[Merged] Immortality & Bayesian Statistics

[Kevin_Lowe]

Kevin and the gang,

- That's the point. 4 aces is no more improbable than any other specific hand. But then, 4 aces does raise a somewhat plausible possibility that the other guy's hand is not simply random -- when most other specific hands would not raise that concern… That’s the point.

Yes, because based on previous experience with people dealing cards, we have a rationally-arrived-at prior probability that any given person dealing cards is up to some kind of monkey business.

- My claim is that my own existence at this time does raise that kind of plausible possibility in regard to human mortality/immortality. And, if I stick that possibility into the Bayesian formula, I see that my existence right now is probably not a random hand dealt from a one-short-lifetime-deck.

You can't know anything of the sort, because you don't have prior experience with multiple universes in which conditions differ. You just don't have prior knowledge of how the universe might be, but I do have prior knowledge of how people get up to monkey tricks with cards.

- So now, besides dropping the whole issue, how would you guys like me to proceed?

Stop abusing Bayesian statistics please?

They aren't a magic wand. The GIGO rule (garbage in, garbage out) applies to all forms of statistics be they frequentist, Bayesian or Callathumpian. Framing a dumb idea ("I'm immortal because I think I'm special!") in Bayesian terms does not make it less dumb, even by an iota.

 

[Jabba]

For me, this is the most crucial part of your argument. You have completely failed to show any reason to suspect that you are in any way special, or that anyone is in any way special.

It's the question I keep asking you.

How do you know that you are the ace of spades rather than the three of clubs?

Claim it all you want, but your conclusion is totally founded on this assumption.

If you cannot show this to be true beyond any doubt then your conclusion will never be a proof. It will, at best, be an interesting mental exercise.

Wollery,

- I am not the ace of spades. I am the hand of aces -- as are we all... I'm claiming that we are all "special."

- Firstly, I'm claiming that my conscious existence right now is extremely improbable.
- But then, what specific event is not extremely improbable? Extremely improbable events happen all the time. They HAVE to happen. I should get used to it.
- Yet, certain "kinds" of improbable events rightfully give us pause... We think, "This is an interesting coincident..." I'm claiming that such events give us pause because we can think of at least one other somewhat plausible explanation. We don't have to accept that such events are random and just happen. I'm claiming that such is the case right here. Your existence right now, according to the "scientific" model, is SO improbable that almost any potential explanation would rightfully supercede it. And, in determining the odds, we compare the probability of a single (or less) short lifetime in all of eternity (whatever that is) to its "compliment" -- the probability of one of all the other possibilities being the right one.

- The following are at least some of the other possible explanations.
1) Reincarnation.
2) Each of us is an eternally basic part of reality.
3) "Now" isn't what we think it is.
4) We aren't as smart as we think we are.

- Actually, I'm not sure that answers your question...
- So far, I'm claiming that we are each special because each of our existences suggest possible explanations that -- certainly, in combination -- are much more probable explanations than is chance. But then, there seems to be another, this time "ineffable," question behind that answer. Hopefully, that's temporary, or specific to me -- maybe, you can express that question...
-I'll leave it there for now -- I have miles yet to go.

--- Jabba

 

[zooterkin]

Wollery,

- I am not the ace of spades. I am the hand of aces -- as are we all... I'm claiming that we are all "special."

Why? It's the same principle; why are we a hand of aces as opposed to any other hand, which is equally likely?

 

[Pixel42]

I am not the ace of spades. I am the hand of aces -- as are we all... I'm claiming that we are all "special."

You are making the same fundamental mistake as the puddle that thinks the hole it sits in was designed specifically for it, as the odds of it ending up the exact shape required to fit it are so astronomically small.

 

[Squeegee Beckenheim]

- Yet, certain "kinds" of improbable events rightfully give us pause... We think, "This is an interesting coincident..." I'm claiming that such events give us pause because we can think of at least one other somewhat plausible explanation.

...which brings us back to doh, oh, oh, oh...

Because human beings have an in-built biological tendency to see significance where there is none, and to see patterns where there are none.

 

[Recovering Agnostic]

- Firstly, I'm claiming that my conscious existence right now is extremely improbable.
- But then, what specific event is not extremely improbable? Extremely improbable events happen all the time. They HAVE to happen. I should get used to it.
- Yet, certain "kinds" of improbable events rightfully give us pause... We think, "This is an interesting coincident..." I'm claiming that such events give us pause because we can think of at least one other somewhat plausible explanation. We don't have to accept that such events are random and just happen. I'm claiming that such is the case right here. Your existence right now, according to the "scientific" model, is SO improbable that almost any potential explanation would rightfully supercede it. And, in determining the odds, we compare the probability of a single (or less) short lifetime in all of eternity (whatever that is) to its "compliment" -- the probability of one of all the other possibilities being the right one.

So explain to me how you deduce the probabilities.

The bare facts are that however likely or not you consider it, we exist. The only questions of relevance are conditional on that fact. We don't care if the bare probability of us existing is a googolplex to one - all that matters is the conditional probability of our existence being explained by the scientific model, as you put it, given that we exist. It's a completely different question.

And as far as I can see, none of this has anything to do with immortality.

 

[Jabba]

Wollery,

- I am not the ace of spades. I am the hand of aces -- as are we all... I'm claiming that we are all "special."

- Firstly, I'm claiming that given the scientific model that we each have -- at most -- one short life to live, my conscious existence right now is extremely improbable...

Wollery,

- I just inserted the blue part. I claim that this probability approaches zero. And then, if I insert that probability into the Bayesian formula, I get that the probability of the scientific model being true also approaches zero. It's my understanding that this kind of situation is exactly what the Bayesian formula is about -- how the addition of new evidence affects the probability of an existing model.

- I keep catching stomach viruses from my grandchildren, my own lap top is being repaired, and I wanted to show that I'm still around... I'll try to clean up my whole story and present it as concisely as possible in my next post.

--- Jabba

 

[Kevin_Lowe]

- I just inserted the blue part. I claim that this probability approaches zero. And then, if I insert that probability into the Bayesian formula, I get that the probability of the scientific model being true also approaches zero. It's my understanding that this kind of situation is exactly what the Bayesian formula is about -- how the addition of new evidence affects the probability of an existing model.

Like I said before, garbage in = garbage out.

If you start with a dumb prior probability, then of course you can generate a dumb posterior probability.

 

[Baffled]

Day 6 of my current experiment into immortality.

I want to be immortal too! Where's my deck of cards!

...uh-oh, I drew the Ace of Spades, but it's upside down. Does that mean I was never born? hold me, I'm scared

Still appear to be alive.

Have interviewed two witnesses to my alleged birth, both of whom wish to remain anonymous. Their testimonies appear consistent with few disparities and a low probability of having been falsified. To judge from their sworn statements, I either looked like a large meatball or a small moose. Not sure how to converge those images....

 

[Giordano]

Jabba,

The chance of any one sperm fertilizing the ovum that became you was less than one in a million, yet here you are. Your sperm was the winner, and the millions of other sperm lost out. Another mating, and your sister was the winner, and the sperm that might have produced an almost-you lost out. How is that unlikely, and how does that prove immortality?

 

[AdMan]

Jabba, your argument is the same as if you had a thoroughly shuffled deck of 52 cards in a particular order, and you worked backward to figure out what the chances were that that specific order would come up. Because the chances of 52 cards being arranged by chance in that particular order are so incredibly small, your argument sounds like you could then claim that it had to be predestined.

What you don't seem to understand is that you are working backward from that specific order, or in your case, from your own existence. In either case, the argument for predestination or God or immortality--whatever you want to call it--is wrong.

If the cards started off in new deck order, you thoroughly and fairly shuffled them and then arrived at an order that had been previously forecast, you would have something (even though that could still happen purely by chance, small as it is). But working backward, no.

 

[Pixel42]

I claim that this probability approaches zero.

The probability that a hole would happen to end up exactly the right shape to fit the puddle of water that now sits in it also approaches zero. Except it doesn't, does it? The shape of the puddle is determined by the shape of the hole so no matter what shape the hole is, and how unlikely it was that it would end up that particular shape, the puddle will still fit it exactly.

You keep making the same fundamental mistake. Why don't you try to grasp the nature of that mistake, instead of constantly repeating it whilst ignoring those who are explaining what you're doing wrong?

 

[pakeha]

Jabba, is this related to the Drake Equation?
http://www.youtube.com/watch?v=qMybbu_1hAY

 

[wollery]

Wollery,

- I just inserted the blue part. I claim that this probability approaches zero. And then, if I insert that probability into the Bayesian formula, I get that the probability of the scientific model being true also approaches zero. It's my understanding that this kind of situation is exactly what the Bayesian formula is about -- how the addition of new evidence affects the probability of an existing model.

You keep claiming that the probability approaches zero, but you have yet to show it. In fact you have yet to get within the same galaxy as showing it.

Bayes theorem is only reliable when applied to test measurements, i.e. actual measurable results of experiments.

You have no measurable results, just a belief. Let's assume, just for the sake of argument, that Baye's theorem applies to this case. Now look at what happens if you assume that the probability of you existing right now is 99.99999% based on the scientific model. Instead of assuming that P(Alive now based on scientific model) = 0.000001% try plugging 99.99999% into your equations and see what happens.

Then get back to us.

 

[Humots]

Wollery,

- I just inserted the blue part. I claim that this probability approaches zero. And then, if I insert that probability into the Bayesian formula, I get that the probability of the scientific model being true also approaches zero. It's my understanding that this kind of situation is exactly what the Bayesian formula is about -- how the addition of new evidence affects the probability of an existing model.

--- Jabba

Jabba: let's look at the math.

In your link http://messiahornot.com/Act2Scene2.php, you state:

Given
...k = all background knowledge,
...P = the probability of,
...NR = Non-Religious hypothesis,
...| = given,
...me = me (my existence),
...R = Religious hypothesis.
The formula for this probability is
...P(NR|me & k) = P(me|NR)P(NR|k) / (P(me|NR)P(NR|k) + P(me|R)P(R|k)).

Shouldn't the initial Bayes formula for your probability be

P(NR|me & k) = P(me & k|NR)P(NR) / (P(me & k|NR)P(NR) + P(me & k|R)P(R)) = P(me & k|NR)P(NR) / P(me & k)

assuming NR = not R, so P(NR) + P(R) = 1.0?

Note that this is a straightforward application of Bayes Theorem to determine the conditional probability P(A|B) (see http://en.wikipedia.org/wiki/Bayes'_theorem)

P(A|B) = P(B|A) P(A) / ( P(B|A) P(A) + P(B|not A) P(not A) ) = P(B|A) P(A) / P(B)

and assigning
A = NR
and
B = (me & k) | NR and (me & k) | R = me & k

again assuming NR = not R, so P(NR) + P(R) = 1.0

If this initial Bayes formula is not correct, please explain why.

If this initial Bayes formula is correct, please explain the steps leading from the initial Bayes formula to your formula.

After that, we can talk about the values you assume for P(NR), P(R) and especially P(me & k). And your definition of "|" as "given" in a non-mathematical sense.

 

[Jabba]

Jabba: let's look at the math.

In your link http://messiahornot.com/Act2Scene2.php, you state:


Shouldn't the initial Bayes formula for your probability be

P(NR|me & k) = P(me & k|NR)P(NR) / (P(me & k|NR)P(NR) + P(me & k|R)P(R)) = P(me & k|NR)P(NR) / P(me & k)

assuming NR = not R, so P(NR) + P(R) = 1.0?

Note that this is a straightforward application of Bayes Theorem to determine the conditional probability P(A|B) (see http://en.wikipedia.org/wiki/Bayes'_theorem)

P(A|B) = P(B|A) P(A) / ( P(B|A) P(A) + P(B|not A) P(not A) ) = P(B|A) P(A) / P(B)

and assigning
A = NR
and
B = (me & k) | NR and (me & k) | R = me & k

again assuming NR = not R, so P(NR) + P(R) = 1.0

If this initial Bayes formula is not correct, please explain why.

If this initial Bayes formula is correct, please explain the steps leading from the initial Bayes formula to your formula.

After that, we can talk about the values you assume for P(NR), P(R) and especially P(me & k). And your definition of "|" as "given" in a non-mathematical sense.

Humots,

- This will take me awhile to figure out...
- Maybe, old dogs can learn new tricks.
- But, just to get us started,

-From http://www.answers.com/topic/bayes-theorem#Bayesian_interpretation
In the special case of a binary partition,
P(A|B) = P(B|A)P(A)/(P(B|A)P(A)+P(B|non-A)P(non-A)

- I assume that in this formula, the "background knowledge" is also just assumed.
- And, isn't ours a case of "binary partition"?

--- Jabba

 

[Humots]

Jabba: let's look at the math.

In your link http://messiahornot.com/Act2Scene2.php, you state:


Shouldn't the initial Bayes formula for your probability be

P(NR|me & k) = P(me & k|NR)P(NR) / (P(me & k|NR)P(NR) + P(me & k|R)P(R)) = P(me & k|NR)P(NR) / P(me & k)

assuming NR = not R, so P(NR) + P(R) = 1.0?

Note that this is a straightforward application of Bayes Theorem to determine the conditional probability P(A|B) (see http://en.wikipedia.org/wiki/Bayes'_theorem)

P(A|B) = P(B|A) P(A) / ( P(B|A) P(A) + P(B|not A) P(not A) ) = P(B|A) P(A) / P(B)

and assigning
A = NR
and
B = (me & k) | NR and (me & k) | R = me & k

again assuming NR = not R, so P(NR) + P(R) = 1.0

If this initial Bayes formula is not correct, please explain why.

If this initial Bayes formula is correct, please explain the steps leading from the initial Bayes formula to your formula.

After that, we can talk about the values you assume for P(NR), P(R) and especially P(me & k). And your definition of "|" as "given" in a non-mathematical sense.

Humots,

- This will take me awhile to figure out...
- Maybe, old dogs can learn new tricks.
- But, just to get us started,

-From http://www.answers.com/topic/bayes-theorem#Bayesian_interpretation
In the special case of a binary partition,
P(A|B) = P(B|A)P(A)/(P(B|A)P(A)+P(B|non-A)P(non-A)

- I assume that in this formula, the "background knowledge" is also just assumed.
- And, isn't ours a case of "binary partition"?

--- Jabba

Good, you're looking at a reference.

(1) Bayes' Theorem is a mathematical equation. Every event referenced in this equation is a specific event with an associated probability, even if we don't know what that probability is. "Background knowledge" is not a specific event with a probability. It is not "just assumed", it has no meaning in this context.

(2) It is a binary partition if NR has the same meaning as non-R.

 

[carlitos]

The probability that a hole would happen to end up exactly the right shape to fit the puddle of water that now sits in it also approaches zero. Except it doesn't, does it? The shape of the puddle is determined by the shape of the hole so no matter what shape the hole is, and how unlikely it was that it would end up that particular shape, the puddle will still fit it exactly.

You keep making the same fundamental mistake. Why don't you try to grasp the nature of that mistake, instead of constantly repeating it whilst ignoring those who are explaining what you're doing wrong?

Jabba will continue to ignore this and play with math instead.

 

[dafydd]

Would it possible for the mod to delete the word.''immortality'' from the thread title?

 

[Acleron]

Would it possible for the mod to delete the word.''immortality'' from the thread title?

In case it comes true? (for the thread, not the premise)