Why Malerin is Wrong About Bayes Theorem

[Malerin]

I beg to differ Malerin. There is, according to your own definition, evidence for the FSM. In fact, I had a subjective spiritual experience with the FSM just yesterday. Thus - again, by your own definition & arguments - you must accept the agnostic probability of the FSM's existence to be 0.5.

Prove me wrong. Go on, I dare you.

In the meantime, may you be touched by His Noodly Appendage. RA-men!

No offense, but you're not very believable. Anyway, I have had spiritual experiences that I believe to relate to God, as have other people I trust. You're the only person I've ever heard admit to having a spiritual experience about the FSM, and you're obviously lying to make a point.

So, evidence for God? Yes. Evidence for FSM? No. FSM is unlikely given my background knowledge of biology (taking off my skeptical hat for the moment)? Yes.

 

[Ivor the Engineer]

<snip>

Yes, but if that were the case, I think Bayes' theorem itself would fall down.

It would be an interesting challenge to attempt to draw the tree diagram.

 

[Ichneumonwasp]

Malerin, could you remind us again why you chose 0.5 for the likelihood of God's existence?

I don't recall any longer.

 

[Frozenwolf150]

Wow, Malerin, a million dollar challenge! If you truly believe what you've been saying in other threads, and you are willing to consistently apply your arguments, then why not take RandFan's challenge?

C'mon Malerin, put your money where your mouth is.

Throwing one's money onto the piles of refuse found on the floor of a bull pen, where the mouths of creationists may be found, is not a sound investment option. For example:

There is no evidence for the FSM, and it would seem to violate basic biological truths and evolutionary truths. As CJ has pointed out, there is evidence for God (subjective spiritual experiences, NDE accounts), though not the kind of evidence everyone finds convincing, and the existence of God does not violate any basic scientific truths. Therefore, one can be agnostic about God (maybe the experiences are genuine, maybe the NDE accounts are accurate) while rejecting the FSM (no evidence for it, how would it be possible biologically).

When you need reliable firsthand accounts to support your woo, look no further than brain damage patients. Um, yeah.

No offense, but you're not very believable. Anyway, I have had spiritual experiences that I believe to relate to God, as have other people I trust. You're the only person I've ever heard admit to having a spiritual experience about the FSM, and you're obviously lying to make a point.

So, evidence for God? Yes. Evidence for FSM? No. FSM is unlikely given my background knowledge of biology (taking off my skeptical hat for the moment)? Yes.

Even evidence that undermines your stance is better than no evidence at all! Am I right?

Never mind that this is only evidence for people's belief in God, and not evidence for God's actual existence.

 

[rocketdodger]

And we know from upthread that they do not; we've got the examples both of a possibly unfair coin and of a child of indeterminate sex who may or may not be named Sue.

Since we know that it's possible that they don't add up to the necessary number, it makes no sense whatsoever to assert that, in all cases where we don't have information enough to confirm that they don't, they must.

But the crux of my argument is that in both of these other cases you have much more information than you do in Malerin's example.

With the coin, you know at least that a fair coin lands on heads/tails 50% of the time.
With the child, you know at least all the other names of children you remember encountering or at the very least are familiar with human naming conventions.

If you didn't know that already, you would not know it is possible to have a child named anything but Sue or for a coin land on anything but heads.

That is why I feel my number example is the best analogy -- it represents exactly what Malerin is getting at in his example. Suppose

H = a number chooser that is likely to choose a number satisfying constraints C exists
E = a number satisfying constraints C was choosen somehow

Furthermore, suppose that out of a single trial -- and only a single trial -- we observe a number satisfying those constraints.

What can you do besides say P(E|H) + P(E|~H) == 1? All we have is a single trial and we know nothing else that could lead to a better estimate. Is there something we can do that is more rational?

I really want to understand this.

 

[drkitten]

What can you do besides say P(E|H) + P(E|~H) == 1?

Not say anything. You're essentially making up a number here -- as you point out, you're going from three free variables to two, which is equivalent to silently assigning a value to a variable, a value pulled out of thin air.

Which is where Malarin's problem comes in. It's not with the explicitly agnostic value assigned to P(H). It's with the implicit and falsely confident gnosis in assigning meaningless but precise values to everything else.

 

[Malerin]

Not say anything. You're essentially making up a number here -- as you point out, you're going from three free variables to two, which is equivalent to silently assigning a value to a variable, a value pulled out of thin air.

Which is where Malarin's problem comes in. It's not with the explicitly agnostic value assigned to P(H). It's with the implicit and falsely confident gnosis in assigning meaningless but precise values to everything else.

Why do you think Pr(E/H) is imprecise? Given a universe creator (fine-tuner) sympathetic towards life exists, we should not be surprised at all that life exists.
Pr(E/H) >> .5

 

[GreedyAlgorithm]

What can you do besides say P(E|H) + P(E|~H) == 1? All we have is a single trial and we know nothing else that could lead to a better estimate. Is there something we can do that is more rational?

I really want to understand this.

My reasoning:

1) rocketdodger is making a mistake in saying we can do nothing besides "say P(E|H) + P(E|~H) == 1"
2) rocketdodger has some fundamental confusion about probability maths which is causing this mistake
3) from the perspective of the nonconfused, his choice of the constant 1 is arbitrary, so his confusion must be causing his choice of the constant 1
4) his constraint is supposed to arise because we know that P(E) == 1, which perhaps not so coincidentally is our magic constant?

So what's the smallest confusion that could cause this? I'm going with 69dodge's earlier guess, that somehow rocketdodger is confusing P(E|H) + P(E|~H) with P(E & H) + P(E & ~H). This second quantity has the property that P(E) = P(E & H) + P(E & ~H), so upon setting P(E) to 1, we find that instead of P(E & H) + P(E & ~H) being some unknown number, it has suddenly been set to 1 as well. Furthermore if you substitute "P(E & H) + P(E & ~H)" into all of his posts in for "P(E|H) + P(E|~H)", the ludicrous parts are suddenly tautological and uninteresting, but correct.

 

[Malerin]

Even evidence that undermines your stance is better than no evidence at all! Am I right?

In the same way that .0000009 is greater than .0000008, sure.

 

[Dunstan]

What can you do besides say P(E|H) + P(E|~H) == 1? All we have is a single trial and we know nothing else that could lead to a better estimate. Is there something we can do that is more rational?

I really want to understand this.

As has been pointed out repeatedly, your equation simply isn't justified by the premises. Instead of asking everyone "what's wrong with saying this," you need to explain why you think that equation is right. Because it's just not a statement that makes any sense.

Maybe try putting into words what it is you think that the equation P(E|H) + P(E|~H) = 1 means?

 

[drkitten]

My reasoning:
4) his constraint is supposed to arise because we know that P(E) == 1, which perhaps not so coincidentally is our magic constant?

So what's the smallest confusion that could cause this?

I think the smallest confusion is the idea that P(E) must equal 1, simply because we've observed the data E.

That's equivalent to saying that anything that did happen must have happened; there are no contingent events or free will. If I named my cat Lucretius, then there were no other possible names I could have given my cat.

 

[GreedyAlgorithm]

Why do you think Pr(E/H) is imprecise? Given a universe creator (fine-tuner) sympathetic towards life exists, we should not be surprised at all that life exists.
Pr(E/H) >> .5

There are many reasons. Here's a bound: Given that

1) you are using the datum "Given a universe creator (fine-turner) sympathetic towards life exists, we should not be surprised at all that life exists" to set P(E|H)>>0.5
2) since the fine-tuner is already life, then if we should not be surprised that more life exists given that he exists, then we can conclude that he is likely to value quantity of life at least somewhat
3) a different universe construction than is likely given the evidence we have collected on earth would result in significantly more life than we currently have evidence for
4) we have no good reason to suppose that the amount of life we currently have evidence for is for some reason optimal for a sympathetic fine-tuner

we can conclude that 5) a completely agnostic prior P(H)=0.5 is leaving out at least some known data. The ability to be specific about P(E|H) and not have any ability to be more specific about P(H) than 0.5 is very unlikely, so that either you are throwing away data about P(H) or being overly specific about P(E|H). There's a reason; there are more.

 

[drkitten]

Why do you think Pr(E/H) is imprecise?

I don't. I think our knowledge of it is imprecise.

Given a universe creator (fine-tuner) sympathetic towards life exists, we should not be surprised at all that life exists.
Pr(E/H) >> .5

Not at all. I'm fond of ice cream, but the probability that I'm eating it at any given instant is rather low. The universe creator may be fond of life, but that doesn't mean that everything He does produces life.

The fact is that you have no actual data to support any particular estimate of P(E|H), and all you have is a rather vague hand-wave, but you want to turn that into an actual number.

 

[Frozenwolf150]

In the same way that .0000009 is greater than .0000008, sure.

Actually, I meant that the evidence you claim for God is evidence that ends up weakening your stance. People have "experiences" but they're never the same between individuals, and these experiences have applied to many different competing gods in the past. Furthermore, your reliance on NDE accounts as evidence for your woo is nearly as exploitative of victims of tragedy as the 9/11 truth movement is. Congratulations, you've shown that every new refuge apologists try to hide under winds up being a new low.

 

[MattusMaximus]

No offense, but you're not very believable. Anyway, I have had spiritual experiences that I believe to relate to God, as have other people I trust. You're the only person I've ever heard admit to having a spiritual experience about the FSM, and you're obviously lying to make a point.

So, evidence for God? Yes. Evidence for FSM? No. FSM is unlikely given my background knowledge of biology (taking off my skeptical hat for the moment)? Yes.

My subjective spiritual experience of the FSM is just as valid as your subjective spiritual experience of God. Hence the term "subjective". Who are you to say that my experience isn't true? Heretic!

Besides, we all know your "evidence" for God is just a fantasy. My evidence for the FSM is real - because I say so given my background knowledge of all things pasta.

Malerin, I implore you to see the light. Cease worshiping this false "God" of yours, and embrace the Truth that is Pastafarianism. Allow yourself to be touched by His Noodly Appendage, and you shall surely enter into Heaven where you can enjoy the endless pleasures of the Beer Volcano and Stripper Factory for all eternity.

If you insist upon keeping to your heretical beliefs, you will surely suffer in the afterlife. The FSM will cast you into a lake of boiling Ragu, where you shall simmer for all eternity with the rest of the unbelievers. You've been warned.

It's true - my subjective spiritual experience with the FSM told me so.

Prove me wrong.

ETA: Besides, I'm not the only one. There is even a FaceBook group for the FSM - millions of us have been touched by His Noodly Appendage! RA-men!!!

 

[69dodge]

I think the smallest confusion is the idea that P(E) must equal 1, simply because we've observed the data E.

That's equivalent to saying that anything that did happen must have happened; there are no contingent events or free will. If I named my cat Lucretius, then there were no other possible names I could have given my cat.

No, it's just saying that what is known to have happened is known to have happened. If you named your cat "Lucretius", its name is "Lucretius". Of course, before you named it, its future name was unknown. And, of course, you are free to give your next cat a different name. Still, the probability is now 1 that your current cat is named "Lucretius".

Are we assuming that we know E has happened, or are we assuming that we don't know that?

We'll get very confused if we don't specify which, and especially if we unconsciously switch back and forth, saying simply "P(E)" for both.

Probabilities conditioned on different conditions are quite incomparable, which is why I find it especially strange that rocketdodger is adding them together, in P(E|H) + P(E|~H). What, conceptually, is that sum supposed to represent? Well, it's the sum of two probabilities, so presumably it's also a probability. But of what? And, more importantly, on what condition? (I guess those are questions for him, not you.)

 

[MattusMaximus]

I wasn't talking about the Christian God. I was talking about be agnostic about God (the existence of some powerful supernatural deity).

Again, you're thinking of the Christian God. Plenty of people believe there might be a God that started the whole thing and doesn't really have much to do with it now.

Little do you know, Malerin, that this "God" of which you speak is simply the end-point of one of His Divine Noodly Appendages. The FSM is so much more than this false, vague "God" you keep mentioning - this "God" is merely an illusion, put in place to test your faith in the One, True path of Pastafarianism.

For your sake, I encourage you to seek out His Holy Prophet, Chef Boy-R-Dee, in an attempt to save your appetite from certain disappointment! Oh yeah, and that whole pit of flaming Ragu would totally suck as well...

Really, Malerin, I'm not making this stuff up. My subjective spiritual experience justifies everything I'm saying.

ETA: RA-men!!!

 

[Malerin]

I don't. I think our knowledge of it is imprecise.



Not at all. I'm fond of ice cream, but the probability that I'm eating it at any given instant is rather low. The universe creator may be fond of life, but that doesn't mean that everything He does produces life.

The fact is that you have no actual data to support any particular estimate of P(E|H), and all you have is a rather vague hand-wave, but you want to turn that into an actual number.

Or the fine-tuner was created after the universe began. Good points.

I don't know if you can formulate H so that Pr(E/H) is high and Pr(E/~H) is low.

If H is strong (e.g., "God created the universe"), Pr(E/H) is high (if you define God as a life-loving kind of deity), but ~H is "God did not create the universe". That doesn't entail nothing created the universe. Maybe an evil genius did, or the FSM, or a super-intelligent alien. Pr(E/~H) won't be as low as the argument needs it to be.

If H is weak (e.g., "Something created the universe"), you get a nice negation (~H= "Nothing created the universe") but I can't see
Pr(E/H) >.5. That leaves all the heavy lifting to Pr(E/~H).

 

[RandFan]

No, because I have other information based on other closets and other people.

But don't you have information about the world in general? Don't you know that most binary choices are not really 50/50? Don't you know about statistics and deviation from the mean?

In a sense, you're actually proving Malerin's case for him.

In a sense, no. It's demonstrable that when we look at the world, many if not the majority of dialectical choices are not 50/50.

So you are actually "proving" (demonstrating) my case for me. Anyone who has experience with the world would know that a .5 probability for an unknown dialectic is nonsense. The world clearly does not work that way.

If you knew nothing about money and closets, then it would be a good bet for you.

Based on what theory? An argument from ignorance? If we could magically take away everyone's memory would all dialectic decisions magically become .5? No. We know that they are not so weighted. Hell, we know that even a coin is likely biased.

 

[69dodge]

Hell, we know that even a coin is likely biased.

Yeah, but if you don't know whether it's biased towards heads or towards tails, do you have any more reason to bet on heads than on tails?