everything we could calculate all events(or something like that). But we use probability to fill in our ignorance, which is randomness for us.


The extraordinary evidence idea is from Hume. Craig mentions a book by Earman called "Hume's Abject Failure" which is a detailed refutation of this idea. There are several papers on the net about Earman's book which are interesting, some claiming Earman is wrong.

I read somewhere a long time ago that Swinburne was being a bit tongue in cheek with the book. He was just showing that if you use probability it can be made to support the existence of God, the resurrection, etc. The whole thing comes down to prior probability, and that of course is an impossible number to assign. I once asked DS about his pp and got no answer.

One of the points of this whole exercise is to show that Hume didn't consider how other factors besides the initial improbability of an event should affect our evaluation of it's likelihood. The nature of the documents, the empty tomb, the testimony of the disciples, the prophecies fulfilled, the failure to counter the resurrection claim by people with both the motive and power to do so, and others.

There is no evidence that Jesus was the son of God. Therefore it is reasonable to say that the chance of him be resurrected is the same as any other tom dick and harry. 0

The example about the people today was just to show how high that number is.

If we are defining a resurrection event, by it's nature if we assign it a probability greater than 0 then we are assuming supernatural causes are allowed. How about if I frame the question this way, what are the chances that a random person will be supernaturally resurrected by God? My starting assumption that Jesus was a random person is valid here. In order to show that Jesus was not ordinary, we must first assume he is and then take into account the evidence to the contrary. If the evidence is strong enough then we can say he was not ordinary. But we can't use the idea that he is the Son of God when we're evaluating the evidence for his being the Son of God.

The math is good. It's the assumptions that might be bad. That's what this this thread is all about.

Ok I'll try to do that eventually. I'm taking way too many courses and teaching as well so it could be a while. I think it'll be worth it if I post a review afterward so we can get some discussion going.

Do you mean to say that miracle events are not random for us, despite our ignorance?

All of those pieces of evidence are very important. What's more important is how strong that evidence is compared with how improbable the event is. I've never read Hume but it seems obvious that prior probabilities need to be evaluated in conjunction with all of the evidence available.

It should be pretty clear that your assumption about the probability of a "random" person being "supernaturally resurrected by God" (or however you want to frame it -- it doesn't really matter) is pulled out of a convenient bodily orifice, bringing a certain level of pointlessness to the entire model.

You're still facing the battle of a "random" act (what I called 'naturalistic' in the previous post) against the acts of a sentient being.

This thread is about making bad assumptions? You could have fooled me:

How do you figure? Pick a random person from the last five thousand years. What is the probability that that person was resurrected after death? Of course once we know someone has been resurrected they are no longer random/ordinary, but we're talking about prior probabilities.

See my post to NR. Our lack of information causes it to appear to us to be a random event. It doesn't matter if the event was actually random or not.

Sigh. I meant the thread is about discussing assumptions that I've put forth in my argument, just as we've been doing.

I think we're going to reach a point where we simply disagree with the use of Bayesian analysis for this type of situation. Under this scheme, you would conclude that any extreme outlier from any sample probably didn't actually exist.

such a silly post

Only in the sense that I can take any poker flop and conclude that it probably didn't happen.

The probability that the board goes A78TA is quite small; that doesn't imply that it didn't just happen.

The straightforward application of Bayes' theorem applies in cases of limited information. That is its nature. Obviously, if you identify any unlikely event (I'm going to assume "extreme outlier" implies an unlikely event), then the independent probability of that event happening is quite small. That's what it means for an event to be unlikely. If an extreme outlier didn't have a low probability, then you'd be ****ing up the math.

If you don't know what my poker hand is, it would be foolish of you to bet that it's A78TA. It would be wise to bet against that particular board. Obviously post-hoc I can say you would have lost if you did that in my last hand, but that kind of post-hoc reasoning isn't statistically valid. And it's exactly the type of reasoning you're trying to apply to the resurrection of Jesus.