Comments on: Pascal’s Wager and the Ad Baculum Fallacy https://www.richardcarrier.info/archives/13810 Announcing appearances, publications, and analysis of questions historical, philosophical, and political by author, philosopher, and historian Richard Carrier. Wed, 04 Nov 2026 18:08:11 +0000 hourly 1 https://wordpress.org/?v=7.1.4 By: Richard Carrier https://www.richardcarrier.info/archives/13810#comment-45873 Wed, 04 Nov 2026 18:08:11 +0000 https://www.richardcarrier.info/?p=13810#comment-45873 In reply to Istvan.

you omitted a minus sign

Good catch. Fixed.

Nick Bostrom published in 2009, called Pascal’s mugging

Lovely. Thank you for reporting that. There is even a Wikipedia page on that now.

I haven’t read it all through but it does look like he’s arguing the same point I am. I’m glad to see someone else noticed what I did.

She genuinely believes…But I shouldn’t believe her

That’s my point, too.

Although, just a side note, I do think this is more often lying—no one actually knows anything about hell or who goes there, and preachers and apologists do seem to make stuff up about that, whatever conveniently suits the outcome they want. It’s like any activists who emotionally over-exaggerate a problem to try and “scare” people into caring because the truth didn’t persuade them: quite a lot of the time this is a tactic, not an honest mistake. It’s self-defeating. But people aren’t rational enough to know that lying is generally a far worse idea than they’ve convinced themselves it is.

do you think it is rational for alcoholics and drug addicts to turn to Christianity in the hope of overcoming their addiction

No. Because it is toxic. It is functionally trading one poison for another. And that is indeed the scam: to rope suffering people in with snake oil and hope the placebo effect will get them to approve all the other evil policies and controls they stack on top.

A rational move (though remember, most people, especially addicts, are not rational, so this isn’t a prediction but a conditional) would be to turn to science and evidence-based treatment, which doesn’t stack on the victim’s back evil commitments and false beliefs as preconditions for receiving care. Secular Recovery or REBT are rational, for example. Christianity is not. It’s no different than getting clean by going to Jonestown or joining Heavens’ Gate. The only difference is the immediacy or directness of the bad outcome. Christianity is the boiled frog or weaponized frog to Jonestown’s crushed frog. But either way is a poisonous outcome. And the slow roll does far more widespread harm (think sexism, homophobia, child abuse, religious trauma, voting against the interests of the poor).

For more on that see What’s the Harm? vs. Dear Christian: You Might Be Worshiping the Antichrist and The Real War on Christmas: The Fact That Christmas Is Better Than Christ.

It’s ultimately always the same question: does Reiki cure disease or evidence-based medicine? Replace “Reiki” with any religion and the answer remains the same. I close with this point at the end of Let’s Be Honest about Why You Are a Christian or a Muslim.

]]>
By: Istvan https://www.richardcarrier.info/archives/13810#comment-45851 Wed, 04 Nov 2026 08:09:05 +0000 https://www.richardcarrier.info/?p=13810#comment-45851 Dr. Carrier,

If I may, I have a few things to say and ask about this post.

First, I just want to point out that you omitted a minus sign: “The expected utility of buying a raffle ticket is minus one dollar ($0.99 – $1.99 = $1).”

Second, the thought experiment you describe in the paragraph beginning with “We can call this the Liar’s Wager” is similar to the one Nick Bostrom published in 2009, called Pascal’s mugging. Before reading your post, I knew something was wrong with Pascal’s reasoning (I mean the character Pascal from Bostrom’s thought experiment), but I didn’t know what. The mugger tricks Pascal by saying he can give him any finite amount of money and happy days, and by asking Pascal to specify just how unlikely he thinks it is that he (the mugger) will deliver on his promise. Once Pascal specifies the probability, the mugger makes an amazing promise, and Pascal hands him his wallet. I think Pascal should have remained skeptical of the mugger’s promise no matter how amazing it was. He didn’t believe the mugger would deliver on a lesser promise, so he shouldn’t have believed he would deliver on an amazing one (at least not without a lot of evidence).

If you’ll ever read Bostrom’s thought experiment, I would like to know what you think about it. https://nickbostrom.com/papers/pascal.pdf

Third, you wrote, “Maybe Pascal is just a fool, and thus simply in error, and not actually a liar.” I think most Christians genuinely believe that God sends people to Heaven as long as they believe in Him. So, alongside thought experiments such as the Liar’s wager and Pascal’s mugging, we could also use thought experiments that don’t involve ill-intentioned people. Let’s say I have a relative who genuinely cares about me, but who is also a member of a cult. She genuinely believes the cult’s doctrines and tells me that, based on these doctrines, the souls of those who don’t join the cult by dedicating their lives to it will be annihilated, while the souls of those who do will go to Heaven. She asks me to believe her and join the cult, but I have already looked into the cult’s doctrines, and I find them very unlikely. There is a nonzero chance my relative is right, and any finite probability multiplied by an infinite gain remains an infinite gain. But I shouldn’t believe her because my skepticism about the cult’s doctrines means I wouldn’t have believed her if she had made a less amazing claim.

Fourth, do you think it is rational for alcoholics and drug addicts to turn to Christianity in the hope of overcoming their addiction?

]]>
By: Richard Carrier https://www.richardcarrier.info/archives/13810#comment-44630 Mon, 13 Jul 2026 19:12:44 +0000 https://www.richardcarrier.info/?p=13810#comment-44630 In reply to Ulka.

That doesn’t matter.

That’s what my article explains.

It literally cannot matter whether I “just make something up,” because anything logically possible has a nonzero probability of being true, and any nonzero probability multiplied by infinity is infinity—and not just infinity, but exactly the same infinity.

Do you understand the problem now?

If we rely on the wager, its own faulty logic entails even the most improbable religion logically conceivable is just as good as any sect of Christianity you want to settle on. Because the wager converts all probabilities into infinite net goods. All. Not “the higher probabilities.” All of them. They all flatten out as the same.

This is what I am pointing out is broken mathematically with the wager. It can’t do the thing you want to do. Its own mathematics rules your point out of court. Which should make us suspicious of its mathematics.

Which is what the rest of my article exposed and debunks: the wager’s math is wrong. You can’t make a religion attractive by increasing the absurdity of its promises. That’s as true of Christianity as Islam or Buddhism or Taoism or Cthulhuism or Pastafarianism or [insert made-up religion here].

So all we have left is the evidential status of the religion compared to alternatives. And on that, Christianity always loses. But that’s not because of the wager. It’s because of precedent, parsimony, and evidence. And turning Christianity into a cosmic conspiracy theory doesn’t save it.

]]>
By: Ulka https://www.richardcarrier.info/archives/13810#comment-44626 Mon, 13 Jul 2026 18:09:14 +0000 https://www.richardcarrier.info/?p=13810#comment-44626 In reply to Richard Carrier.

Your wager’s God doesn’t come from any religion or holy book, it’s just made up to make the argument work. But Christianity or Islam have thousands of years of philosophy, scholars, and followers behind their God. Doesn’t that make the Christian or Islamic God more likely to be real than your wager’s God, which feels kind of made-up?

]]>
By: Richard Carrier https://www.richardcarrier.info/archives/13810#comment-44617 Mon, 13 Jul 2026 14:56:27 +0000 https://www.richardcarrier.info/?p=13810#comment-44617 In reply to Ulka.

That’s not two hypotheses it’s “infinitely many.” And it includes organically constructed religions not reverse engineered from Christianity at all but often even predating it. But even the engineered examples have the same epistemic status because Christianity (and all its tens of thousands of mutually refuting sects) is a made-up false religion, so “making up” more false religions has no effect on the relative likelihoods.

Which may signal that you don’t understand Pascal’s Wager. The point of the wager is to address the implausibility of Christianity: whereby yes, even if the evidence signals it is false, the wager is the last resort to ensuring you believe it anyway. The flaw in that is that it assumes it is the only available hypothesis of long-term outcomes. Once you reintroduce all infinitely many alternative theories, the wager washes out as warranting no belief at all. So you have to go back to the original basis for believing things: the thing with the most real evidence. A contest Christianity always loses.

Now, back to the point of this article: even granting the wager’s invalid logic, we get the result…

These two Wagers have the exact same utility value (even if it’s infinite)—and therefore negate each other. Therefore all such Wagers are invalid. That wouldn’t necessarily be the case, but it is the case for Pascal’s Wager because of that invalid premise of an irrational utility function we just examined. If we grant that the utility function for a God is infinite, then it is infinite for every contrary God; and the mathematical result is zero net utility

Notice that the wager’s own logic entails that differences in likelihood for alternative hypotheses can never matter because they multiply that by infinite utility functions, and any finite number multiplied by infinity equals the same value: infinity (in particular aleph-1, the set of all real numbers). So even if I pair Christianity against a made-up religion that is ten billion times less likely to be true than Christianity, they both end up infinitely preferable, and so it is impossible to choose based on the wager’s logic. In short, the wager’s own logic eliminates any function that differences in epistemic probability would have had, because it makes all theories infinitely rewarding no matter how unlikely they are.

Hence the rest of my article addresses what is going wrong here: you can’t just multiply a probability by infinity. The more outrageous the promise, the lower the probability. So the epistemic value of eternal reward is not infinity, but infinitesimal. For reasons the article explains. But when we use the correct math, the wager gets no usable result. Hence it is fallacious.

]]>
By: Ulka https://www.richardcarrier.info/archives/13810#comment-44605 Sat, 11 Jul 2026 20:15:51 +0000 https://www.richardcarrier.info/?p=13810#comment-44605 In reply to Richard Carrier.

Since there are infinitely many possible diverse outcomes (infinitely many possible gods, systems of gods, and systems lacking gods), they all negate each other. For example, for any “believe in God(x) or else” Wager, there is a contrary Wager that is equally valid, such as “God(y) will punish you for believing in God(x), therefore you should believe in God(y).” These two Wagers have the exact same utility value (even if it’s infinite)—and therefore negate each other. Therefore all such Wagers are invalid.

My point is: If you construct an ‘anti-Pascal God’ specifically to counter the Christian God, that anti-Pascal God is reverse engineered.

]]>
By: Richard Carrier https://www.richardcarrier.info/archives/13810#comment-44593 Fri, 10 Jul 2026 22:37:41 +0000 https://www.richardcarrier.info/?p=13810#comment-44593 In reply to Ulka.

I am not sure what you mean by “both hypotheses.” Can you quote the exact text of the article you are asking about?

]]>
By: Ulka https://www.richardcarrier.info/archives/13810#comment-44568 Sun, 05 Jul 2026 14:40:31 +0000 https://www.richardcarrier.info/?p=13810#comment-44568 I understand your point, but I want to ask something. You keep saying both hypotheses have the exact same probability — but I’m not sure that’s fully correct. By Occam’s razor, shouldn’t a purpose-built counter-hypothesis (like the Anti-Pascal God) get some penalty for complexity, since it wasn’t built from evidence but reverse-engineered specifically to counter the other claim? A hypothesis built backward from a wanted conclusion feels different from one that comes from actual observation, like in science.
Also, in Bayesian reasoning, likelihood can be different from person to person, depending on their background knowledge and assumptions. So how can you say both hypotheses have exactly the same probability for everyone, when Bayesian reasoning itself allows for different priors and likelihoods between people?

]]>
By: albattani https://www.richardcarrier.info/archives/13810#comment-40274 Tue, 25 Mar 2025 08:44:38 +0000 https://www.richardcarrier.info/?p=13810#comment-40274 Excellent Blog, as usual! But still, the best answer to Pascal’s wager is the poem by Jacques Prévert…

]]>
By: Richard Carrier https://www.richardcarrier.info/archives/13810#comment-39973 Sat, 18 Jan 2025 19:48:10 +0000 https://www.richardcarrier.info/?p=13810#comment-39973 In reply to secretlytastemaker867ef0866b.

Refuted in the article you are commenting on.

Please stop commenting without addressing the article. That will get you permanently banned if you do it again.

]]>