Consider the boolean question 'will the die come out as a six?'.

Obviously there is a 1 in 6 chance it will.

Contrast that to the 'is he guilty?' question, where we start with a 1 in 2 (50%) chance of guilt/innocence.

So boolean questions have an answer distribution space inherent within them - and we have to take that distribution into account when combining with other probabilities - it is not always the same distribution.

Hello, @Devans99?
As (also already) mentioned, your and my belief don't matter; sorry, but we don't get to tell it all what it is, rather we adjust our belief accordingly.
Quantumatics and relativity are established (already posted an example).
You're free to keep re-re-repeting your assertions, call the subject matter experts names, and dismiss established theories with a hand-wave, none of which make your belief so.
If that's the extent of your investigation, then maybe the opening post is sort of not really in good faith? :-/

I'm not criticising relativity, but QM seems immature - a work in progress. And as you say, the two are incompatible anyway. So any theories that combine relativity and QM to give us insights into the early universe are highly speculative (to say the least).

I'd bet my house on causality, can't say the same for quantum cosmology.

ANYONE reading this thread quickly realizes — Frank Apisa

Part of Devans schtick is playing with statistical ideas. But he does it wrong. And he has presented the exact same arguments even years ago on this site and its predecessor site, all refuted multiple times by multiple people, with no apparent effect on Devans. He is essentially intellectually dishonest, which means just plain dishonest.

Typically he says, well, there's a God or there isn't, right there it's 50-50. Throw in some design and some pseudo-whatever, and by yollee, it's almost certain that God exists. Though never a mention of what any of these terms mean.

One big problem with his argument - of the many - is that whatever he can say about the existence of God can with equal validity be said about an uncountable infinity of other non-present things for which there is no evidence. As with, for example, every conceivable size and shading of purple flying hippopotamus. And that's just the flying purple hippopotami. The "odds" on God, then, go to infinitesimal. That is, a number greater then zero but infinitely smaller than the smallest real number greater then zero.

Devans, then, is just a waste of time.

No. If we are considering a question with a 90% / 10% initial distribution of probabilities then we can't just ignore that distribution and start at 50%. — Devans99

That's precisely the point. You cannot just say "Since we do not know if it is this guy, the probabilities are 50/50". And yet that is exactly how you proceeded.

Contrast that to the 'is he guilty?' question, where we start with a 1 in 2 (50%) chance of guilt/innocence. — Devans99

That is fine because "guily" or "not guilty" are a complete set of options. There are uncountably many more options than "God did it", which is a silly non-option. Saying "God did it" is 50/50 is exactly the same as saying the probability of the unseen die roll yielding a 1 is 50/50.

No. If we are considering a question with a 90% / 10% initial distribution of probabilities then we can't just ignore that distribution and start at 50%. — Devans99

We roll a 10-sided dice. If it rolls a 1 or a 2 I will place a red ball in a box.

Before you check the box you consider that the initial probability that the dice rolls a 1 is 10%.

You check the box and find a red ball. You know from this that there's a 50% chance that the dice rolled a 1 and a 50% chance that the dice rolled a 2.

Do you then add the initial 10% to this second 50% using your method to come up with some new likelihood that it rolled a 1? Or do you just accept that it's 50%?

@Devans99, you don't get to brush established theories aside with a hand-wave because they don't accommodate your belief.

He is essentially intellectually dishonest, which means just plain dishonest. — tim wood

Harsh. This is what religious indoctrination does to most people. It puts up mental barriers to true things that don't fit the pedagogically-derived rules while maintaining an open-door policy to doctrine that is clearly unjustifiable or incompatible with evidence. Abusing or inventing mathematics to make it fit the desired answer is probably a perfectly understandable thing to do if you start from the idea that any mathematics that yields the wrong answer is necessarily incorrect. Every creationist I've met thinks this way; I don't think it's a choice.

That's precisely the point. You cannot just say "Since we do not know if it is this guy, the probabilities are 50/50". And yet that is exactly how you proceeded. — Kenosha Kid

We have no data on the distribution of the answer space for 'was the universe a creation?' - so assuming it is normally distributed (50%/50%) is correct.

We have no data on the distribution of the answer space for 'was the universe a creation?' - so assuming it is normally distributed (50%/50%) is correct. — Devans99

So that is inventing the "fact" that it is binary. That is mathematically invalid.

We roll a 10-sided dice. If it rolls a 1 or a 2 I will place a red ball in a box.

Before you check the box you consider that the initial probability that the dice rolls a 1 is 10%.

You check the box and find a red ball. You know from this that there's a 50% chance that the dice rolled a 1 and a 50% chance that the dice rolled a 2.

Do you then add the initial 10% to this second 50% using your method to come up with some new likelihood that it rolled a 1? Or do you just accept that it's 50%? — Michael

This does not really seem related to my calculation. My calculation is for combining separate pieces of non-overlapping evidence that support a particular conclusion into a single, combined probability estimate.

So that is inventing the "fact" that it is binary. That is mathematically invalid. — Kenosha Kid

Its a binary question, we have no data on the likely distribution of the answer space, which of the following initial assumptions is most reasonable:

1. Assume a 0% chance that the universe is a creation
2. Assume a 100% chance that the universe is a creation
3. Assume 50% / 50%

If you go for [1] or [2], then you are biasing one way or the other based on no evidence. We really can't do that for 'is the universe a creation?' - we have no evidence - we can't for example say that 90% of known universes are actually creations - there is only one universe we know of - and we don't know whether it's a creation or not - so 50% / 50% is the only reasonable assumption.

Then with my calculation, I adjust this initial estimate to allow for the evidence we do have, which gets me to about 95% likely that the universe is a creation.

This does not really seem related to my calculation. My calculation is for combining separate pieces of non-overlapping evidence that support a particular conclusion into a single, combined probability estimate. — Devans99

It's exactly related to your calculation:

Assume the initial distribution is 50% guilty/ 50% innocent, and the first piece of evidence is 50% likely that he is guilty:

1) 50% guilty + 50% innocent x 50% = 75% chance he is guilty — Devans99

In my example the initial distribution is 90% the dice roll is > 1 and 10% the dice roll = 1 and the first piece of evidence (the red ball) is 50% likely that the dice roll = 1, so using your reasoning above there's either 10% + 90% x 50% = 55% probability that dice roll = 1 or 90% + 10% x 50% = 95% probability that dice roll = 1 (which one is it exactly? Do I multiply the 50% by the 10% chance of 1 or by the 90% chance of not 1)?

Quantumatics and relativity are established (already posted an example). — jorndoe

Forgive me for interrupting, but what exactly is the analogy to causation? Are you guys relating quantum physics to probability/the likelihood of a type of Heraclitus Logos, an objective cosmic law, etc.?

Otherwise, there will always be an element of uncertainty in nature; determinism and indeterminism, chance and choice, changing and unchanging, etc. Sure, QM as an element of both. The quantum uncertainty of an electron cannot have a well defined position and momentum at the same time. You can make measurement of the position to obtain a value, but the value of the momentum is completely uncertain. So for a general quantum state, it is apparently impossible to say in advance what value will be obtained by a single measurement, only probabilities can be assigned.

In that sense, yes, the Logos of causation appears that there are only ranges of outcomes available. But here's the important part. The system of QM is indeterministic as well as contingent. Yet, on the other hand, the experimenter determines whether the measurement shall be of position or momentum, so the range of alternatives is decided by an external agent. And so as far as the electron is concerned, the nature of alternatives is fixed necessarily, and the actual alternative adopted is contingent.

To me, one of many implications of the quantum state of affairs is that it eliminates complete randomness and chaos from the universe/existence. It combines choice and chance/contingency and necessity. It suggests yet another dipolar driving force in nature. The world is neither wholly determined or arbitrary.

This also suggests, in my view, that a participatory anthropic involvement of causation is at work. Kind of like cognitive science/our stream of consciousness. Random thoughts appear to us; it is our choice to choose which thoughts are of value. Causation, in this way, has intrinsic metaphysical value to humans, no?.

In my example the initial distribution is 90% the dice roll is > 1 and 10% the dice roll = 1 and the first piece of evidence (the red ball) is 50% likely that the dice roll = 1, so using your reasoning above there's either 10% + 90% x 50% = 55% probability that dice roll = 1 or 90% + 10% x 50% = 95% probability that dice roll = 1 (which one is it exactly? Do I multiply the 50% by the 10% chance of 1 or the 90% chance of not 1)? — Michael

Its really not related - the first piece of evidence - the probability of 10% because it is a 10-sided die, is replaced completely by the second piece of evidence - the probability of 50% because of the red ball - so we have a probability of 50% that it was a one.

My method is for combining separate, unrelated pieces of evidence into a single probability estimate.

The example you have given does not require any combining of probabilities - the second piece of evidence overlaps and supersedes the fact that it is a 10-sided die.

Its a binary question — Devans99

It's the same binary question as "Is the die showing a 1" The answer then is not 50/50.

Saying "the probability of God creating the universe is 50/50" is identical to saying: "there are precisely two ways the universe can have been created and we know not which". Not knowing the possible means of the universe being created is not leave to invent the non-fact that there were precisely two. It sounds like this has been explained to you before.

The rest of your "calculation" proceeds from this error and introduces myriad more. There is no point treating it, since the root of your problem is right at the start.

Its really not related - the first piece of evidence - the probability of 10% because it is a 10-sided die, is replaced completely by the second piece of evidence - the probability of 50% because of the red ball - so we have a probability of 50% that it was a one.

My method is for combining separate, unrelated pieces of evidence into a single probability estimate. — Devans99

So why doesn't the first piece of evidence (the finger prints on the knife) replace the initial, uninformed probability of 50% guilt based on lack of any evidence one way or the other?

And do you even know what it means to say that the evidence suggests a 50% chance of guilt? It means that it is equally likely that there is an innocent explanation as there is a guilty explanation, and as such cannot be used to either suggest guilt or to suggest innocence.

It's like taking a 50% chance of heads as evidence to increase the likelihood that it's heads. That's nonsense.

Saying "the probability of God creating the universe is 50/50" is identical to saying: "there are precisely two ways the universe can have been created". Not knowing the possible means of the universe being created is not leave to invent the non-fact that there were precisely two. It sounds like this has been explained to you before. — Kenosha Kid

I assigned an initial probability estimate of 50% / 50% to the question 'is the universe a creation?'.

Let me ask you, what initial probability estimate would you yourself assign to this question?

Devans, then, is just a waste of time. — tim wood

Yep.

I assigned an initial probability estimate of 50% / 50% to the question 'is the universe a creation?'. — Devans99

Yes, I know, that's what I and a bunch of other people are telling you is wrong. It is an invalid starting point for Bayesian inference.

Let me ask you, what initial probability estimate would you yourself assign to this question? — Devans99

Me personally? Zero, since only possible causes should be included and God has not been shown to be even possible. But if I were to attempt Bayesian inference and include the God hypothesis as one, I would need to know all of the possible options, not just those known today, but those known in the future and those never figured out. These would need to be cast into mutually exclusive categories. Then whatever number of categories N I ended up with, the probability of intelligence creation would be generously assigned the value 1/N.

The extent to which this cannot be done is the extend to which your methodology is invalid. It doesn't become valid just because one of the options you do know about happens to be the one you want.

So why doesn't the first piece of evidence (the finger prints on the knife) replace the initial, uninformed probability of 50% guilt based on lack of any evidence one way or the other? — Michael

1) There is an initial estimate of 50%/50% based on the fact that people at trials are 50% likely to be guilty.

2) Then the initial estimate is adjusted to reflect the first piece of evidence.

3) Then that estimate is adjusted to reflect the first piece of evidence

So no piece of evidence replaces any prior piece of evidence - that would not be combining probabilities into a single estimate - instead the probability estimate is adjusted in the light of subsequent evidence.

And do you even know what it means to say that the evidence suggests a 50% chance of guilt? It means that it is equally likely that there is an innocent explanation as there is for a guilty explanation, and as such cannot be used to either prove guilt or to prove innocence. — Michael

'prints on the knife' implies a 50% chance someone is guilty. The fact that there are 'prints on the knife' does not imply a 50% chance of innocence.

Think what you are saying - there are prints on the knife so we can increase the probability estimate he is innocent - that's clearly wrong.

The fact that there are 'prints on the knife' does not imply a 50% chance of innocence.

'prints on the knife' implies a 50% chance someone is guilty. The fact that there are 'prints on the knife' does not imply a 50% chance of innocence.

Think what you are saying - there are prints on the knife so we can increase the probability estimate he is innocent - that's clearly wrong.

The fact that there are 'prints on the knife' does not imply a 50% chance of innocence. — Devans99

Tell me what you think it means for evidence to imply a 50% chance that someone is guilty.

1) There is an initial estimate of 50%/50% based on the fact that people at trials are 50% likely to be guilty.

2) Then the initial estimate is adjusted to reflect the first piece of evidence.

3) Then that estimate is adjusted to reflect the first piece of evidence

Thank God you're not a juror. This is crazy.

Me personally? Zero, since only possible causes should be included and God has not been shown to be even possible. — Kenosha Kid

Thats just bonkers - spacetime cannot have existed forever - so how exactly do you have it a 'not a creation'?

Thats just bonkers - spacetime cannot have existed forever - so how exactly do you have it a 'not a creation'? — Devans99

I didn't say it doesn't have a creation, I said it wasn't created by an intelligent deity.

Tell me what you think it means for evidence to imply a 50% chance that someone is guilty. — Michael

I am trying to combine separate pieces of evidence into a combined probability estimate:

1) We have a completely independent first piece of evidence. If this piece of evidence was presented on its own, it would imply a 50% chance of guilt.
2) Then another completely independent second piece of evidence implies separately a 25% chance of guilt.

So [1] by itself implies a 50% chance he is the killer. And [2] separately implies a 25% chance he is the killer. The question is how do you combine these into a single probability estimate?

It is clear the combined probability estimate must be higher than the 50% alone we have for the first piece of evidence. I can see no other way of doing the calculation than:

50% guilty + 50% innocent X 25% = 62.5% guilty.

Think about it geometrically, we can represent the 50%/50% probability space as a square - half of it marked guilty, the other half marked innocent.

We then take the 50% innocent and say well actually we have evidence that he is 25% likely to be guilty - so we take 25% of the innocent half of the square and add it to the 50% guilty part of the square - giving 62.5% guilty.

Thank God you're not a juror. This is crazy. — Michael

Well at least I have a clue how to combine separate pieces of evidence into an overall probability estimate - you seem to have no idea whatsoever.

I didn't say it doesn't have a creation, I said it wasn't created by an intelligent deity. — Kenosha Kid

I don't believe in random so that just leaves the creation of spacetime as a deliberate act.

So [1] by itself implies a 50% chance he is the killer. And [2] separately implies a 25% chance he is the killer. The question is how do you combine these into a single probability estimate?

It is clear the combined probability estimate must be higher than the 50% alone we have for the first piece of evidence. I can see no other way of doing the calculation than:

50% guilty + 50% innocent X 25% = 62.5% guilty. — Devans99

Before we continue, clarify something for me. If the initial distribution is 60% chance he is guilty, which of these is correct:

1. 60% guilty + 40% innocent x 25% = 70% guilty
2. 40% innocent + 60% guilty x 25% = 55% guilty

I don't believe in random so that just leaves the creation of spacetime as a deliberate act. — Devans99

Then if that were true, the only possibility would be intelligent creation which would be 100%, not 50%. And your argument reduces to:

Given that the universe had a beginning
And I don't believe anything other than an intelligent creator could've done it
God exists

Not very compelling.

Devans, then, is just a waste of time. — tim wood

Could be.

I try to converse with everyone, but some conversations ARE harder than others.

In my opinion, any discussion of whether any gods exist or not...

...is pretty much a waste of time.

But I am in those kinds of discussions all the time, which belies that opinion.

Oh well.

BOTTOM LINE in my opinion: We do not know and cannot logically estimate the likelihood in either direction.