Is ESP More Probable Than Advanced Alien Life?

[jaydeehess]

Let's say the bachelor goes into an air balloon with his fiancée and some official and they get married in the air.
Now they land. Although the man is still the same person, he no longer is a bachelor, he is a married man and as such he lands.

Him being the same person doesn't matter in your scenario.

This is a straight forward, almost trivially known mechanism by which a bachelor becomes a married man.
In the case of a two headed coin becoming a two tailed coin the mechanism is much more complex and includes several disparate scenarios (trickery versus rearranging of molecules).
Thus the former, changing a bachelor into a married man is more probable than a two headed coin becoming a two tailed coin.
Imho of course.

 

[jsfisher]

* http://fitelson.org/probability/eells_bpooe.pdf

That looks like a treatise on Philosophy and not Statistics.

 

[Daylightstar]

This is a straight forward, almost trivially known mechanism by which a bachelor becomes a married man.
In the case of a two headed coin becoming a two tailed coin the mechanism is much more complex and includes several disparate scenarios (trickery versus rearranging of molecules).
Thus the former, changing a bachelor into a married man is more probable than a two headed coin becoming a two tailed coin.
Imho of course.

There are much, much more elaborate mechanisms for a bachelor to become a married man.
But I think you're right, the current bachelor example compared to quantum molecule atom zappers has by far the highest probability.

Even the reason for any one of the more elaborate mechanisms for the bachelor to become a married man can understand this after about 7 or 8 years

 

[Fudbucker]

That looks like a treatise on Philosophy and not Statistics.

"In probability theory and statistics, Bayes' theorem (alternatively Bayes' law or Bayes' rule) relates current probability to prior probability. It is important in the mathematical manipulation of conditional probabilities."


From the link at http://fitelson.org/probability/eells_bpooe.pdf

"Bayesian Problems of Old Evidence"

ETA: Relevant part:

"What I shall call (henceforth) simply "The Problem Of Old Evidence" arises
in cases in which E is learned first, and [H] is formulated subsequent to the learning
of E."

We knew life on earth existed long before we began to seriously wonder when alien life existed. Hence, you have the problem of old evidence.

 

[Fudbucker]

This is a straight forward, almost trivially known mechanism by which a bachelor becomes a married man.
In the case of a two headed coin becoming a two tailed coin the mechanism is much more complex and includes several disparate scenarios (trickery versus rearranging of molecules).
Thus the former, changing a bachelor into a married man is more probable than a two headed coin becoming a two tailed coin.
Imho of course.

Agreed, but it was asserted that it was logically impossible (not just highly improbable) for a two-headed coin to land tails. That's obviously not true. There are many far-fetched scenarios for how a two-headed coin could land tails. I wouldn't bet on any of them happening, but they are still possible.

 

[Slowvehicle]

Agreed, but it was asserted that it was logically impossible (not just highly improbable) for a two-headed coin to land tails. That's obviously not true. There are many far-fetched scenarios for how a two-headed coin could land tails. I wouldn't bet on any of them happening, but they are still possible.

Fud, you are still, simply, wrong about this.

Do try to follow:

A coin lands on the ground, with one face showing "tails". The only way for that to happen is if at least one of the faces is, in fact, a "tails". If at least one (out of a possible two) faces is a "tails", the coin is not, cannot be, a "two-headed coin". It is, in the inestimable turn of phrase of a poster on this thread, "...something entirely else..." than a coin-with-two-heads, as demonstrated by the observation that at least one of its faces (out of a possible two) is a "tails".

...you're welcome.

 

[jsfisher]

"In probability theory and statistics, Bayes' theorem (alternatively Bayes' law or Bayes' rule) relates current probability to prior probability. It is important in the mathematical manipulation of conditional probabilities."


From the link at http://fitelson.org/probability/eells_bpooe.pdf

"Bayesian Problems of Old Evidence"

Had you bothered to look up, you might have seen the point I was making as it whooshed over your head.

Nonetheless, I repeat: The treatise looks like one of Philosophy. Its reference to Bayes' theorem no more makes it a study of Statistics than its reference to evidence places it in a court of law.

 

[jaydeehess]

Agreed, but it was asserted that it was logically impossible (not just highly improbable) for a two-headed coin to land tails. That's obviously not true. There are many far-fetched scenarios for how a two-headed coin could land tails. I wouldn't bet on any of them happening, but they are still possible.

Yet getting back to the OP, any discussion of the impossible or possible is irrelevant since those two conditions are not the subject. Instead its relative probability.
Advanced life has a known occurrence and thus although complex, it has a demonstrated mechanism. ESP otoh , while one could speculate on all manner of far fetched mechanisms , there has never been one demonstrated to have taken place.

 

[The Norseman]

I did one for advanced alien life. The hypothesis (H) is that "alien life exists". E is our evidence that there's life on Earth. Pr(E/H) is either through the roof (i.e., not surprising at all, and therefore not confirming) or, if looked at counterfactually, a number can't be assigned because we don't know the necessary conditions for which life is possible, and therefore don't know how surprising it would be for life to occur on Earth if we were looking at Earth 4 billion years ago, without the knowledge that life would eventually occur, and wondering how surprising it would be for life to occur. (counterfactuals get a little tricky, but you have to do something like that if the evidence is already known to exist (problem of old evidence*)).

More devastating to the calculus is that Pr(E) is either so high that no confirmation occurs, or, again, counterfactually impossible to assign.

So you have a case where there's either no confirmation, because the evidence is already known to exist, or unknown confirmation, because the surprisingness of the evidence can't be determined.

I didn't get a chance to do a Bayesian ESP calc. No one would agree my Bayesian one on alien life was correct (although it is).

* http://fitelson.org/probability/eells_bpooe.pdf

*"The problem first emerges when considering cases in which pr(E) is taken to have the extreme value of 1. In those circumstances, E is treated probabilistically just as if it were a tautology. It cannot confirm anything, since pr(H/E) will be equal to pr(H). Once one is absolutely certain about one’s evidence, it no longer is evidence

...

The most immediately appealing, and most widely criticized, direct approach to the synchronic problem relies on counterfactual degrees of belief. For an agent who knows E, this approach would analyze confirmation with the standard positive relevance definition; however, it would be applied not to the agent’s actual probabilities, but to what those probabilities would have
been in circumstances where the agent did not know E.

http://www.brown.edu/Departments/Philosophy/onlinepapers/christensen/MeasuringConfirmation.pdf

Then show your work with a proper Baye's Theorem when applied to ESP and advanced alien life, without priors. It should be simple enough to show that the theories of ESP and AAL have an approximately equal probability. It's not necessary to go into heavy detail, just put your numbers down and let's all go from there.

You said that both ESP and advanced alien life have the same probability. How then did you arrive at this statement, other than an erroneous application of the null hypothesis?

 

[Fudbucker]

Fud, you are still, simply, wrong about this.

Do try to follow:

A coin lands on the ground, with one face showing "tails". The only way for that to happen is if at least one of the faces is, in fact, a "tails". If at least one (out of a possible two) faces is a "tails", the coin is not, cannot be, a "two-headed coin". It is, in the inestimable turn of phrase of a poster on this thread, "...something entirely else..." than a coin-with-two-heads, as demonstrated by the observation that at least one of its faces (out of a possible two) is a "tails".

...you're welcome.

This is true, however, you're still not getting it. The claim isn't that the two-tailed coin IS a two-headed coin. The claim is that the two-tailed coin WAS a two-tailed coin, prior to the toss. Just like the married man ISN'T a bachelor when he lands, but he WAS a bachelor before he landed.

If a bachelor can land a married man, then a two-headed coin can land tails. There's no way around it. It is impossible for a bachelor to BE a married man, it's not impossible for a bachelor to LAND a married man. It is impossible for a two-headed coin to BE a two-tailed coin. It is not impossible for a two-headed coin to LAND a two-tailed coin.

For any X, before X lands, X can turn into Y.

 

[jaydeehess]

http://www.nasa.gov/ames/kepler/nas...planet-in-the-habitable-zone-of-another-star/
The search gets closer.
How's it going with ESP research?

 

[jsfisher]

That looks like a treatise on Philosophy and not Statistics.

A follow-up:

Ellery Ells, the author of the paper, was part of the Philosophy department faculty at the University of Wisconsin until his untimely death in 2006. Branden Fitelson created the link Fudbucker provided for students in his course, Philosophy 148: Probability and Induction, in 2008. Dr. Fitelson is a professor of Philosophy at Rutgers University.

This thread belongs in a different forum.

 

[jaydeehess]

This is true, however, you're still not getting it. The claim isn't that the two-tailed coin IS a two-headed coin. The claim is that the two-tailed coin WAS a two-tailed coin, prior to the toss. Just like the married man ISN'T a bachelor when he lands, but he WAS a bachelor before he landed.

If a bachelor can land a married man, then a two-headed coin can land tails. There's no way around it. It is impossible for a bachelor to BE a married man, it's not impossible for a bachelor to LAND a married man. It is impossible for a two-headed coin to BE a two-tailed coin. It is not impossible for a two-headed coin to LAND a two-tailed coin.

For any X, before X lands, X can turn into Y.

Which again, says nothing at all about the relative probability of the two examples.

 

[jaydeehess]

A follow-up:

Ellery Ells, the author of the paper, was part of the Philosophy department faculty at the University of Wisconsin until his untimely death in 2006. Branden Fitelson created the link Fudbucker provided for students in his course, Philosophy 148: Probability and Induction, in 2008. Dr. Fitelson is a professor of Philosophy at Rutgers University.

This thread belongs in a different forum.

I accept some responsibility for it being here. Fudbucker suggested the Philosophy section, a suggested it stay in math & science.

 

[Fudbucker]

Had you bothered to look up, you might have seen the point I was making as it whooshed over your head.

Nonetheless, I repeat: The treatise looks like one of Philosophy. Its reference to Bayes' theorem no more makes it a study of Statistics than its reference to evidence places it in a court of law.

JS, the article is about the problem of old evidence, which is a problem for a Bayesian analysis where life on Earth is being used as evidence for the existence of alien life. It's a problem for any statistical analysis of the confirmation of a piece of evidence when the evidence is already known to exist.

I'm not forcing you to read the article. It's a reference for anyone interested in the problem of old evidence.

 

[Fudbucker]

​

You said that both ESP and advanced alien life have the same probability. How then did you arrive at this statement, other than an erroneous application of the null hypothesis?

Do you agree we can't assign a probability to the existence of alien life?

 

[Fudbucker]

I accept some responsibility for it being here. Fudbucker suggested the Philosophy section, a suggested it stay in math & science.

Well, I agreed, but I thought these problems would show up.

Still, I doubt we have many professional philosophers posting in the philosophy forum, so the outcome probably would have been the same.

 

[jsfisher]

JS

I'd prefer "jsfisher".

the article is about the problem of old evidence, which is a problem for a Bayesian analysis where life on Earth is being used as evidence for the existence of alien life. It's a problem for any statistical analysis of the confirmation of a piece of evidence when the evidence is already known to exist.

It is a philosophical argument, not one of statistics.

 

[Fudbucker]

Which again, says nothing at all about the relative probability of the two examples.

Nor was it meant to. Just that both are logically possible (i.e., both have a non-zero chance of happening).

 

[jaydeehess]

Do you agree we can't assign a probability to the existence of alien life?

That is not required for the OP, instead it concerns relative probability.