Comparing Theism and Naturalism

Bayes machine goes BRR

Graham Oppy, famous comparer of the theories.

I’ve elsewhere written about the evidence both for theism and atheism. Here, I’ll provide some extremely back-of-the-envelope calculations for the odds of theism vs atheism. I’m going to use the following to represent what’s going on: ~ denotes the odds ratio on theism vs naturalism. So, for instance, 5~1 means some piece of evidence is five times as likely on theism as on naturalism. The article I’ve linked describes my reasoning, so here I won’t explain why I regard these as as strong as I do, it will just give the numbers:

• Prior 1~100

• Physical stuff exists 1~1

• It does stuff 5~1

• Laws apply to constants 4~1

• The stuff it does is interesting and valuable 1,000~1

• Constants are finely tuned 2~1

• Life originated and went through the many stages needed to get us 2~1

• There are psychophysical laws 100~1

• Souls remain consistent over time 2~1

• There’s psychophysical harmony 50~1

• The stuff that the psychophysical laws make conscious exists (the psychophysical laws make brains conscious and then there happen to be brains) 3~1

• People have generally true beliefs about morality, metaphysics, math, modality, and more 13~1

• Anthropic stuff 100:1

• Lots of people have religious belief, powerful religious experiences, and feel a deep connection with God. 5~1

• Miracles 5~1

• The world has tons of beauty, happiness, and love 1.5~1

• There are lots of NDEs 1.2~1

• Evil 1~100,000

• Hiddenness 1~5

• The absence of any super clear evidence for a miracle 1~5

• The universe is big and old—conscious beings with valuable experiences are a flash in the pan. 1~3

• The fundamental physical laws are simple 1~2

(The other stuff got funneled into the prior, and I’ve since become convinced that most of it isn’t hugely significant).

Now, as to which view is more probable: the conclusion is that theism is 468,000 times more plausible than naturalism. And that’s all while taking evil very seriously and being conservative about much of the evidence. Don’t take these numbers too seriously, but this is, I think significant.

Here I’ve just compared theism to naturalism because otherwise it gets trickier, for then the probabilities start getting funneled to the small shared of modal space that solve most of the problems. Obviously my credence in theism is not actually 468,000 times that of naturalism—the model could be wrong!

Though I don’t think it’s crazy to think that the correct calculation will turn up a ridiculously high probability of theism. First, there’s a lot of evidence for theism. If some view has 17 decent pieces of evidence in its favor, then even if none of those pieces of evidence are crushing, collectively they provide ridiculously strong evidence. In contrast, atheism only has one good argument for it: the problem of evil.

Furthermore, it might be that theism has a really low prior on the best way to calculate prior probability. But I don’t really know how to calculate prior probability, and the contingency argument or something might work. So, therefore, one’s prior in theism can’t be too low, for there are plausible arguments for why it’s a necessary truth. Even if you’re 90% sure that theism has a prior of near zero, if you think there’s some chance the contingency argument or some ontological argument works, that’s enough to get theism to have a non-terrible prior.

Definitely don’t take this too seriously. The odds that I’m massively confused and the evidence is much weaker than I think is much greater than 1/468,001.