Nineteen Names and a Historical Starting Point
“That Jesus’ followers (and later Paul) had resurrection experiences is, in my judgment, a fact. What the reality was that gave rise to the experiences I do not know.”
— E. P. Sanders (scholar), The Historical Figure of Jesus (1993)
If you were serving on jury duty and not one, but nineteen individuals were brought before the court as named eyewitnesses, each independently placing the defendant at the scene of the crime, would you consider that strong evidence that the defendant was, in fact, there? Especially if those witnesses came from different backgrounds, testified at different times, and had nothing obvious to gain by their testimony?
In ordinary life, we would not dismiss such evidence lightly. We would weigh it, test it, and ask whether alternative explanations (collusion, mistake, deception) were more plausible than the conclusion the witnesses themselves drew. That is precisely the kind of reasoning required when we approach the resurrection of Jesus.
The resurrection of Jesus is not introduced in the New Testament as a vague spiritual experience or an anonymous rumor. It begins with names: nineteen identifiable individuals who are either explicitly stated or directly included in post-crucifixion appearances of Jesus. Mary Magdalene, Mary the mother of James, Salome, Joanna, Peter, Andrew, James son of Zebedee, John, Philip, Bartholomew, Matthew, Thomas, James son of Alphaeus, Simon the Zealot, Judas son of James, Cleopas, James the brother of Jesus, Paul, and Nathanael form the initial core of resurrection witnesses. Before appealing to unnamed individuals or large groups, historical reasoning requires that we stop and ask a foundational question: what do these nineteen people have in common that could plausibly explain their shared testimony?
What quickly becomes clear is that they do not share a single psychological profile, social role, or prior expectation that would naturally lead to a uniform claim. Some were devoted followers, others were fearful deserters, at least one was a skeptic who rejected Jesus as Messiah during His lifetime, and one was an active persecutor of the early Christian movement. They encountered Jesus at different times, in different places, and under different circumstances. And yet they converge on the same core claim: Jesus, who had been publicly executed, was later encountered alive.
“What is certain is that Jesus’ disciples believed that he appeared to them after his death.”
— Bart Ehrman (agnostic scholar), The New Testament: A Historical Introduction (2016)
The Explanatory Options on the Table
When the evidence is taken seriously, the range of plausible explanations narrows quickly. The question is not whether alternative explanations are logically possible, but whether they are probable given the kind of evidence we actually have. At bottom, the options reduce to a small and manageable set:
• Deliberate fabrication — the witnesses knowingly lied about seeing Jesus alive after His crucifixion. This explanation requires sustained deception across numerous individuals, consistency over time, and silence under pressure, all without any clear incentive and in the face of persecution.
• Sincere mistake or misidentification — the witnesses genuinely believed they saw Jesus but were mistaken. This must account for repeated encounters, recognition by close companions, extended conversations, physical interaction, and the inclusion of skeptics and hostile opponents who were not psychologically primed to believe.
• Deception or hoax by others — the witnesses were fooled by an external scheme. This requires a missing body, successful long-term coordination, no exposure by enemies, and no internal collapse of the deception, despite strong motivation to expose it.
• Psychological phenomena — hallucinations, delusions, or grief-induced experiences. This must explain convergence across many individuals, consistency of content, group encounters, physical interaction, and the transformation of skeptics and opponents, features that do not align with how hallucinations are known to function.
• Actual resurrection appearances — the witnesses encountered Jesus alive after His death. This explanation predicts multiple witnesses, varied settings, skeptics converted, opponents transformed, early public proclamation, and appeal to living eyewitnesses.
The force of the resurrection case does not come from asserting that alternative explanations are impossible. It comes from comparing them and asking which explanation best accounts for the totality of the evidence with the fewest additional assumptions.
This is where Bayesian reasoning becomes essential. Bayesian probability does not ask whether a resurrection would be extraordinary. It asks a more precise and historically responsible question: given the specific pattern of evidence we have, which explanation makes that evidence more likely? Even a hypothesis with a low prior probability can become rationally compelling if its rivals require multiplying independent improbabilities to survive.
“I know in their own terms what they saw was the raised Jesus. That’s what they say, and then all the historical evidence we have afterwards attests to their conviction.”
— Paula Fredriksen (Jewish scholar) Interview, PBS Frontline: From Jesus to Christ
Why Bayesian Probability Matters Here
Bayesian reasoning weighs hypotheses by comparing how well they predict the observed evidence. Even if one assigns a low prior probability to miracles, a hypothesis can still become rationally compelling if the likelihood of the evidence under that hypothesis is vastly greater than under its rivals. In other words, the question is not whether miracles are rare, but whether nineteen named individuals, along with many more unnamed and grouped witnesses, are more likely to testify as they did if Jesus rose from the dead or if He did not.
To make this concrete, consider two hypotheses. Let R represent the claim that the witnesses were broadly correct in reporting encounters with the risen Jesus, and let W represent the claim that the witnesses were wrong in the strong sense required by the alternative explanations, whether through coordinated deception, sincere but repeated mistake, external deception, or psychological phenomena. Bayes compares the probabilities P(R|E) and P(W|E) by combining the prior odds P(R) divided by P(W) with the likelihood ratio P(E|R) divided by P(E|W). The decisive factor is that P(E|R) is naturally high, because if Jesus truly rose, multiple witnesses across time and place is exactly what one would expect. By contrast, P(E|W) becomes extremely small once we ask what would have to be true for nineteen individuals, some devoted, some doubtful, some skeptical, and at least one openly hostile, to converge on the same core claim without an actual resurrection.
To illustrate this numerically, assume values that intentionally favor the possibility that the witnesses were wrong. Suppose that for each named witness there is as much as a ten percent chance that he or she could arrive at a false resurrection claim through some combination of error, suggestibility, deception, or psychological experience. That assumption is deliberately generous.
If we treat these as roughly independent chances, the probability that all nineteen witnesses were wrong equals 1 in 1019, or 0.0000000000000000001. Even if we greatly relax the assumption and allow a twenty percent chance per witness of being wrong, the probability that all nineteen were wrong becomes 0.219, which is approximately one in nineteen trillion.
By contrast, if we assume under the resurrection hypothesis that each witness had a reasonably high chance of being correct, for example eighty percent per witness, allowing for absence, misunderstanding, or incomplete recognition, the probability that all nineteen were broadly correct is 0.819. While that number is modest, the comparison is what matters. Under conservative assumptions that favor skepticism, the evidence is billions to quadrillions of times more expected if the witnesses were broadly correct than if all nineteen were wrong in the way the alternative explanations require.
Consider the Odds: Under assumptions that already favor skepticism, the odds that all nineteen witnesses were wrong are approximately 1,000,000,000,000,000,000 to one against that being the case.
To get a sense of how small these probabilities are, consider a simple illustration. Imagine a stadium filled with trillions of identical coins, stretching far beyond what the eye can see. One coin is secretly marked. You are blindfolded, spun around, and allowed to reach in once. The odds that you pick the marked coin on your first try would still be far better than the odds that nineteen independent witnesses all arrived at the same resurrection claim through error, deception, or delusion under the assumptions already granted to skepticism. Even that comparison understates the point, because the calculation above only concerns the nineteen named witnesses. When unnamed individuals and large group appearances are added, the probability that the entire witness tradition arose through naturalistic failure modes shrinks further still. Numbers at this scale do not merely indicate improbability. They indicate that the explanation has ceased to function as a serious competitor.
“The disciples believed that God had raised Jesus from the dead.”
— Michael Goulder (skeptical scholar), The Evangelists’ Calendar (1978)
Assessing the Naturalistic Alternatives
The idea that all nineteen knowingly lied quickly collapses under scrutiny. Sustained conspiracies become less likely as the number of participants grows, not more likely. Human deception fractures under pressure, fear, and incentive. Yet we find no retractions, no confessions, and no internal exposure, even as the witnesses faced persecution rather than reward. Attempts to reduce the problem by suggesting only a few lied simply shift the explanation to legend or rumor, which introduces new difficulties, especially given the early appeal to named eyewitnesses.
The possibility of sincere mistake fares no better. Misidentification might explain a single ambiguous encounter, but it cannot plausibly account for repeated recognition, extended conversations, physical interaction, and the transformation of skeptics. This explanation becomes especially implausible when applied to James, who rejected Jesus during His ministry, and Paul, who actively opposed the movement. These individuals were psychologically resistant, not primed for belief.
Deception or hoax theories multiply improbabilities. They require a missing body, flawless coordination, repeated successful staging, and silence from enemies who had every incentive to expose the fraud. Bayesian reasoning penalizes such cumulative low-probability assumptions heavily.
Hallucination theories are often presented as the most scientific alternative, but they struggle even more when examined closely. Hallucinations are private mental events, shaped by expectation and emotion. They do not generate shared, repeatable, multi-person encounters involving physical interaction. Even granting generous probabilities for grief-induced experiences among devoted followers, the inclusion of skeptics and a hostile persecutor dramatically lowers the likelihood. When these probabilities are multiplied across nineteen individuals, the odds approach zero.
A common objection is that we do not possess surviving writings from all nineteen named witnesses themselves. But this standard, if applied consistently, would eliminate most of ancient history. The ancient world did not preserve events through personal memoirs from every eyewitness. It preserved them through early proclamation, named testimony, multiple independent sources, and public appeal to living witnesses. The New Testament does not present anonymous claims. It names individuals, situates them in known communities, and openly invites verification, even while hostile authorities were present.
• Paul explicitly appeals to living witnesses
• The resurrection proclamation was public, not private
• Hostile parties never deny the claim that disciples believed they saw Jesus
• Critics instead propose explanations for the experiences, which already concedes the testimony
In other words, even critics concede the belief and testimony. The debate is not whether the witnesses existed or believed, but how to explain what they believed they experienced.
Historians do not ask whether every eyewitness wrote a book; they ask whether the testimony was early, named, public, and open to challenge. By that standard, the resurrection accounts are among the best-attested events of antiquity.
“The resurrection was not a belief that grew up gradually. It was there from the beginning.”
— James D. G. Dunn (scholar), Jesus Remembered (2003)
Beyond the Nineteen: The Witness Pool Expands
The argument does not stop with named individuals. Scripture expands the witness base to include unnamed women accompanying Mary, the unnamed disciple on the road to Emmaus, groups of disciples in Jerusalem, witnesses present at the Ascension, and a group of more than five hundred people, many of whom were still alive when publicly referenced. Each expansion compounds the explanatory burden on alternative hypotheses. Hallucination models do not scale to groups. Conspiracy models weaken as numbers grow. Mistake theories collapse under repeated, multi-modal encounters.
Bayesian reasoning does not merely ask whether these explanations are conceivable. It asks whether they remain probable as the evidence accumulates. They do not.
“The disciples were convinced against all odds that Jesus had risen from the dead.”
— Geza Vermes (Jewish scholar), The Resurrection (2008)
Why the Resurrection Best Explains the Evidence
If Jesus actually rose from the dead and appeared to people, the pattern of evidence we observe is precisely what we would expect: multiple witnesses across time and place, skeptics converted, opponents transformed, early public proclamation, and appeal to living eyewitnesses. Under the resurrection hypothesis, the likelihood of the evidence is high. Under competing hypotheses, it remains stubbornly low unless one continuously adds auxiliary assumptions to keep them viable.
This is not an argument from ignorance. It is an argument from explanatory superiority. Once the alternatives are assessed on their own terms, the question becomes unavoidable.
The Logical Core of the Case
The reasoning can be stated simply. If one hypothesis explains a wide and diverse body of evidence with fewer ad hoc assumptions than its competitors, it is the most rational explanation. The resurrection hypothesis explains the convergence of named individuals, unnamed witnesses, skeptics, opponents, and large groups more naturally than explanations involving deception, mistake, or delusion. The alternatives require multiplying improbabilities to fit the same data. Therefore, the resurrection of Jesus is the best explanation of the available evidence.
The resurrection case does not rest on a single testimony or a single moment. It rests on a widening circle of witnesses whose agreement becomes more difficult to dismiss as the numbers grow. At some point, rejection of the resurrection ceases to be driven by probability and becomes driven by prior philosophical commitments. If the supernatural is ruled out in principle, no amount of evidence will ever be sufficient.
But if one is willing to follow the evidence where it leads, the resurrection does not emerge as a desperate last resort. It emerges as the most coherent explanation of the facts.
That is not blind faith. It is disciplined reasoning applied to history.
“It is historically certain that Peter and the disciples had experiences after Jesus’ death in which Jesus appeared to them as the risen Christ.”
— Gerd Lüdemann (atheist scholar), The Resurrection of Jesus (1994)
(The above figures in this blog are not illustrative guesses. The probabilities and odds were generated by running the stated assumptions through a Bayesian probability model and odds-ratio calculations using standard computational tools.)
Tom's Theology Blog 
Leave a comment