Imagine you have a line of 100 friends playing a game of “telephone” in which the first person in line whispers a message to the second person, who in turn whispers the message to the third person…until the message reaches the hundredth person.

How do you assess the probability that the message will be transmitted without error through the 100-person chain? There are two possible methods.

- Reason the probability of each of your friends either mishearing or intentional distorting the message to be negligible, and multiply zero by 100 to arrive at the conclusion there is zero chance the message you receive at the end of the chain is corrupted.
- Reason the probability of each of your friends either mishearing or intentionally distorting the message to be slight, and calculate that slight probability, say 1%, over 100 events (0.99^100) to arrive at the conclusion that there is only a 36.6% chance of an uncorrupted message.

This failure to perceive this substantial probability of corruption (63.4%) intrinsic to a long series of events, each with low probability of corruption, seems to be at the heart of Christian illogic when considering the probability of a human resurrection based on the evidences available.

In this case, the probabilities of “corruption” are not chronologically linear, but remain probabilistically additive, and can therefore be calculated the same way.

Imagine a case in which you have an unreliable friend Tom who has 1) a mendacious character, 2) an appetite for hallucinagens, 3) a gullibility towards pranksters, and 4) a poor understanding of standards of evidence. Tom tells you his dog rose from the dead. You consider alternative explanations. For the sake of simplicity, imagine that, in this scenario, the probability of 1) a lie, 2) a hallucination, 3) a deception, and an unconsidered natural explanation are all 10%. We don’t, of course, suppose that the rather low probability of each of these contingencies warrants concluding his dog most certainly did rise from the dead. We instead calculate the final probability his claim is true given all the alternative explanations. This would be 0.9^4, or about 65%.

However, does an honest assessment stop here? Do we therefore conclude that there is a 65% probability that Tom’s dog rose from the dead? No.

We have only assessed the probability that Tom’s claim that his dog resurrected is wrong based on the alternative explanations we can identify. We have not yet assessed the probability of a dog rising from the dead. And this is a step that many Christians seem to ignore in respect to the alleged resurrection of Jesus.

Though we can not demonstrate with any degree of certainty that any of the four possible alternative explanations for the apparent resurrection of Tom’s dog we have introduced are true, we don’t conclude that their individual unlikelihoods, even when properly combined to arrive at the 65% probability Tom is a reliable source for any claim, demand that we place a 65% degree of certainty in the notion that Tom’s dog did, in fact, resurrect.

We very seldom see dogs resurrect. I don’t think it would be a stretch to place, for my readers, the probability of a dog actually resurrecting based on our inductive experience to be far below one in a million.

So though the assessment of the probabilities of the alternative explanations we have identified suggests Tom is still more than likely to be relating a truth, we intuitively understand that such an assessment is incomplete until we include the probability of a dog rising from the dead.

We now need to reflect on disparity between our 65% degree of certainty we have in Tom based on the alternative explanations we have thought of, and our induction-based 0.0001% degree of certainty that a dog can rise from the dead. This, if we are rational, will lead us to conclude there may also be other unknown possible explanations that we have not thought of.

Perhaps these explanations are of such a nature that we could not possibly currently think of them such as in the non-demonic neurological explanation of epileptic fits unavailable to us prior to scientific knowledge of neurology. In that context, demon possession (an explanation reenforced by other contemporary “supernatural” events such as lightening) would have seemed very likely absent the understanding there may be natural explanations we may not yet have considered or have access to.

We all correctly conclude that even multiple claims from even quite reliable sources do not trump our inductive assessment of the probability of physical events. If Tom and ten others of high reputation tell us Tom’s dog resurrected, we rightfully remain quite skeptical, even when there is no alternative explanation that comes to mind.

I strongly recommend Christians do this when considering the claim of a resurrected Jesus.

First, we should properly calculate the probabilities of the alternative explanations we can think of. A partial list is as follows.

- The entire story was fabricated by one or a few individuals.
- The resurrection emerged after many retellings of the life of Jesus.
- The body was never in the tomb or was stolen.
- The events were hallucinated.

Note that we shouldn’t make the error that some do and ignore remote probabilities of individual events that become a substantial probability when the probability of those events are combined.

(Something else to consider is whether you would believe an evidentially equivalent story from an non-biblical source.)

Once we have determined the probability of the Biblical account being in error, we then more importantly weigh that probability against the probability that someone will rise from the dead based on our inductive experience. It is this step that is far too often missing from the analysis of the Christian.

This becomes quite salient when we consider the Muslim claim that Muhammed split the moon. We may consider the reliability of the Koran and collaborating sources, but we rightfully more importantly simply consider the physical possibility that the moon can be split. The same holds for Muhammed’s flying on a horse. The fact that we have never seen a horse fly makes an assessment of the reliability of the eye-witnesses quite unnecessary.

We don’t need to waste much time assessing the reliability of the sources of such claims when the claims run so counter to our inductive experience. So also for claims of the resurrection. For a Muslim to suggest we must give an account of all the eye-witnesses to these miracles or believe those miracles happened is silly. So also for the claims of the resurrection.

If Tom, as unreliable as he is, tells me his dog chased a rabbit yesterday, I may still be rational in giving him the benefit of a doubt. But if he claims his dog rose from the dead, it does not fall on me to come up with a definitive alternative explanation that would explain his claim. I can dismiss his claim of a resurrected dog as highly improbable and get on with life without knowing what the actual explanation is.

This recent attempt by apologists to shift the burden of proof to those who are highly doubtful a man named Jesus came back from the dead is silly. Those same apologist employ different standards of evidence to reject equally evidenced claims of resurrecting dogs, splitting moons and flying horses.

I’m willing to accept that many Christians have never given the proper assessment of probabilities in such context much thought. But it does appear some apologists intentionally 1) avoid combining improbable explanations that suggest the far more probable conclusion a miraculous claims is in error, and more importantly 2) dismiss the essential step of inductively assessing the probability of the actual event in question.

A source with a fair degree of reliability can claim a) a dog caught a rabbit, and b) a dog resurrected. We justifiably accept (a) and reject (b) if we understand the proper and honest way to assess such claims.