Proof of Immortality, the IV league

[aleCcowaN]

Why do you continue to ignore the scientific model which indicates that your "self" and your body are the same thing?

Because his epistemological hedonism, what else? Branding oneself in different ways gives a lot for speculation and fringe exploit.

Come to a certain point it should be clear that it wasn't the intellectual aspect the only one related to the issue.

 

[Darat]

Several posts have been moved to "Deep Storage" whilst the mod team discuss confidential matters, there was nothing wrong with the posts themselves and hopefully the matter can be quickly resolved and they are returned.

Posted By: Darat

Please remember that discussions about moderation belong in the Forum Management section. In regards to this matter you of course can bring it up in the Forum Management section but the mod team will not be providing any further details at this time.

Replying to this modbox in thread will be off topic  Posted By: Darat

 

[Jabba]

Indeed, but what evidence do you have that it actually is currently occupying a body?

Mojo,
- I don't get your question. Somebody's fingers are (slowly) typing this message...

 

[theprestige]

Mojo,
- I don't get your question. Somebody's fingers are (slowly) typing this message...

Some process is slowly typing, you mean.

 

[Mojo]

Mojo,
- I don't get your question. Somebody's fingers are (slowly) typing this message...

Yes, but is it you or somebody else? According to you it could be any one of an infinite number of potential selves, only one of which is you.

 

[Mojo]

Mojo,
- I don't get your question. Somebody's fingers are (slowly) typing this message...

I notice that you don't say that they are your fingers.

 

[jt512]

This actually works quite well for the "desert survivors" problem. There is no God, so any evidence which implies there is is wrong.

(posterior odds of God hypothesis) = (Bayes factor*) × (prior odds of God hypothesis)

0 = 850 × 0

Seems to be quite reasonable, actually.

_______________

*Bayes factor being the weight of the evidence in favor of the God hypothesis over a particular alternative hypothesis.

 

[Loss Leader]

Is this debate, Jabba?

Is this what you asked for? A single discussion of each indivisible detail with one person asking you one question at a time - that's what we had. But you chose to abandon that rather than attempt to work through your problem.

Some have followed you in your great circle. I, however, do not believe I will.

We were stuck at you needing to show evidence that some sort of immaterial perception is at least possible. The closest you got was begging that we read some stories you found as though they legitimately introduced doubt.

I asked you for the name of the patient who recounted another amputation in an adjacent room. I asked you for the work phone number of the nurse who found a penny on top of a cabinet. If these doesn't make the anecdotes repeatable, at least they can help rule out some confounding variables.

And that is when you abandoned me. Please come back with that information because that, and nothing else, is the step we are on.

 

[Nonpareil]

*Bayes factor being the weight of the evidence in favor of the God hypothesis over a particular alternative hypothesis.

Which isn't necessarily 33,000.

Again, for those keeping score at home, the 33,000 jt512 is using arises from Bayes' theorem:

p(cheat|royal) = p(royal|cheat) * p(cheat) / p(royal)

That is, given that I have a royal flush, the probability that I cheated is equal to the probability that I have a royal flush if I cheat, times the probability that I cheated, divided by the probability of getting a royal flush fairly.

The odds that I get a royal flush, if I cheat, are one. The odds of getting a royal flush fairly are .003%. Given this, we can simplify to:

p(cheat|royal) = 1/.003% * p(cheat) = 33,333 * p(cheat)

But 33,333 is not necessarily the Bayes factor.

The Bayes factor is a comparative measure between two alternative hypotheses. It is denoted as k, and is calculated as:

k = p(flush|cheating) / p(flush|other hypothesis)

Or, "The Bayes factor for any two hypotheses about how I got this royal flush is equal to the probability of getting a flush in one hypothesis divided by the probability of getting a flush in the other."

We have already established that the probability that I get a flush if I cheat is one, so:

k = 1 / p(flush|other hypothesis)

Now, if we just say the other hypothesis is "drawing random cards", then yes, we end up with 1/.003%, and our original result.

But the thing is that you can put literally anything else in there for "other hypothesis", and so long as it ends with saying "...so I get a royal flush", p(flush|other hypothesis) is one. This is up to, and including, "the dealer happened to shuffle the deck in exactly this way, entirely unintentionally".

So the Bayes factor is not, in fact, 33,000. It can be literally anything, including 1/1, even if the alternative hypothesis is that it happened entirely randomly. There is literally nothing preventing this sort of manipulation.

Which is why we go back to what I said before. In order for any sort of Bayesian analysis to have any weight whatsoever, you must be able to provide evidence that indicates your hypothesis is actually true.

Otherwise, you just end up with the same old issues.

 

[jt512]

Which isn't necessarily 33,000.

Not that it really matters, since it's getting multiplied by 0, but I meant to use the Bayes factor for the God hypothesis calculated earlier, which was about 850, but inadvertently used the one from the poker problem. Will edit my post accordingly.

 

[Nonpareil]

Not that it really matters, since it's getting multiplied by 0, but I meant to use the Bayes factor for the God hypothesis calculated earlier, which was about 850, but inadvertently used the one from the poker problem. Will edit my post accordingly.

Which is fair enough, but it still doesn't change anything.

You can just as easily frame the hypotheses in the desert wanderers hypothetical in such a way that the Bayes factor is one. It still comes down to providing evidence that any given hypothetical is actually true, whether it's p(cheating) or p(god).

 

[jt512]

Which is fair enough, but it still doesn't change anything.

You can just as easily frame the hypotheses in the desert wanderers hypothetical in such a way that the Bayes factor is one. It still comes down to providing evidence that any given hypothetical is actually true, whether it's p(cheating) or p(god).

What you're saying makes no sense, as usual. I can't figure out what your misconception about Bayesian inference is, so I can't help.

 

[jt512]

The odds that I get a royal flush, if I cheat, are one. The odds of getting a royal flush fairly are .003%. Given this, we can simplify to:

p(cheat|royal) = 1/.003% * p(cheat) = 33,333 * p(cheat)

So far, so good.

The Bayes factor is a comparative measure between two alternative hypotheses. It is denoted as k, and is calculated as:

k = p(flush|cheating) / p(flush|other hypothesis)

Or, "The Bayes factor for any two hypotheses about how I got this royal flush is equal to the probability of getting a flush in one hypothesis divided by the probability of getting a flush in the other."

Still good.

We have already established that the probability that I get a flush if I cheat is one, so:

k = 1 / p(flush|other hypothesis)

Now, if we just say the other hypothesis is "drawing random cards", then yes, we end up with 1/.003%, and our original result.

Still good.

But the thing is that you can put literally anything else in there for "other hypothesis", and so long as it ends with saying "...so I get a royal flush", p(flush|other hypothesis) is one. This is up to, and including, "the dealer happened to shuffle the deck in exactly this way, entirely unintentionally".

Now you start to go off the rails.

So the Bayes factor is not, in fact, 33,000. It can be literally anything, including 1/1, even if the alternative hypothesis is that it happened entirely randomly. There is literally nothing preventing this sort of manipulation.

And now you have gone completely off the rails. First you say that the probability of getting a royal flush randomly is .003%, and therefore for the BF for cheated vs random is 33,000; and then you completely contradict yourself, saying that the Bayes factor could be 1, for comparing those very same hypotheses using the very same data.

 

[Loss Leader]

I'm starting to come around to the opinion that a person needs to be licensed by the World Health Organization in order to use Bayes' Theorem.

 

[JoeMorgue]

It is largely seeming like Bayes is becoming the Mathematical equivalent to Quantum Theory, a real fascinating branch of science... that for some reason attracts a certain type of person who looks at it all and they can see is "Well this sciency thing can be used to explain seemingly unlikely things, so obviously I can invoke it as a magic word to excuse away my unlikely claim and those big mean skeptics can't say anything because I'm using their precious science!"

 

[Pixel42]

Some have followed you in your great circle. I, however, do not believe I will.

Agreed. I actually started to write a post explaining the Texas sharpshooter fallacy for the umpteenth time, but then came to my senses. It's time we insisted Jabba actually completes a discussion about one of his sub sub points instead of switching to another or just going right back to the beginning as if the last four years never happened.

 

[jt512]

Agreed. I actually started to write a post explaining the Texas sharpshooter fallacy for the umpteenth time, but then came to my senses. It's time we insisted Jabba actually completes a discussion about one of his sub sub points instead of switching to another or just going right back to the beginning as if the last four years never happened.

Why not just ignore him.

 

[Pixel42]

Why not just ignore him.

Because he interprets giving up responding to him as conceding that he has won the argument. Last time everyone just gave up trying to get him to understand the fallacies he was spouting he started talking about how he was going to win the MDC with his now vindicated proof of immortality. Needless to say he got some responses to that, and here we all still are a couple of years later.

 

[jt512]

Because he interprets giving up responding to him as conceding that he has won the argument. Last time everyone just gave up trying to get him to understand the fallacies he was spouting he started talking about how he was going to win the MDC with his now vindicated proof of immortality. Needless to say he got some responses to that, and here we all still are a couple of years later.

I think the world will continue to turn if some guy on the internet is allowed to think he can use Bayes' Theorem to prove he's immortal.

 

[JoeMorgue]

Why not just ignore him.

Because this death spiral that dares to call itself a discussion stopped being for Jabba's benefit a thousand posts, 4 years, a hundred nested arguments, dozens of fringe resets, and an entirely different message board ago.

Every few weeks someone pokes their head into this thread and tries to tut-tut us over still arguing with Jabba and I don't really get why. What exactly do you think we don't get? That we aren't getting through to him? Yeah we get that. I don't think any of us are trying to at this point.

We're not engaging with Jabba at this point. We're just engaging at him.