Proof of Immortality, VI

Status
Not open for further replies.
Jabba said:
Dave,
1. I’m back!
2. Can’t help myself.
3. Simple Bayesian formula: P(I|E)=P(E|I)*P(I)/P(E)
4. I: I’m immortal
5. E: I currently exist
6. If I allow for
6.1. a 1% prior probability for my immortality, and
6.2. an unimaginably small number for the prior probability of me currently existing, and
7. If P(E|I) is NOT an example of the Texas Sharpshooter fallacy,
7.1. P(I|E)=1*.01/.00000000000…1=.9999999…9, and
7.2. I must be immortal.
8. Am I using the formula properly?
Click to expand...

No. You're not using the formula properly. It is impossible for the denominator P(E) to be less than the numerator P(E|I)P(I). The fact that it is is proof that you're not doing anything Bayesian at all; you're just making up numbers. Now go away.
 
Last edited:
jt512 said:
No. You're not using the formula properly. It is impossible for the denominator P(E) to be less than the numerator P(E|I)P(I). The fact that it is is proof that you're not doing anything Bayesian at all; you're just making up numbers. Now go away.
Click to expand...

I seriously don't know if I want to get involved in this discussion at all... but...

By what reasoning would anyone make the ridiculous assumption that P(E) is 'unimaginably small'? Does Jabba believe that he doesn't exist? Or does he have really incredibly serious doubt about whether he exists? Gotta say, this is some really strange abuse of poor Mr. Bayes... :boggled:
 
Emily's Cat said:
I seriously don't know if I want to get involved in this discussion at all... but...

By what reasoning would anyone make the ridiculous assumption that P(E) is 'unimaginably small'? Does Jabba believe that he doesn't exist? Or does he have really incredibly serious doubt about whether he exists? Gotta say, this is some really strange abuse of poor Mr. Bayes... :boggled:
Click to expand...

Oh no you didn't!

Now it's gonna start all over again!eleventy!!!!
 
Emily's Cat said:
I seriously don't know if I want to get involved in this discussion at all... but...

By what reasoning would anyone make the ridiculous assumption that P(E) is 'unimaginably small'? Does Jabba believe that he doesn't exist? Or does he have really incredibly serious doubt about whether he exists? Gotta say, this is some really strange abuse of poor Mr. Bayes... :boggled:
Click to expand...
It is is unlikely that anything exists, which is not immortal. Jabba exists, therefore it is likely that Jabba is immortality. If we compute the likelihood of existence without immortality to infinitesimal, then we can compute the likelihood of immortality to be virtually certain, for anything that exists. The only other requirement for immortality being that the thing which exists must have a soul with a sense of self.
 
theprestige said:
It is is unlikely that anything exists, which is not immortal. Jabba exists, therefore it is likely that Jabba is immortality. If we compute the likelihood of existence without immortality to infinitesimal, then we can compute the likelihood of immortality to be virtually certain, for anything that exists. The only other requirement for immortality being that the thing which exists must have a soul with a sense of self.
Click to expand...

Lol. That's entertaining.

Seriously though, this is kind of the opposite of the problem I usually run into. I end up seeing people really over-weighting the likelihood of an event, because they're looking at it after the fact. So the fact that it did happen ends up translating in their brains to it being highly likely to happen.

This though... this is taking a thing that is known and actual and then ascribing an infinitesimal likelihood to it's actuality? What is this, psychedelic statistics? Mathing while high? Reasoning under the influence?
 
Emily's Cat said:
I seriously don't know if I want to get involved in this discussion at all... but...

By what reasoning would anyone make the ridiculous assumption that P(E) is 'unimaginably small'? Does Jabba believe that he doesn't exist? Or does he have really incredibly serious doubt about whether he exists? Gotta say, this is some really strange abuse of poor Mr. Bayes... :boggled:
Click to expand...


P(E) = P(E|H1)P(H1) + P(E|H2)P(H2) + . . . .

In other words, P(E) is a weighted average of the conditional probabilities of E given each possible hypothesis, with the weights given by the prior probabilities of those hypotheses. E is the observed evidence, so we know E occurred. The question is how likely was its occurrence under a set of competing hypotheses. If E is unlikely under all hypotheses, then P(E) can be small. It just means that E was unlikely to have occurred. However, it is clear from the above equation that P(E) ≥ P(E|H_i)P(H_i) for all i.
 
Last edited:
Emily's Cat said:
Lol. That's entertaining.

Seriously though, this is kind of the opposite of the problem I usually run into. I end up seeing people really over-weighting the likelihood of an event, because they're looking at it after the fact. So the fact that it did happen ends up translating in their brains to it being highly likely to happen.

This though... this is taking a thing that is known and actual and then ascribing an infinitesimal likelihood to it's actuality? What is this, psychedelic statistics? Mathing while high? Reasoning under the influence?
Click to expand...
Your hilite stops short of the second, equally important clause in that sentence.
 
Jabba said:
7. If P(E|I) is NOT an example of the Texas Sharpshooter fallacy,
Click to expand...

Of course it isn't, the TS fallacy has nothing to do with using your own existence in a conditional and there are no such things as "valid targets" or "invalid targets", just ignore all that.

The TS fallacy is the fallacy of switching the conditional.

Let A be "shot the side of the barn such that it counts as a hit"
Let B be "drew a difficult target on the side of the barn"

Then a sharpshooter can be considered as P(sharpshooter) = P(A|B), ie by the probability that you can score a hit given that a difficult target has been drawn. Someone using the TS fallacy would be arguing P(sharpshooter) = P(B|A), ie that they're a sharpshooter because they can draw a difficult target given a shot at the side of the barn that needs to count as a hit.

In your case, if you were making a TS fallacy you'd be arguing "the probability of me being immortal is P(E|I)" and you're making a lot of errors, but you're not making that particular one.

Your actual problem is this assertion: P(E|~H) > P(E|H)
 
Dave Rogers said:
The assertion is actually more like P(E|~H)>>P(E|H). Jabba's not just arguing that his existence is more likely under ~H; he's arguing that it's so much more likely that its probability can be neglected under H.

Dave
Click to expand...

True, but I decided against putting >> because it might give the impression that just > would be fine. And while it's true that Jabba is making that particular error in gigantic proportions, that still doesn't stop it from being an error even if it were made in more milder proportions.
 
caveman1917 said:
True, but I decided against putting >> because it might give the impression that just > would be fine. And while it's true that Jabba is making that particular error in gigantic proportions, that still doesn't stop it from being an error even if it were made in more milder proportions.
Click to expand...

Fair point. It's a very good example of something I coined the term "unevaluated inequality fallacy" to describe, quite a few years ago: the unsupported assertion that, though the value of a specific quantity is unknown, common sense dictates that it must be greater than the value of some other specific quantity, which in the strong form of the fallacy is also unknown. It's surprising how often it gets invoked.

Dave
 
theprestige said:
Your hilite stops short of the second, equally important clause in that sentence.
Click to expand...

Meh. Seemed like the second clause was tautological, and didn't merit scrutiny.

Claim(s): It's unlikely that anything exists
[subclaim] unless that thing is immortal

Conclusion: I exist, therefore I am immortal

The immortality element of it seemed to be whatever the technical term is for assuming the conclusion to be true, then using that assumed conclusion as part of the argument for why the conclusion is true. There's a term, but I can't recall it.

Pretty much, you don't have to go beyond the assumption that the likelihood of anything existing is really small. That's not even a good assumption, it's obviously incorrect in a pretty dramatic fashion. Nothing beyond that false assumption bears consideration.
 
Last edited:
Emily's Cat said:
The immortality element of it seemed to be whatever the technical term is for assuming the conclusion to be true, then using that assumed conclusion as part of the argument for why the conclusion is true. There's a term, but I can't recall it.
Click to expand...

Circulus in demonstrando, aka circular reasoning.
 
Status
Not open for further replies.

Back
Top Bottom