[Merged] Immortality & Bayesian Statistics

[dafydd]

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

Good point, Jabba's evidence-free shroud thread seems never ending.

 

[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.

Humots,
- Re #1, I still don't understand. Isn't the probability of NR (or A) based upon "background (or "prior") knowledge"?
- Re #2, the two do have the same meaning.
--- Jabba

 

[dafydd]

Humots,
- Re #1, I still don't understand. Isn't the probability of NR (or A) based upon "background (or "prior") knowledge"?
- Re #2, the two do have the same meaning.
--- Jabba

Please at least try to answer a question, for once. Have you dropped the immortality fantasy?

 

[wollery]

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.

Jabba, have you tried this yet?

If not, why not?

It's very easy to do, and may give you a revealing insight into the problem.

 

[Humots]

Humots,
- Re #1, I still don't understand. Isn't the probability of NR (or A) based upon "background (or "prior") knowledge"?
- Re #2, the two do have the same meaning.
--- Jabba

Jabba: Again, Bayes' Theorem is a mathematical equation. It is not some kind of syllogism.

The expression P(A|B) is the probability of event A "given" event B in a strictly defined technical sense. It means the probability of event A happening given that event B has happened. Event A is in some way affected by event B.

Event B is not "background knowledge". It is a specific event with its own probability.

If events A and B have nothing to do with each other, then P(A|B) = P(A). The event A = "I draw an ace" has nothing to do with the event B = my wife is 7' 8" tall. For a standard, unstacked deck, P(I draw an ace) = 4/52 whatever my wife's height.

I am talking about the syntax and technical meaning of Bayes' Theorem.

P(A|B) = P(B|A) P(A) / P(B)

is based on a specific mathematical syntax, that A and B are specific events with specific probabilities.

You are still getting the basics wrong.

We haven't even gotten to whether P(NR) or P(R) make any sense.

 

[shuttlt]

Jabba, have you tried this yet?

If not, why not?

It's very easy to do, and may give you a revealing insight into the problem.

I'm going to guess 99.9999999999999999999999%, but since that's so close to 99.99999% I'm sure it won't make a much of a difference to the answer.

This is the same maths that has been used to conclusively demonstrate that Amanda Knox is innocent, so I have high hopes.

 

[Jabba]

Jabba: Again, Bayes' Theorem is a mathematical equation. It is not some kind of syllogism.

The expression P(A|B) is the probability of event A "given" event B in a strictly defined technical sense. It means the probability of event A happening given that event B has happened. Event A is in some way affected by event B.

Event B is not "background knowledge". It is a specific event with its own probability.

If events A and B have nothing to do with each other, then P(A|B) = P(A). The event A = "I draw an ace" has nothing to do with the event B = my wife is 7' 8" tall. For a standard, unstacked deck, P(I draw an ace) = 4/52 whatever my wife's height.

I am talking about the syntax and technical meaning of Bayes' Theorem.

P(A|B) = P(B|A) P(A) / P(B)

is based on a specific mathematical syntax, that A and B are specific events with specific probabilities.

You are still getting the basics wrong.

We haven't even gotten to whether P(NR) or P(R) make any sense.

Humots,

- I seem to get more confused with each exchange.
- Perhaps, I should be referring to "Bayesian inference," instead of the "Bayes Theorem"...
- The following is what I'm talking about -- but, it doesn't seem to be what you're talking about.

- From http://www.answers.com/topic/bayes-theorem
(mathematics) A theorem stating that the probability of a hypothesis, given the original data and some new data, is proportional to the probability of the hypothesis, given the original data only, and the probability of the new data, given the original data and the hypothesis. Also known as inverse probability principle.

- And, from http://en.wikipedia.org/wiki/Bayesian_inference
In statistics, Bayesian inference is a method of inference in which Bayes' rule is used to update the probability estimate for a hypothesis as additional evidence is learned.

- Does that help?

--- Jabba

 

[shuttlt]

Jabba,

Isn't the problem that the uncertainty associated with the "probability estimate for [your] hypothesis" and the uncertainty associated with the "additional information" are so enormous that your answer is entirely dependent on what numbers you happen to pick with out any particularly solid justification.

 

[Jabba]

Jabba,

Isn't the problem that the uncertainty associated with the "probability estimate for [your] hypothesis" and the uncertainty associated with the "additional information" are so enormous that your answer is entirely dependent on what numbers you happen to pick with out any particularly solid justification.

Shuttit,
- I think that I agree with you except that I think I can justify the numbers I insert. As I understand Bayesian, it does deal with rather subjective probabilities.
- I think that Wollery has been asking me for that justification, and I just haven't really gotten around to it except to refer people to my website. Let me know if you want the link. I'd give it now except that everybody fusses about it when I refer anyone to my own site.
--- Jabba

 

[Squeegee Beckenheim]

- I think that Wollery has been asking me for that justification, and I just haven't really gotten around to it except to refer people to my website. Let me know if you want the link. I'd give it now except that everybody fusses about it when I refer anyone to my own site.
--- Jabba

This is because you should have posted your entire argument in the opening post of this thread. Here it is, 10 days after you started the thread, and you're sill making excuses for why you haven't posted your argument, rather than simply posting your argument.

Now, please, stop faffing around and wasting everybody's time and instead post your entire argument, clearly and concisely. In this thread. Not anywhere else. Here.

 

[Humots]

Humots,

- I seem to get more confused with each exchange.
- Perhaps, I should be referring to "Bayesian inference," instead of the "Bayes Theorem"...
- The following is what I'm talking about -- but, it doesn't seem to be what you're talking about.

- From http://www.answers.com/topic/bayes-theorem
(mathematics) A theorem stating that the probability of a hypothesis, given the original data and some new data, is proportional to the probability of the hypothesis, given the original data only, and the probability of the new data, given the original data and the hypothesis. Also known as inverse probability principle.

- And, from http://en.wikipedia.org/wiki/Bayesian_inference
In statistics, Bayesian inference is a method of inference in which Bayes' rule is used to update the probability estimate for a hypothesis as additional evidence is learned.

- Does that help?

--- Jabba

Yes, in that it shows that you still don't understand what you are talking about.

The two statements you reference are essentially saying the same thing in different ways: (1) that Bayes' Theorem (or rule) can be used to create an equation giving the probability of a hypothesis based on current evidence and (2) that this equation can be updated to include new probability values for existing evidence and new evidence as it appears.

 

[Giordano]

Jabba,

Here's a clue for you: What is the probability of me drawing the ace of spades from a deck of cards? And, separate question, after having drawn the ace of spades, what is the probability of the card I drew being the ace of spades? The probability of your existence is the latter.

Here is another hint: the Earth is the right distance from the Sun for human life: much further or closer we could not survive. Creationists therefore conclude this is an incredibly unlikely event that must be evidence of God's design. Can you see the flaws in this argument?

 

[tsig]

Shuttit,
- I think that I agree with you except that I think I can justify the numbers I insert. As I understand Bayesian, it does deal with rather subjective probabilities.
- I think that Wollery has been asking me for that justification, and I just haven't really gotten around to it except to refer people to my website. Let me know if you want the link. I'd give it now except that everybody fusses about it when I refer anyone to my own site.
--- Jabba

We know how conception takes place and how the embryo grows so when does this immortality thing get implanted?

 

[abaddon]

- I seem to get more confused with each exchange.

Urge to go for obvious repost rising...rising...

- Perhaps, I should be referring to "Bayesian inference," instead of the "Bayes Theorem"...

Make up your mind mister statistician.

- The following is what I'm talking about -- but, it doesn't seem to be what you're talking about.

Urge to go for obvious repost rising further.

- From http://www.answers.com/topic/bayes-theorem
(mathematics) A theorem stating that the probability of a hypothesis, given the original data and some new data, is proportional to the probability of the hypothesis, given the original data only, and the probability of the new data, given the original data and the hypothesis. Also known as inverse probability principle.

answers.com Mr. statistician?

- And, from http://en.wikipedia.org/wiki/Bayesian_inference
In statistics, Bayesian inference is a method of inference in which Bayes' rule is used to update the probability estimate for a hypothesis as additional evidence is learned.

Then a follow up from Wiki? No pro statistician you.

- Does that help?

--- Jabba

Nope.

Shuttit,
- I think that I agree with you except that I think I can justify the numbers I insert. As I understand Bayesian, it does deal with rather subjective probabilities.

Yet you won't justify them. Why?

- I think that Wollery has been asking me for that justification, and I just haven't really gotten around to it except to refer people to my website. Let me know if you want the link. I'd give it now except that everybody fusses about it when I refer anyone to my own site.
--- Jabba

Ya think?

 

[jt512]

Jabba: let's look at the math.

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

[...]

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)

Both of those equations are wrong. If you're going to explicitly include the background information k in Bayes Theorem, then it has to appear in every probability term and always behind the bar '|', like this:

Jay

 

[dafydd]

What does all this have to do with immortality?

 

[abaddon]

What does all this have to do with immortality?

Well, you see....OK I give up.

 

[AdMan]

It's increasingly apparent Jabba is way over his head, and has little actual understanding of what he thinks he is talking about.

 

[Pixel42]

Here is another hint: the Earth is the right distance from the Sun for human life: much further or closer we could not survive. Creationists therefore conclude this is an incredibly unlikely event that must be evidence of God's design. Can you see the flaws in this argument?

He hasn't responded to previous attempts to explain the fundamental mistake he is making using Douglas Adams' puddle analogy, so I'm betting he'll ignore this one as well.

ISTR there's a wizard in one of Terry Pratchett's Discworld books who, starting with the observation of the amazing coincidence that the sun rises in the morning at just the right time for people to wake up and go to work, ends up "proving" that the entire universe was created in order to bring about his own existence.

 

[Squeegee Beckenheim]

It's increasingly apparent Jabba is way over his head, and has little actual understanding of what he thinks he is talking about.

So what else is new?