Can You Put a Number on the Divinity of Jesus?

Can we assign a probability to the claim that Jesus was the son of God? The question sounds glib, but it is a useful exercise - not because the number that comes out is meaningful in itself, but because working out where the number comes from exposes exactly how much of your belief is evidence and how much is prior commitment.

Why a Clean Number Is Hard

"Son of God" is a metaphysical claim about the fundamental nature of reality, not an ordinary empirical proposition. The machinery of probability - a sample space, evidence to update on, a reference class - fits it badly. Any estimate is therefore driven mostly by your starting prior rather than by anything specific to Jesus. That is worth saying up front, because a confident-sounding percentage tends to disguise the fact that the work was all done before the evidence arrived.

Decomposing the Claim

The claim is really a chain, and its total probability is roughly the product of the links:

1. Does a God of the relevant kind exist at all?
2. Given that, does this God intervene personally in history?
3. Given that, is the specific Christian account of Jesus the true incarnation - as opposed to mistaken, mythologized, or one of many competing religious claims?

Because the links multiply, the total compounds downward fast unless you begin with strong theistic priors. Even a generous 50% at each link lands you at 12.5%, and link three is where the real damage happens: it is not one hypothesis against its negation, but one hypothesis against thousands of mutually exclusive alternatives.

What History Can and Cannot Settle

There is broad scholarly agreement that a man named Jesus existed, preached, was baptized, and was crucified. But "existed and was crucified" is a completely different claim from "was the son of God." No document or artifact can establish divinity. The gospels were written by committed believers decades after the events - which is exactly what you would expect whether the claim is true or whether it is the theology of a devout community. Evidence that is equally likely under both hypotheses carries zero information. This is the crux: the historical record, however solid, simply does not bear on the question people actually care about.

Where the Number Lands

• Committed Christian: near 1, held as faith-grounded certainty.
• Committed naturalist: near 0, because the framework is rejected at link one.
• Genuine agnostic: low single-digit percent once you multiply through the chain and adjust for the many competing claims on the same slot.

If forced to give a single number from an agnostic starting point, something around 3% is defensible - with the strong caveat that the number is mostly the prior, not the evidence.

The Argument About a Zero Prior

The interesting disagreement is not about the object-level claim. It is about whether a committed naturalist should set the prior to exactly zero, or merely to something very small. The case against a literal zero:

• Zero is unmovable. Bayesian updating multiplies. Anything times zero stays zero. That is not extreme skepticism, it is a claim of impossibility, and it makes the belief immune to any conceivable evidence.
• Cromwell's rule. For any proposition that is not a logical contradiction, a zero prior leaves you unpersuadable even in worlds where the claim is true. Setting your prior to 0 mirrors the young-earth creationist who sets theirs to 1 - the same unfalsifiable move, pointed the other way. "There is a God" is not a contradiction in the way "square circle" is.
• Two bundled priors. Epsilon on Jesus specifically - this man, this tradition, among thousands of contenders - is easy to defend. But epsilon on the whole claim is really a near-zero prior on theism in general, which requires near-certainty on a genuinely open philosophical question.
• Model uncertainty. You should reserve some probability for your entire framework being wrong in ways you cannot currently see. If you are 99.9% confident that naturalism is complete, then all supernatural hypotheses combined cannot sit below 0.1%. That is a floor, and it is not epsilon. Open puzzles - why there is something rather than nothing, fine-tuning, the hard problem of consciousness - prove nothing, but they have not been cleanly closed either.

And the fair counterpoint: model uncertainty need not privilege theism over the infinitely many other frameworks that might replace naturalism. The probability mass leaking specifically to "Christianity is correct" can stay vanishingly small. This is why the bottom line can remain tiny while still not being exactly zero.

The Bottom Line

Two separable questions, with two different answers.

On the object level - is the divinity claim true? - the naturalist is basically right. History does not nudge the divinity of Jesus anywhere meaningful, and the multiplication through the chain drives the result into the ground.

On the prior - should it be exactly zero rather than merely tiny? - no. A literal-zero prior abandons empiricism and adopts the one habit that makes a belief permanently immune to correction. That habit is precisely what we criticize in the believer. It is no better when we adopt it ourselves.

The number should be small. Low single digits or less on the framework, multiplied far down for the specific claim. But not literally zero, and the difference matters more for what it says about how we hold our beliefs than for what it says about Jesus.