Existence of God is odds on at 2-1

[Marquis de Carabas]

Always round to the nearest number whose factors' sum is itself. It's the best way to maintain a consistent record of wild inaccuracy and still call yourself perfect.

 

[Iacchus]

I'm kinda partial to rounding to the nearest multiple of 37, myself.

Yes, 37 x 18 = 666 ...

 

[Chemical_Penguin]

Yea I doubt he really trusts his own work. I've seen this kind of thing before when I read the book "The Physics of Immortality" where the author, Frank J. Tipler, used "modern day"(1993) cosmology to prove that a Supreme Being would come into existence in the far future. Too bad that after all his evidence he ends the book with a chapter where he admits that he himself does not believe all of the goofiness he just sold to you, the reader.

And as a double whammy his theory from 1993 was based on a universe that is collapsing, too bad we now know it's expanding.

So I don't think this 2/3's business is the least bit credible because it's using all the information we have NOW. And there's plenty of info about reality that we lack.

 

[kuroyume0161]

Yea I doubt he really trusts his own work. I've seen this kind of thing before when I read the book "The Physics of Immortality" where the author, Frank J. Tipler, used "modern day"(1993) cosmology to prove that a Supreme Being would come into existence in the far future. Too bad that after all his evidence he ends the book with a chapter where he admits that he himself does not believe all of the goofiness he just sold to you, the reader.

And as a double whammy his theory from 1993 was based on a universe that is collapsing, too bad we now know it's expanding.

So I don't think this 2/3's business is the least bit credible because it's using all the information we have NOW. And there's plenty of info about reality that we lack.

You paid for that also!? About third way through the book, suspiscious voices in my head kept nagging me about those mounting suppositions based upon suppositions (ad infinitum, heheh). Post treatment ha helped.

Been so long that I can't remember when I spent most of the read laughing and writhing in contention, but at some point the speculations became too dependent upon far reaching possibilities. As with any predictive situation involving complex systems (in this case, the entire universe and everything), chaotic uncertainty quickly overtakes any significant probability.

Kuroyume

 

[Chemical_Penguin]

lmao well I'm glad I'm not the only one )

 

[ceo_esq]

Who says that "good" is one of God's defining attributes? Where does he have evidence of this?

That's part of his definition of God for the purposes of this proof, I gather. Evidence is typically required for the premises and conclusions of proofs, but not for the definitions - right? He can define what he's trying to prove in any way he likes; strictly speaking, it shouldn't affect the validity of the proof.

 

[Silicon]

Except, CEO, he then goes straight to the fact that we know the difference between good and evil as evidence that god exists.

Circular reasoning.

 

[daenku32]

That's part of his definition of God for the purposes of this proof, I gather. Evidence is typically required for the premises and conclusions of proofs, but not for the definitions - right? He can define what he's trying to prove in any way he likes; strictly speaking, it shouldn't affect the validity of the proof.

Since adding definitions to an object increases it's specificness, doesn't that mean that in his calculation non-specific neutral and indifferent god would have 100% chance of existing?

That seems like a leap of faith to me.

The 50/50 assumption might only apply to an totally non-specific god being. The moment you start applying any attributes to it, the chances go down.

So when you get down to the Christian God with thousands of claimed attributes, you are talking about quite tiny odds.

 

[ceo_esq]

Except, CEO, he then goes straight to the fact that we know the difference between good and evil as evidence that god exists.

Circular reasoning.

I'm not sure that that's circular. Let's say that X is a hypothetical entity whose attributes include (1) being fundamentally good and (2) having created humankind in the image/likeness of X. Let H be the hypothesis X exists. Let O be the observation that there are creatures in the world capable, in principle, of distinguishing good from evil.

Not having read the book, my understanding is that at least one facet of Bayes' Theorem dictates that the probability one would assign to H being true after making observation O can be expressed as a fraction where the numerator is the product of the the probability one would assign to H before making observation O (the "prior probability") and the probability of O being true if H is true, and the denominator is the numerator plus the product of the prior probability of H not being true and the probability that O would be true even if H were untrue.

To condense all this, I think the author runs the foregoing calculation, and then runs it again with other different inputs for O, and concludes in each case that the truth of O actually marginally boosts the probability of H being true. I don't think this argument is circular in the self-referential sense you're thinking of it. As far as I can tell (and I dropped math after freshman-year calculus, so I could well be wrong), this approach - or something similar to it - is essentially how statisticians calculate a lot of conditional probabilities.

 

[kuroyume0161]

I'm not sure that that's circular. Let's say that X is a hypothetical entity whose attributes include (1) being fundamentally good and (2) having created humankind in the image/likeness of X. Let H be the hypothesis X exists. Let O be the observation that there are creatures in the world capable, in principle, of distinguishing good from evil.

I like your logical reasoning here, but, just like Pascal's Wager, the alternatives to the hypothesis do not seem to be introduced (that differentiating good and evil could be an inherited trait caused by evolutionary processes) to balance the calculation. Although humans can vocalize their understanding of these concepts, it is quite assured that many higher mammals also have a concept of 'right and wrong' (from which 'good and evil' follow, I think), at least in a survival context.

Anyway, 'good and evil' are still subjective, relative terms and there is much gray inbetween their dichotomic inference. Labeling things and events as such is usually a matter of culture, tradition, past experience, and future results.

Kuroyume

 

[ceo_esq]

I like your logical reasoning here, but, just like Pascal's Wager, the alternatives to the hypothesis do not seem to be introduced (that differentiating good and evil could be an inherited trait caused by evolutionary processes) to balance the calculation.

Well, I can't vouch for whatever numbers the author assigns to such things, but using Bayes' Theorem certainly would require that some prior probability be estimated for the likelihood that awareness of good and evil would develop fortuitously in a world without God, because that's one element of the equation (it's the part I referred to as "the probability that O would be true even if H were untrue"). I agree with you that there is certainly some probability that O would be true even if H were untrue, but I agree with the author (or at least with what I understand to be his position) that such probability, even if high, is lower than the (very high) prior probability of O being true if H is, in fact, true.

 

[jlakbj]

No one else was concerned that this guy thought that without God humans would only have a 10% chance of being able to differentiate good from evil? What a misanthrope!

 

[metacristi]

The impossibility,in many practical cases,to calculate the apriori probabilities (of being true) for the competing hypotheses under investigation is one of the main arguments used by the supporters of the objective interpretation of probabilities to reject the subjective bayesian interpretation.Still,as practice proved,it's fair to set them to 50/50 initially,the results of this assignation has proved fruitful in many practical cases.I'd that this is justified even in the case studied by Unwin since science neither deny nor confirm the existence of such a God.I don't think this is a crucial problem,the main difficulty in Unwin's approach is to establish what deserve to be considered new evidence pro/con God and especially the lack of agreement in the evaluation of this new data.The same facts can be interpreted in totally different ways,for example there where one sees a sign of God's goodness another might be totally unimpressed,considering it as a normal fact of life;the interpretation of the same new data leads to totally different conclusions.This is why the bayesian interpretation of probabilities does not lead to conclusion which to be accepted by all would be rational people (of course with the apriori assumption of the validity of the bayesian perspective).

We have to agree in the interpretation of the new evidence to obtain some meaningful conclusions valid for all would be rational people.I think an example would be suggestive here.Suppose that you play dice with a friend you know is not reliable (each of you have your own die when throwing) because he used previously in many occasions (other possibilities being negligleable I will exclude them conventionally in this example) an unfair die which had 6 on all sides (assume for the sake of this example that you can see only the upper part of your friend's die when it settles down).What subjective probabilities can you assign for the competing hypotheses that [she is cheating again using the die having 6 on all sides/she is fair play this time] based on the result of a throw?

Let

H1=the hypothesis that his die is fair
H2=the hypothesis that his die is unfair having '6' on all sides
E=the new evidence,new data=the results of his throw with the die.

If the results of his throw is different from 6 then you have almost certainty that H1 is true (if other possibilities are excluded H1 is true-this is the case I treat here) and certainly H2 is false.But let's suppose that a 6 appeared.

A form of Bayes' formula usually used in such cases (only 2 competing hypotheses) is:

A[E]=P[H1/E]/P[H2/E]={P[E/H1]*P[H1]}/{P[E/H2]*P[H2]} (1) where:

A[E]=the Bayes' factor which measures the amount by which the new data favors H1 over H2.
P[H1/E]=the probability of H1 occuring conditioned by the apparition of E.
P[H2/E]=the probability of H2 occuring conditioned by the apparition of E.
P[E/H1]=the probability of appearance of the new evidence E when H1 is true.
P[E/H2]=the probability of appearance of the new evidence E when H2 is true.
P[H1]=the apriori probability of H1 being true,set at 50%=1/2 conventionally.
P[H2]=the apriori probability of H2 being true,also set at 50%=1/2 conventionally.

Now we have:

P[E/H1]=1/6 (for the fair die the probability to appear a '6' is 1/6)
P[E/H2]=1 (for the die having 6 on all sides the probability to appear '6' is of course 1)

Plugging these in (1) --->

A[E]={(1/6)*(1/2)}/{1*(1/2)}=1/6 (2)

Thus this means that the hypothesis H2 is six times more probable than H1 in the light of the new data.All rational people will calculate the same subjective probabilities once they accept the bayesian approach as a valid 'tool'.If a new 6 appears H2 is 36 times more probable than H1 so that the result of the subjective evaluation converge very rapidly toward the conclusion that he cheats again,H2 being disproved [of course we never reach certitudes if we do not check his die].Using the usual objectivist interpretation of probabilities even if there appears a string of 6 at the throws,with the probability of appearance of (1/6)^n,n the number of appearance of a 6,the data is still compatible with both hypotheses so there are no sufficient reasons to believe that H2 was disproved,if n is enough small.In practice the subjecitve approach proved to be much more sensible in many occasions even when evaluating the results of experiments studying paranormal phenomena.