Cont: Proof of Immortality VIII

[Jabba]

This thread continues from Part VII, found here.

Posted By: Loss Leader

That doesn't make sense. The coin may be fair, two-headed or two-tailed, and the probabilities of those three outcomes must sum to 1. Yours sum to 1.5. Show your working, and I'll show you where you've gone wrong.

Dave

Dave,
- I'm not talking about probabilities; I'm talking about likelihoods.
- P(heads|fair) and P(heads|2 headed). P, here, refers to ikelihoods.

 

[Belz...]

Dave,
- I'm not talking about probabilities; I'm talking about likelihoods.

Stop using that as a way to avoid answering other posters. It's very rude, and I have it on good authority that we should avoid rude posters.

 

[Jabba]

I'd like to see you demonstrate that, but you won't.



I'd like to see you demonstrate that, but you won't.

Belz,
- If your using a fair coin, what is the probability that it will come up heads? That's the meaning of "likelihood."

 

[Dave Rogers]

Dave,
- I'm not talking about probabilities; I'm talking about likelihoods.
- P(heads|fair) and P(heads|2 headed). P, here, refers to ikelihoods.

Neither of these is the likelihood that the coin is fair or that it's two-headed. They are conditional probabilities that the coin toss results in heads in either instance. And, call it what you want, likelihood or probability; there is no valid calculation that comes up with the result that, if a single coin is flipped once, it is both 50% likely to be fair and 100% likely to be two-headed. The very suggestion is nonsensical. Are you confusing P(heads|fair) with P(fair|heads), and P(heads|2 headed) with P(2 headed|heads)?

Dave

ETA: Just to remind you:

Mojo,
- Say that we do have a penny -- i.e., the probability that this penny exists is 1. We flip the penny and it comes up heads. What is the Bayesian likelihood that this penny is two headed?

Dave,
- Yes we do.
- The likelihood of a fair coin is .5 -- the likelihood of a 2 headed coin is 1.

You're confusing the likelihood that a fair coin comes up heads with the likelihood that a coin that comes up heads is fair. If you don't understand the difference between those, then you haven't grasped the fundamentals of conditional probability.

 

[Belz...]

Belz,
- If your using a fair coin, what is the probability that it will come up heads? That's the meaning of "likelihood."

That's a semantics point, and you just did exactly what I accused you of doing, which is rude, and I have it on good authority that we should avoid rude posters.

 

[JesseCuster]

Dave,
- Yes we do.
- The likelihood of a fair coin is .5 -- the likelihood of a 2 headed coin is 1.

There's a 50% chance the coin is fair while simultaneously there's a 100% chance the coin has 2 heads?

I don't think you've thought this through at all and/or your grasp of the basics of probability is wonky.

 

[JayUtah]

That's the meaning of "likelihood."

So far you haven't demonstrated a working understanding of likelihood. You've just quoted from textbooks and ignored all follow-up questions and requests to demonstrate proficiency. Most egregiously, you've given us an example and told us what you think the answer should be, but you haven't actually worked the problem to show that you are proficient in the means used to get the answer. It is rude for you to lecture others from a position of purported expertise while being unwilling to demonstrate the expertise when asked.

Please work your sample problem and show your work at each step. Don't just declare what you believe the answer should be.

 

[theprestige]

I've never seen a 2-headed penny. As far as I can tell, they're pretty rare. If I have a penny, it's almost certainly a fair one, regardless of whether I toss it or not, and regardless of what side is up after the toss.

"That coin toss came up heads, so probably the coin is two headed," said nobody ever.

 

[JoeMorgue]

- If there is such a thing as reincarnation, at least most of us would not know that the previous person was "us."

So, and this is the sort of thing an intellectually honest person would address at some point prior to spending 5 years stalling, what exactly is immortality to you Jabba?

If I'm the reincarnation of.... a young Irish man killed in the trenches of WWI during the Battle of the Somme but we don't look the same, have the same names, don't share any personality traits, and I don't have any of his memories then what exactly has been reincarnated?

I noticed the weasel world "most" in there so I'm going to assume some "highly evolved" people such as yourself will be able to remember.

 

[Mojo]

You're confusing the likelihood that a fair coin comes up heads with the likelihood that a coin that comes up heads is fair. If you don't understand the difference between those, then you haven't grasped the fundamentals of conditional probability.

It's a confusion that is at the heart of his 'proof'.

 

[The Sparrow]

So, and this is the sort of thing an intellectually honest person would address at some point prior to spending 5 years stalling, what exactly is immortality to you Jabba?

If I'm the reincarnation of.... a young Irish man killed in the trenches of WWI during the Battle of the Somme but we don't look the same, have the same names, don't share any personality traits, and I don't have any of his memories then what exactly has been reincarnated?

I noticed the weasel world "most" in there so I'm going to assume some "highly evolved" people such as yourself will be able to remember.

You know...its the "observer" part. The watcher. Though, without memories, how would you know it is the same watcher............ZOMG!!!?

 

[JayUtah]

The likelihood of a fair coin is .5 -- the likelihood of a 2 headed coin is 1.

You asked for "Bayesian likelihood," which isn't really a thing. It appears everyone here has assumed you meant to ask what the posterior probability was, the solution of Bayes' theorem for the example. The answers you're giving us suggest that you're thinking only about the likelihood ratio that is a term in Bayes' theorem. There is such a thing as likelihood, obviously, outside the context of Bayes' theorem. And the posterior probabilities for each of two events, given a common event, are directly comparable only if you have normalized the inverted likelihood via Bayes. But the likelihood ratio by itself doesn't mean what you're ascribing to it. Its only use is to scale the prior probabilities of each of those two events, and your example doesn't contain any priors. You're dissecting Bayes in a way that neither produces correct, useful results nor fixes your proof for immortality.

 

[Dave Rogers]

You asked for "Bayesian likelihood," which isn't really a thing. It appears everyone here has assumed you meant to ask what the posterior probability was, the solution of Bayes' theorem for the example. The answers you're giving us suggest that you're thinking only about the likelihood ratio that is a term in Bayes' theorem.

And, Jabba, even if that is what you think you're talking about, you're still getting it wrong; P(heads|2-headed) is not, for any meaning of the terms used, the Bayesian likelihood that a coin that comes up heads is double-headed, which is what you claim it is. You've mixed it up with the likelihood that a double-headed coin comes up heads - unless, that is, you've made some even more bizarre elementary error.

Dave

 

[Dancing David]

This creates an interesting conundrum. This thing you want to be immortal has no aspect that you observe. So in all of the cycling between your existence and your sense of self, never are you hitting on evidence of immortality.

yet another self contradiction to throw on the bonfire!

 

[Monza]

Mojo,
- If there is such a thing as reincarnation, at least most of us would not know that the previous person was "us."

This creates an interesting conundrum. This thing you want to be immortal has no aspect that you observe. So in all of the cycling between your existence and your sense of self, never are you hitting on evidence of immortality.

If the "self" has no memories or personality but the one contained in the brain of whatever body it inhabits, then although the "sense of self" would be the same, it'd have no way to recognise itself! So not only no way to test whether the theory is correct, but also it makes no difference to the individual person.

In a way, I suspect that jabba's sense of self and mine are in many ways identical, like "going 60 mph" is the same for two cars. It's the body that makes the difference.

So, and this is the sort of thing an intellectually honest person would address at some point prior to spending 5 years stalling, what exactly is immortality to you Jabba?

If I'm the reincarnation of.... a young Irish man killed in the trenches of WWI during the Battle of the Somme but we don't look the same, have the same names, don't share any personality traits, and I don't have any of his memories then what exactly has been reincarnated?

I noticed the weasel world "most" in there so I'm going to assume some "highly evolved" people such as yourself will be able to remember.

Jabba,

Please explain what is the difference between being immortal and being mortal? From your comments, I tend to agree with the people I have quoted above. There is no discernible difference between a mortal person and an immortal person. For example, you have previously stated that (despite being immortal) you did not exist in the year 1888. I then asked what is the likelihood that you would exist in the year 2119. You didn't answer this one, but I assume it is pretty small... say virtually zero.

If a mortal body lives for about a hundred years, and an immortal being cannot be distinguished from a mortal body, the likelihood of immortality is the same as that for mortality. In other words, immortality cannot be observed or measured. Is it really any different than the materialist model?

 

[JoeMorgue]

Jabba,

Please explain what is the difference between being immortal and being mortal?

We all know he'll do no such thing.

 

[MRC_Hans]

Belz,
- If your using a fair coin, what is the probability that it will come up heads? That's the meaning of "likelihood."

What has this to do with the discussion at hand?

Hans

 

[JayUtah]

What has this to do with the discussion at hand?

He's trying to figure out a way to make his critics actually look as ignorant as he seems to believe they are. He's posed an an example he believes demonstrates his purportedly superior understanding of the methods he uses in his proof. Whether through imprecise wording or more egregious error, it's not working the way he expected. But the point was to say, "See, I know likelihoods and you guys don't, so I can therefore dismiss all your statistics rebuttals as unlearned and probably wrong."

 

[jt512]

You asked for "Bayesian likelihood," which isn't really a thing. It appears everyone here has assumed you meant to ask what the posterior probability was, the solution of Bayes' theorem for the example. The answers you're giving us suggest that you're thinking only about the likelihood ratio that is a term in Bayes' theorem. There is such a thing as likelihood, obviously, outside the context of Bayes' theorem. And the posterior probabilities for each of two events, given a common event, are directly comparable only if you have normalized the inverted likelihood via Bayes. But the likelihood ratio by itself doesn't mean what you're ascribing to it. Its only use is to scale the prior probabilities of each of those two events, and your example doesn't contain any priors. You're dissecting Bayes in a way that neither produces correct, useful results nor fixes your proof for immortality.

I agree with you that he must have been asking for the likelihoods, not the posterior probabilities.

I have no idea what Jabba was planning to do with the likelihood ratio, but the likelihood ratio by itself is a meaningful quantity in Bayesian inference. The likelihood ratio is sometimes called the "weight of the evidence," and it is independent of the prior odds of the hypotheses. For any prior odds, the likelihood ratio times the prior odds of some hypothesis versus some other hypothesis equals the posterior odds of those hypotheses. So, the likelihood ratio is the amount by which the evidence changes the odds of the hypotheses. In Jabba's example the likelihood ratio was 2 in favor of the 2-headed coin vs a fair coin. So whatever the prior odds were that the coin was 2-headed (vs fair), the posterior odds will be double the prior odds.

 

[JayUtah]

So, the likelihood ratio is the amount by which the evidence changes the odds of the hypotheses.

I'm on board with that. Unless I've grossly misunderstood you, the effect of the likelihood ratio on some other term is the point I was trying to emphasize in the same way you appear to have. I don't disagree with anything you said.