[Merged] Immortality & Bayesian Statistics

Status
Not open for further replies.
Jabba said:
Dave,
- I'll see if I can find an official statement, but sure think we're expected to "speculate" in Bayesian statistics.
Click to expand...
Dave,
- Doing a quick search, I found the following in Wikipedia: "Bayesian statistics is a subset of the field of statistics in which the evidence about the true state of the world is expressed in terms of degrees of belief or, more specifically, Bayesian probabilities." To me, "belief" implies speculation. If it doesn't imply speculation to you, let me know, and I'll see if I can find something more substantial.
 
Jabba said:
Dave,
- Doing a quick search, I found the following in Wikipedia: "Bayesian statistics is a subset of the field of statistics in which the evidence about the true state of the world is expressed in terms of degrees of belief or, more specifically, Bayesian probabilities." To me, "belief" implies speculation. If it doesn't imply speculation to you, let me know, and I'll see if I can find something more substantial.
Click to expand...

Yeah it's just using judgement, making estimates and applying them to a formula. It's how the world is run.
 
Toontown said:
Every freshman in my genetics class used the test repeatedly. They didn't seem to have any problem with the concept of expected frequencies,
Click to expand...
i am not sure how to interpret this as evidence (for or against :)) but reading the link you kindly provided http://www.colby.edu/biology/BI17x/freq.html we find:
In this example, our null hypothesis is that the coin should be equally likely to land head-up or tails-up every time. The null hypothesis allows us to state expected frequencies. For 200 tosses, we would expect 100 heads and 100 tails.
Click to expand...
this is a clear definition of the null, but i would not say "we would expect 100 heads and 100 tails.", in fact i would bet you this will not be the result. and i'd offer 2:1 odds. care to take that wager?
http://www.colby.edu/biology/BI17x/freq.html Using probability theory, statisticians have devised a way to determine if a frequency distribution differs from the expected distribution. To use this chi-square test, we first have to calculate chi-squared.
Click to expand...
technically this is already askew, as we never "determine if a frequency distribution differs from the expected distribution". firstly the "expected" was a number and is now said to be a distribution, secondly the best one can ever do it look for consistency, or the lack there of: the probability of some previously specified statistics given the particular obs and the stated null. (admittedly the first might be a pedagogical nitpick, but the second is a fairly fundamental violation of the lessons of Statistics 101.)
http://www.colby.edu/biology/BI17x/freq.html Because the chi-squared value we obtained in the coin example is greater than 0.05 (0.27 to be precise), we accept the null hypothesis as true and conclude that our coin is fair.
Click to expand...
one would never ever "accept the null hypothesis as true". neither frequentist nor bayesian.

one either rejects the null, or one fails to reject the null. would you really believe a coin was fair after three flips?

if you know who wrote this page, and s\he would like to discuss improving it with a statistician, i can happily arrange for that to happen if you send me their email via PM.

Toontown said:
I'm simply using the chi-square test to prove the validity of the concept of expected frequency (which many people in these kinds of threads are invariably in ignorance and/or denial about), and to provide an example of how expected frequencies may be compared to observed results to test hypotheses.
Click to expand...
OK: we agree that a given null will assign probabilities to all outcomes allowed by the null. here we agree.


Toontown said:
I'm simply pointing out that the expected frequency of an observation does not become void after the observation.
Click to expand...
that is true: but the expected frequency (sic) is a property of the null, not a property of the system being studied. it is P(X=x|null is true)

and the probability of the observations before it is made... well as i do not believe the null is true i do not know how to estimate that... once i do, then that becomes my prior on that observation.

and the probability of a observation after the observation has been taken? well P(X=x | X=x) = 1
Toontown said:
Not if the hypothesis from which the expected frequency was derived is true.
Click to expand...
it does not matter if the null is true or false. P(X=x|null is true) remains the same...

Toontown said:
The hypothesis determines the value of the expected frequency. The expected frequency does not change depending on when it is calculated, or what is observed. The expected frequency is fixed by the hypothesis.
Click to expand...

agreed.

Toontown said:
The only thing that changes is your belief in the hypothesis, if the observations fail to align sufficiently with the calculated expectations.
Click to expand...
well, not quite the only thin: what also changes is that the observation now has been made, and the probability of the observation being what is goes to one.

so if the observation is "you exist", and you are making an argument about your existence, then the probability you exist is equal to one. agreed?

this is a special case where in order to make an argument about your existence, you are already in the case Prob(you exist) = 1.
 
Toontown said:
No issue. You now have your expected frequency, under the unique brain assumption.

The expected frequency is identical to the probability of chance.

The probability of chance is the probability that the unique brain assumption accounts for your current sentient experience, because that is the probability that the presumably required unique brain would occur by chance, and your brain is that specific prerequisite brain.
Click to expand...

You're mixing up two different probabilities. The probability of mine, or anyone else's, brain existing under the unique brain hypothesis is not the same as that hypothesis being true.

The comet ISON wouldn't exist if not for a specific sequence of unlikely events. Does that mean the naturalistic explanation for the existence of comets is unlikely?

Toontown said:
The probability of chance is sufficiently small to reject the unique brain hypothesis with very high confidence.
Click to expand...

Why? Why would you reject a hypothesis when the observed results equal the results predicted by that hypothesis?
 
Jabba said:
Dave,
- Doing a quick search, I found the following in Wikipedia: "Bayesian statistics is a subset of the field of statistics in which the evidence about the true state of the world is expressed in terms of degrees of belief or, more specifically, Bayesian probabilities." To me, "belief" implies speculation. If it doesn't imply speculation to you, let me know, and I'll see if I can find something more substantial.
Click to expand...

I don't think that means you can just make up numbers with no justification and plug them into a calculation.
 
Jabba said:
Dave,
- Doing a quick search, I found the following in Wikipedia: "Bayesian statistics is a subset of the field of statistics in which the evidence about the true state of the world is expressed in terms of degrees of belief or, more specifically, Bayesian probabilities." To me, "belief" implies speculation. If it doesn't imply speculation to you, let me know, and I'll see if I can find something more substantial.
Click to expand...

A proof of the existence of immortality would come in handy. Got any?
 
godless dave said:
You're mixing up two different probabilities. The probability of mine, or anyone else's, brain existing under the unique brain hypothesis is not the same as that hypothesis being true.
Click to expand...

Actually, it is, in this case.

A couple of things to keep in mind:

1. You exist.

2. There is a corollary to the unique brain hypothesis: experiencable sentience does not require a specific brain. Just some brain that happens tautologically to be 'you'.

Thus, the probability of chance, in this case, is the probability that the specific brain mysteriously required by the unique brain hypothesis exists. Because you do exist either way, and if your brain is not the one and only required by the hypothesis, then the unique brain hypothesis has failed.

Yeah, I know it's counterintuitive. How could just any random brain be 'you'? Well, the way the one that is you is you. No particular reason. It just tautologically is.

Actually, it's no more counterintuitive than the alternative, which is that, for some inconceivable, totally tautological reason, there was only one brain that could ever be 'you', and that one just happens to exist against inconceivably long odds.

godless dave said:
Why? Why would you reject a hypothesis when the observed results equal the results predicted by that hypothesis?
Click to expand...

The result predicted by the hypothesis is nothing forever, with a certainty approaching 1. That is not the observed result. In this case, the observed result is so unlikely under the hypothesis that the hypothesis fails the test spectacularly.
 
Last edited:
godless dave said:
I
Yes. My question is, why are you equating 0.00000........1 with impossibility?
Click to expand...


My question is, why are you arguing about conditional probabilities that depend on something as lame and unscientific as this so-called "unique brain hypothesis"?
 
Toontown said:
http://www.internationalskeptics.com/forums/showthread.php?postid=9623828#post9623828
Click to expand...

This?
Toontown said:
"You" are a one-off, because your existence is entirely dependent upon a unique brain occurring at unique spacetime coordinates, as an extremely indirect result of the chaotic quantum shuffle shortly after t=0+10-43
"You" is roughly defined as a sentient experience which is actually experienced, rather than remote sentient experiences occurring in other brains, which are not experienced, but may be assumed to be experiencing themselves.
Click to expand...
What's the actual hypothesis? It can't simply be that we're all different, because that seems trivially true. And how does it relate to the topic of this thread? Where does immortality come in?

The corollary of the hypothesis is that experiencable sentience is not dependent upon the existence of one unique brain.
Click to expand...
Er, what? More than one person is sentient? Again, we're in Sybil Fawlty territory.

Whenever someone is lucky enough to exist, skeptical enough to be suspicious, and decides to question the hypothesis. It's happened a few times that I am aware of, but it's a low probability event.

I'm aware of it having happened in 1933, 2006, and 2013.
Click to expand...
What happened in 1933, 2006 and 2013?
 
Toontown said:
Actually, it is, in this case.

A couple of things to keep in mind:

1. You exist.

2. There is a corollary to the unique brain hypothesis: experiencable sentience does not require a specific brain. Just some brain that happens tautologically to be 'you'.

Thus, the probability of chance, in this case, is the probability that the specific brain mysteriously required by the unique brain hypothesis exists. Because you do exist either way, and if your brain is not the one and only required by the hypothesis, then the unique brain hypothesis has failed.
Click to expand...

That's pretty much a word salad. "Probability of chance" is redundant.

Toontown said:
Actually, it's no more counterintuitive than the alternative, which is that, for some inconceivable, totally tautological reason, there was only one brain that could ever be 'you', and that one just happens to exist against inconceivably long odds.
Click to expand...

It is not inconceivable that only one brain could ever be me, any more that it is inconceivable that only one planet can be Saturn. Nor is my existence against inconceivably long odds. Long odds, yes. Inconceivably long, no.


Toontown said:
The result predicted by the hypothesis is nothing forever, with a certainty approaching 1.
Click to expand...

This is flat out untrue.

If someone buys a lottery ticket, and it wins, does that mean the lottery is rigged? I mean, it's pretty long odds that a ticket would win.
 
godless dave said:
I don't think that means you can just make up numbers with no justification and plug them into a calculation.
Click to expand...
Dave,
- I've plugged numbers into four parts of the equation: 1) the likelihood of me, given the scientific model; 2) the prior probability of the scientific model; 3) the prior probability of the complementary model; 4) the likelihood of me, given the complementary model. I provided what I think is justification for each of them in post #1172 on page 30. Please let me know what isn't clear or isn't convincing.
 
Status
Not open for further replies.

Back
Top Bottom