The case against, weighed

The short version

1. Every objection a well-read critic would raise is here, in the critic's own words.

   Twelve items are scored; the rest are assigned to the page that scores them. Nothing was left out because it lacked a reply.

   The ledger ↓
2. Each item answers one question: how much less expected is this if the Bible's God exists than if he does not?

   That ratio is a judgement, given as a range with its reasons, so you can disagree with a number and see what follows.

   The rules ↓
3. The heavy items: suffering, the conquest and slavery texts, mind-brain dependence, hiddenness, the imminent-end sayings.

   Each has a serious reply. None of the replies clears the item. The hardest is animal suffering across half a billion years before any human existed.

   The evidence ↓
4. The light items: rival miracles, religious diversity, the track record of science, prayer trials, the cognitive science of religion, contradictions, the size of the universe.

   Each counts against, but only a little, once its reply is weighed.

   The lighter items ↓
5. Together, with overlaps discounted, they cut the odds by about 20 to 1 (range about 9 to 130).

   That does not show the Bible's God is false. It is the weight the positive case has to carry, stated before the verdict, not after.

   The total ↓

The question

We ask whether facts about the world and about the biblical text are less expected if the Bible's account is true than if it is false, and by how much. We were skeptics before we were anything else, and this is the part of the case that weighed on us most. It belongs in the main argument, before the verdict, not in a FAQ.

H, the hypothesis. The God of the Bible (all-powerful, all-knowing, perfectly good, personal and loving) made the universe, spoke through Israel's prophets, and raised Jesus. That is theism plus a claim about revelation, and the counter-evidence hits different parts of it: evil, hiddenness and mind-brain data bear on any theism; the conquest texts and the end-time sayings bear on the Bible in particular; diversity and rival miracles bear on Christianity against other religions.

Which Bible. Two versions of H are kept apart throughout:

• H-inerrant: every statement in the Bible is without error, and every divine command it reports is God's will exactly as stated.
• H-reliable: the Bible is inspired, historically reliable at its core and morally progressive, but uses the genres, idioms and moral starting points of its time. This is the view of much mainstream Christian scholarship, and it is the level at which this site states its hypothesis.

The ledger scores H-reliable, and says where H-inerrant would pay more. We state that choice here, before the scoring, because it is the choice that matters most on the hardest texts.

The alternative. N, naturalism, in the form the philosopher Paul Draper calls the hypothesis of indifference: neither the nature nor the condition of sentient beings on earth results from the good or bad actions of supernatural persons.¹ Where an item separates H from deism or from another religion rather than from N, we say so.

Rules of evidence

Fixed before any item was scored:

1. Same rules both ways. Every rule used elsewhere on the site to score evidence for H applies here to evidence against it. If one veridical near-death experience would count heavily against materialism, one clear failed prophecy counts against an inerrant Bible with the same force.
2. The critic's own strongest form. Each objection is stated from the critics' own writing (Mackie, Rowe, Draper, Schellenberg, Hume, Schweitzer, Allison, Ehrman, Avalos, Morriston), not from an apologist's summary of it.
3. What counts as evidence against. An observation counts against H when H, taken with its own stated content, makes it less expected than N does. A reply counts only if it would be believed without the need to rescue H, or is at least not invented for the purpose. A reply that can absorb any possible observation lowers the objection's force only by also emptying H of content (see the skeptical-theism trade-off).
4. Likelihood ratios, not verdicts. Each item gets a likelihood ratio, LR = P(E | H) / P(E | N): how probable the evidence is if H is true, divided by how probable it is if N is true. Below 1 counts against H. Each LR is a low, central and high value, argued item by item.
5. No double counting. Items that share a premise are grouped into clusters and discounted. Items scored on other pages are summarized here but not scored again.
6. The null space. The full list of objections a well-read critic would raise was drawn up first, from the encyclopedia entries, the critics' books and the skeptics' libraries. Every one is either scored here or assigned to the page that scores it.
7. Exclusions. Objections to Christian practice (church history, scandals, believers' conduct) are not evidence about H's truth, because H predicts sin in the church. Purely conceptual objections (the coherence of omnipotence, the Trinity) bear on the prior, not on the evidence, and are noted but not scored. So do objections to particular doctrines (hell, the atonement, "salvation is too easy"): they concern what H says, not what we observe, and are answered under Objections.

The evidence

First the whole ledger, ranked by how much each item moves the total. Then each item in turn: the critic's case at full strength, the best reply, and what is left over.

The ledger

Clusters: A, "God would act or show himself" (evil, hiddenness, diversity, prayer); B, "the text" (conquest and slavery, end-time sayings, contradictions); C, "naturalism's record" (mind-brain, cognitive science, track record, scale); D, "testimony" (rival miracles).

The heavy items

1. Evil and suffering

2. The conquest and slavery texts

3. Mind-brain dependence

4. Divine hiddenness

5. The imminent-end sayings

The lighter items

6. Rival miracle reports

7. Religious diversity and geography

8. The track record of natural explanation

9. Prayer trials

10. The cognitive science of religion

11. Alleged contradictions

12. The size and age of the universe

Not scored here: Tyre

Ezekiel says of Tyre, "You will be built no more" (Ezekiel 26:14, WEB). Three chapters later the book itself records that Nebuchadnezzar's army "had no wages ... from Tyre" for its siege (Ezekiel 29:18, WEB). Island Tyre fell to Alexander in 332 BC, was rebuilt, and is a city today. Defenders read 26:3's "many nations" as successive waves and "built no more" of the old mainland site. The striking fact for this page is that the book keeps its own partial non-fulfilment, a sign of an unedited text. It is scored on the prophecy page, not here.

The analysis

Items that share a premise are not independent. If you accept that a good God would not allow this much suffering, you will tend to accept that he would not stay this hidden, for the same reason; and one family of replies answers both. So the twelve items are grouped into the four clusters above and combined with a dependence factor ρ (rho):

log L(cluster) = log(strongest LR) + (1 − ρ) × Σ log(other LRs)
ρ = 1: the cluster counts once    ρ = 0: every item multiplies in full
L(total) = L(A) × L(B) × L(C) × L(D)
posterior odds = prior odds × (positive case, areas 01-12) × L(total)

At the central values, the result depends on how much the items overlap:

The Monte Carlo run draws every item from its range, and ρ from 0.3 to 0.8, 200,000 times. Its median (1/39) sits below the point estimate because most ranges stretch further toward the critic's side than the defender's. A median of many draws is not itself a likelihood ratio. Taken as one, the same draws give about 1/33 if the uncertainty is in how well H predicts the data, and about 1/52 if it is in how well N does. The full band of reasonable inputs runs from about 1/9 to about 1/130.

Sensitivity

Which Bible. Score H-inerrant instead (conquest texts central 0.30, contradictions 0.50) and the total falls from about 1/21 to about 1/51 at ρ = 0.5 (1/32 to 1/81 for ρ from 0.7 to 0.3). An inerrant Bible pays about two and a half times as much. We state our H at the reliable level because that is what the site claims ("spoke through Israel's prophets"), and we state it before scoring, not to dodge the hard texts.

Skeptical theism cuts both ways. Cluster A alone (evil, hiddenness, diversity, prayer) is worth about 1/4.3. A reader who accepts skeptical theism strongly enough to set it to 1 removes that factor. But consistency then requires the same humility about predicting God's purposes in the positive case, and the Bible itself teaches that humility: "my thoughts are not your thoughts" (Isaiah 55:8, WEB). The historical evidence for the resurrection as an event survives this, but its support for God does not: inferring God from the event needs a judgement about what God would do, which is exactly what skeptical theism withholds. Applied evenly, full skeptical theism leaves the question of a God who acts at about even odds: it cancels the resurrection's support for such a God together with the cost of evil. Likewise arguments of the form "a God would want X, and X is what we see" (parts of fine-tuning and design) must then shrink. (Prophecy is assessed on its own pages.) We show both switches together, never the discount on one side only.

Double counting. The near-death page and mind-brain dependence must be shown together. The resurrection and rival miracles overlap; so do the origin of life and the track record of science. Tyre belongs to the prophecy page.

What would change the numbers. A clear theodicy for pre-human animal suffering would raise item 1 most. A demonstration that Mark 13:30 cannot refer to AD 70 would lower item 5. A pre-registered, replicated near-death target hit would flip item 3.

Rival traditions

This page uses other religions' miracle reports only as a check on testimony: if early, sincere eyewitness claims are common, testimony alone proves less. The positive evidence offered for other traditions (Judaism, Islam, and others) is weighed on its own terms, against the same rules, on the rival traditions page.

Objections

First, objections to this page's method. Then the questions readers ask most about the replies themselves.

What this shows, and what it doesn't

What we don't claim

• "Plantinga solved the problem of evil." He answered the logical problem, and Mackie conceded that much. The evidential problem is open.
• "Without God there is no evil, so evil proves God." That is the moral argument, a separate argument. It does not make the suffering we see any more expected under H.
• "The Bible never commands violence" or "the conquest was not real killing." The texts say what they say (Deuteronomy 20:16-17; 1 Samuel 15:3). Hyperbole is a reading, not a fact.
• "The Bible condemns slavery." It regulates it; Leviticus 25:44-46 permits permanent ownership of foreigners. The New Testament's seeds of abolition (Philemon; Galatians 3:28) are real, and that is all we claim.
• "The Bible has no contradictions." Overclaim. At most: alleged contradictions have proposed harmonizations of varying strength.
• "Jesus never predicted an imminent end" as settled fact. Many critical scholars think he did; the AD 70 reading is strong and defensible, not the only one.
• "Prayer has been scientifically shown to heal." The largest trial (STEP) and the Cochrane review found no effect. Small positive trials cited alone mislead.
• "No other religion has eyewitness miracle claims." False: Vespasian, the Mormon witnesses, Sai Baba, Fátima.
• "Near-death experiences show the mind works without the brain" without the mind-brain data beside it and without AWARE II's null result on hidden targets.
• "Science is increasingly finding God." The record of explanation runs mostly the other way. We claim only specific findings (the beginning, fine-tuning), and their interpretation is disputed.
• "Death before the Fall is only a young-earth problem." Any view that ties all death and suffering to human sin has it; old-earth Christians owe a separate account of animal suffering.
• "God cannot annihilate a soul." Whatever God keeps in being freely, he can cease to keep; the argument also proves too much. Whether he would is a dispute among Christians, not a proof either way.
• "Animals do not really suffer" as a fact, or as the answer to animal suffering. It is a minority view, and we do not rely on it.
• "Answered prayers in my life show that God acts." Not without base rates: goods that most people have, believers or not, cannot show it. Nor does the Bible promise health and wealth to the faithful (2 Corinthians 12:9).
• "Skeptical theism disposes of evil, and the resurrection still proves God." Applied evenly, it removes both together; only the evidence for the event itself survives.
• "God tried showing himself in the wilderness, and it didn't work." As an experiment God ran, it sits badly with his foreknowledge, and the text has God intend the fear (Exodus 20:20). It illustrates that sight does not produce love; it proves nothing more.
• "It has been shown that Christ came at the only possible time," or that the Christian way of salvation is the only possible one. These are arguments about fittingness, and they are contested.
• Quotations we could not check. Rowe's argument is paraphrased here, not quoted, because we have not checked the journal text.

Sources

1. Draper, P. (1989). "Pain and Pleasure: An Evidential Problem for Theists." Noûs 23(3): 331-350. Quoted via J. J. Lowder, "Draper on Pain and Pleasure: Part One," Secular Frontier, 2014, infidels.org.
2. Mackie, J. L. (1955). "Evil and Omnipotence." Mind 64(254): 200-212. PDF.
3. Rowe, W. L. (1979). "The Problem of Evil and Some Varieties of Atheism." American Philosophical Quarterly 16(4): 335-341. (Paraphrased; the journal text was not checked.)
4. Tooley, M. "The Problem of Evil." Stanford Encyclopedia of Philosophy, §§3, 6.3. plato.stanford.edu/entries/evil.
5. Plantinga, A. (1974). God, Freedom, and Evil. New York: Harper & Row.
6. Mackie, J. L. (1982). The Miracle of Theism. Oxford: Clarendon, p. 154. As quoted in "God and purported logical arguments from evil and suffering," New Blackfriars (2024), Cambridge Core.
7. Wykstra, S. (1984). "The Humean Obstacle to Evidential Arguments from Suffering: On Avoiding the Evils of 'Appearance'." International Journal for Philosophy of Religion 16: 73-93.
8. Bergmann, M. (2001). "Skeptical Theism and Rowe's New Evidential Argument from Evil." Noûs 35(2): 278-296.
9. Hick, J. (1966). Evil and the God of Love. London: Macmillan.
10. Adams, M. M. (1999). Horrendous Evils and the Goodness of God. Ithaca: Cornell University Press.
11. Stump, E. (2010). Wandering in Darkness: Narrative and the Problem of Suffering. Oxford: OUP.
12. World English Bible (public domain), via bible-api.com. All Bible quotations on this page.
13. Southgate, C. (2008). The Groaning of Creation: God, Evolution, and the Problem of Evil. Louisville: Westminster John Knox.
14. Murray, M. (2008). Nature Red in Tooth and Claw: Theism and the Problem of Animal Suffering. Oxford: OUP.
15. Planck Collaboration (2020). "Planck 2018 results. VI. Cosmological parameters." Astronomy & Astrophysics 641: A6 (age 13.787 ± 0.020 Gyr).
16. International Commission on Stratigraphy. International Chronostratigraphic Chart v2023/09 (base of the Cambrian 538.8 Ma).
17. Hublin, J.-J. et al. (2017). "New fossils from Jebel Irhoud, Morocco and the pan-African origin of Homo sapiens." Nature 546: 289-292.
18. Dembski, W. (2009). The End of Christianity: Finding a Good God in an Evil World. Nashville: B&H.
19. Roser, M. "Mortality in the past: every second child died." Our World in Data. ourworldindata.org.
20. Morriston, W. (2009). "Did God Command Genocide? A Challenge for the Biblical Inerrantist." Philosophia Christi 11(1): 7-26.
21. Avalos, H. (2011). Slavery, Abolitionism, and the Ethics of Biblical Scholarship. Sheffield Phoenix.
22. Wolterstorff, N. (2011). "Reading Joshua." In M. Bergmann, M. Murray and M. Rea (eds.), Divine Evil? The Moral Character of the God of Abraham. Oxford: OUP.
23. Copan, P. and Flannagan, M. (2014). Did God Really Command Genocide? Grand Rapids: Baker.
24. Parnia, S. et al. (2023). "AWARE II." Resuscitation. Summary: PSI Encyclopedia, "AWARE NDE Studies," psi-encyclopedia.spr.ac.uk. Counts are checked on the near-death page.
25. Schellenberg, J. L. (1993). Divine Hiddenness and Human Reason. Ithaca: Cornell UP; (2015) The Hiddenness Argument. Oxford: OUP. Quoted, with the replies surveyed, in Howard-Snyder, D. and Green, A. "Hiddenness of God," Stanford Encyclopedia of Philosophy, plato.stanford.edu/entries/divine-hiddenness.
26. Moser, P. (2008). The Elusive God: Reorienting Religious Epistemology. Cambridge: CUP.
27. Bourget, D. and Chalmers, D. J. (2023). "Philosophers on Philosophy: The 2020 PhilPapers Survey." Philosophers' Imprint 23. Results for "God: theism or atheism?": accept or lean toward atheism 66.95%, theism 18.93%, other 14.63% (N = 1,653). survey2020.philpeople.org.
28. Schweitzer, A. (1910). The Quest of the Historical Jesus, tr. W. Montgomery, ch. 19. Project Gutenberg #45422.
29. Allison, D. C. (1985). The End of the Ages Has Come. Philadelphia: Fortress; (1998) Jesus of Nazareth: Millenarian Prophet. Minneapolis: Fortress.
30. Ehrman, B. D. (1999). Jesus: Apocalyptic Prophet of the New Millennium. Oxford: OUP.
31. Wright, N. T. (1996). Jesus and the Victory of God. London: SPCK; Minneapolis: Fortress.
32. Wenham, D. (1988). "The End Is Near. In What Sense?" Themelios 13(3). thegospelcoalition.org.
33. Hume, D. (1748). An Enquiry Concerning Human Understanding, sect. X, §§95-96. Project Gutenberg #9662.
34. Tacitus, Histories 4.81, tr. C. H. Moore (Loeb, 1931). LacusCurtius. Also Suetonius, Vespasian 7.
35. Harper, S. C. (2010). "Evaluating the Book of Mormon Witnesses." Religious Educator 11(2). rsc.byu.edu.
36. Haraldsson, E. (1987; rev. 2013). Modern Miracles: Sathya Sai Baba. White Crow. See also PSI Encyclopedia, "Sai Baba," psi-encyclopedia.spr.ac.uk.
37. Catholic World Report / CNA (16 April 2025). "Lourdes announces 72nd miracle." catholicworldreport.com.
38. Philostratus, Life of Apollonius of Tyana (c. AD 220).
39. Vogel, D. (2002). "The Validity of the Witnesses' Testimonies." In D. Vogel and B. Metcalfe (eds.), American Apocrypha. Salt Lake City: Signature Books.
40. Pew Research Center (9 June 2025). "How the Global Religious Landscape Changed From 2010 to 2020." pewresearch.org.
41. Craig, W. L. (1989). "'No Other Name': A Middle Knowledge Perspective on the Exclusivity of Salvation through Christ." Faith and Philosophy 6(2): 172-188.
42. Draper, P. (2005). "God, Science, and Naturalism." In W. Wainwright (ed.), The Oxford Handbook of Philosophy of Religion. Oxford: OUP.
43. Benson, H. et al. (2006). "Study of the Therapeutic Effects of Intercessory Prayer (STEP) in cardiac bypass patients." American Heart Journal 151(4): 934-942. PubMed 16569567.
44. Roberts, L., Ahmed, I. and Davison, A. (2009). "Intercessory prayer for the alleviation of ill health." Cochrane Database of Systematic Reviews CD000368. cochrane.org.
45. Li, S., Stampfer, M., Williams, D. and VanderWeele, T. (2016). "Association of Religious Service Attendance With Mortality Among Women." JAMA Internal Medicine 176(6): 777-785.
46. Barrett, J. L. (2000). "Exploring the Natural Foundations of Religion." Trends in Cognitive Sciences 4(1): 29-34.
47. Boyer, P. (2001). Religion Explained. New York: Basic Books.
48. Atran, S. (2002). In Gods We Trust. Oxford: OUP.
49. Kelemen, D. (2004). "Are Children 'Intuitive Theists'?" Psychological Science 15(5): 295-301.
50. Wilkins, J. and Griffiths, P. (2013). "Evolutionary Debunking Arguments in Three Domains." In G. Dawes and J. Maclaurin (eds.), A New Science of Religion. Routledge.
51. Launonen, L. (2021). "Debunking Arguments Gain Little from Cognitive Science of Religion." Zygon 56. zygonjournal.org.
52. Jong, J. and Visala, A. (2014). "Evolutionary Debunking Arguments against Theism, Reconsidered." International Journal for Philosophy of Religion 76: 243-258.
53. Murray, M. (2009). "Scientific Explanations of Religion and the Justification of Religious Belief." In J. Schloss and M. Murray (eds.), The Believing Primate. Oxford: OUP.