Atheist v. Agnostic: Friend or Foe?

[Dunstan]

And I assumed that the atheist claim is certainty that there is no God.

A strange assumption for you to make, given that many people have noted in this very thread that they consider themselves atheists but do not claim certainty.

I should have qualified my statement by saying that some atheists are semi atheists hovering in the never-never land between atheist and agnostic and that my opinion only was to be applied to those who claim absolute certainty.

Could you give some examples of atheists who claim certainty? I'm sure there must be some out there, but I'm not aware of any.

Of course I was guided by my penchant for considering people who ridicule the ID idea as claiming certainty. Otherwise, why would they ridicule?

One can acknowledge that absolute certainty does not exist without agreeing that all possibilities are equally likely. Some ideas are absurd and deserving of ridicule. Just because we don't know for certain exactly what "caused" the Big Bang doesn't mean that someone who seriously claims the universe was sneezed out of the nose of the Great Green Arkleseizure can't be laughed at.

But considering the human penchant and predisposition for semantic vagueness as a patina for profundity I guess I might have jumped the gun. Does gloating and ridiculing those who postulate an ID indicate certainty? Perhaps I take it to indicate certainty since that's what it conveys.

I don't why it conveys that to you. Essentially what you're saying is that ridicule is never appropriate, because one cannot be certain of anything. I think that's an uncommon position to take.

BTW
Your response about doctors family and all the other balogna you mentioned to my original post I found irrelevant and unintelligible. Thought about asking for a clarification but let it go.

Actually, you didn't let it go. Letting it go would mean just ignoring it and not making a comment. What you did is just passive-aggressive.

The point was this: None of us can know with absolute certainty who our biological father is. Even DNA tests have a non-zero margin of error, and most of us don't even have the benefit of such a test. But, unless they have some specific reason to question their paternity, practically no one goes around making all the absurd qualifiers about who their father is in my mock dialogue, even in a doctor's office where the conversation is confidential and it's medically important to be accurate. And certainly no one would jump on the assertion "X is my father" with a critique about "you're claiming certainty, and that's impossible!" Everyone understands that it's impossible to know such a thing with absolute certainty, and nobody feels the need to create a special word for "person who is absolutely positively my father."

Yet people apply a completely different standard when it comes to god claims, and accuse atheists of claiming absolute 100% certainty. I don't see why we need a special word "atheist" to connote "one who claims to know with certainty that there is no god," since (a) we don't do it for any other assertion I can think of;* and (b) such people (the certain ones, I mean) seem to be incredibly rare. Whereas we do need a word to describe the comparatively large group of people who do not believe in the existence of a god, but do not necessarily claim certainty on that (or any other) point; "agnostic" doesn't suffice, in my opinion, because it seems to be working overtime covering those who haven't reached even a provisional opinion on the subject, as well as those who claim it is impossible to know, and all the other definitions given in this thread and others like it.

*-indeed, we don't even have a parallel word for someone who claims the existence of god with absolute certainty. When a theist admits that "yes, I could be wrong, it could be that my god doesn't exist," I've seen people exclaim "aha! So you're an agnostic!" But perhaps that kind of thing does go on at other forums.

Well, sorry but I find the atheistic stance far more worthy of marveling at. Similar to a perverse convenient sort of self-inflicted blindness.

Well, we all get to choose at what we will marvel.

 

[Jono]

The terms are even more subjective than "religious" or "deitist", so here enters the personal opinion of that atheist and agnostic mean. Without a doubt, if they are distinct from each other, we will still find self-ascribed atheists who are actually agnostic and vice versa.

Personally, agnostic is the absence of belief or disbelief in God (disbelief can oft be another form of belief), you simply do not entertain an active attachment in "there is no God" or "there is a God", moreso a commonplaced "maybe, who knows".
Atheists would then be those who can relate to the thought; "I believe that there is no God" and Agnostics to "I do not know/care/believe either way".
Science by itself ideally is agnostic, since it does not presume to set out to prove or disprove existence of God, as it does not come into play. Whereas, atheistic approach would be the thought that since science does not provide any evidence for God, there is no God.

It has little to do with 100% certainty, since that is in itself a mythical position one tends to hold when kidding oneself, or so I've heard.

 

[Dunstan]

When a theist admits that "yes, I could be wrong, it could be that my god doesn't exist," I've seen people exclaim "aha! So you're an agnostic!" But perhaps that kind of thing does go on at other forums.

Oops -- I meant to say I haven't seen people exclaim that. Too late to edit now.

 

[fls]

Huh? I'm not sure what you mean by this. What theorems of Fisher are you thinking of? I think the sample space can easily be defined into a binary set of outcomes. A non-interventionist intelligent creator god exists outside of our universe or doesn't. That's a pretty well defined sample space. The universe is a conscious entity or it isn't. That too is a pretty well defined sample space.

Well, with that sort of thinking, I'll have to bow out of this conversation with you. I just realized that I'm going to be struck by lightning or not struck by lightning this afternoon, so I better take care of a few more Christmas presents instead (can't like the look of those odds).

Linda

 

[Beth]

Well, with that sort of thinking, I'll have to bow out of this conversation with you. I just realized that I'm going to be struck by lightning or not struck by lightning this afternoon, so I better take care of a few more Christmas presents instead (can't like the look of those odds).

Linda

Oh, honestly Linda, did you forget we were discussing what you had described as being "indistinguishable ideas"? There is a difference between claiming all binary sample spaces have equal probilities for the two possible outcomes and saying that when there is no evidence to base probability on, it's rational to consider the two binary outcomes as having equal probability until some additional evidence is available. You know the difference between those two situations as least as well as I do, so why the disingenuousness?

 

[Jono]

Oh, honestly Linda, did you forget we were discussing what you had described as being "indistinguishable ideas"? There is a difference between claiming all binary sample spaces have equal probilities for the two possible outcomes and saying that when there is no evidence to base probability on, it's rational to consider the two binary outcomes as having equal probability until some additional evidence is available. You know the difference between those two situations as least as well as I do, so why the disingenuousness?

I believe Chris Langan and his model for the universe set out to establish God's existence in terms of probability as an intelligent designer through binary sample spaces. I haven't really read much on his work, and I do not really care enough to do so, albeit it is a nice curiosa, with one of the allegedly smartest men on Earth and a devout ID'er of his own brand to boot.

 

[Beth]

I believe Chris Langan and his model for the universe set out to establish God's existence in terms of probability as an intelligent designer through binary sample spaces. I haven't really read much on his work, and I do not really care enough to do so, albeit it is a nice curiosa, with one of the allegedly smartest men on Earth and a devout ID'er of his own brand to boot.

I don't believe that's possible, but then again, that's why I'm a strong agnostic on certain definitions of god.

 

[fls]

Oh, honestly Linda, did you forget we were discussing what you had described as being "indistinguishable ideas"? There is a difference between claiming all binary sample spaces have equal probilities for the two possible outcomes and saying that when there is no evidence to base probability on, it's rational to consider the two binary outcomes as having equal probability until some additional evidence is available. You know the difference between those two situations as least as well as I do, so why the disingenuousness?

Yes, I do know the difference between the two. In one case, you know that you are suggesting something that's very silly. In the other case, you don't know that you are suggesting something that's very silly. However, my unwillingness to pretend that ignorance serves to make latter rational is an accurate representation of my thoughts on this matter - i.e. your claim of disingenuousness is unwarranted.

Linda

 

[Jono]

I don't believe that's possible, but then again, that's why I'm a strong agnostic on certain definitions of god.

Hear hear. Or should I say, hear here or hear there?

 

[arthwollipot]

And I assumed that the atheist claim is certainty that there is no God.

Here it is again. My thesis is correct.

To a theist, belief is the default assumption. Theists believe that one must take a postive step to disbelieve.

 

[Beth]

Yes, I do know the difference between the two. In one case, you know that you are suggesting something that's very silly. In the other case, you don't know that you are suggesting something that's very silly. However, my unwillingness to pretend that ignorance serves to make latter rational is an accurate representation of my thoughts on this matter - i.e. your claim of disingenuousness is unwarranted.

Linda

You are right, I don't know that I am suggesting something that's very silly. Instead I find your comparison ludicrous because you are comparing setting the probability of two unknown and undistinguishable conjectures to a situation with well known extreme probalities that have been computed based on estalished and verifiable evidence and saying the one should be evaluated as the other. I've no idea why you would conclude that and felt you understood enough about statistics to known that such an assessment is, well, silly.

I can, if you like, give you a couple of reasons why statisticians like the uniform distribution for setting probabilities when no other evidence exists to distinguish the different possibilities. But I won't bore you with things like the minimax theorem unless you ask me to.

How is it you feel comfortable setting the probability of those particular "indistinguishable ideas" so close to a 0/1 distribution (indicating near certainty) that you find it very silly to use the choice of .5/.5 for such situations?

Earlier you said we didn't "differ all that much on residual probabilities". Do you see now where we differ? Do you find it a significant enough difference to warrant making a distinction between agnostic and atheist?

 

[fls]

You are right, I don't know that I am suggesting something that's very silly. Instead I find your comparison ludicrous because you are comparing setting the probability of two unknown and undistinguishable conjectures

This part needs clarification from me. When I initially mentioned indistinguishable ideas, I made reference to several different god concepts, all of which are formulated in such a way as to make them consistent with what we already know about the nature of the universe. The reason that they are indistinguishable is the same reason that they are not inconsistent with what we already know - a universe in which they are present is the same as a universe in which they are not present. So when I first mentioned assigning probabilities to these various ideas, it was with the idea of distributing probabilities among a hodge-podge assortment of more or less well-formulated ideas. You subsequently changed it to considering each idea individually, as though there is only one possible idea and only one distinction to be made in the set of all possible ideas.

to a situation with well known extreme probalities that have been computed based on estalished and verifiable evidence and saying the one should be evaluated as the other.

I said that in order to attempt (vainly, as it turned out) to illustrate to you where you had gone wrong. If you are talking about using a uniform distribution in order to form a probability distribution, you fill that set with values that are all equally probable. If you don't know the probability for each value, then you fill it with all possible occurrences/examples. For the lightning example, you fill it full of all those times when I am not struck by lightning in addition to adding the one (or maybe more, I guess) time that I am. You don't simply fill it with one example of each. If you wish to understand the probability that the universe is a conscious entity, you fill it with all the examples where it is not a conscious entity in addition to the examples where it is. You don't simply pick one example of each.

I can, if you like, give you a couple of reasons why statisticians like the uniform distribution for setting probabilities when no other evidence exists to distinguish the different possibilities. But I won't bore you with things like the minimax theorem unless you ask me to.

I am already familiar with those ideas. Please don't hide behind the use of technical terms.

How is it you feel comfortable setting the probability of those particular "indistinguishable ideas" so close to a 0/1 distribution (indicating near certainty) that you find it very silly to use the choice of .5/.5 for such situations?

Because I recognize that you have incorrectly filled your uniform distribution.

Earlier you said we didn't "differ all that much on residual probabilities". Do you see now where we differ? Do you find it a significant enough difference to warrant making a distinction between agnostic and atheist?

I wanted to avoid making the distinction the result of faulty reasoning.

Linda

 

[Beth]

This part needs clarification from me. When I initially mentioned indistinguishable ideas, I made reference to several different god concepts, all of which are formulated in such a way as to make them consistent with what we already know about the nature of the universe. The reason that they are indistinguishable is the same reason that they are not inconsistent with what we already know - a universe in which they are present is the same as a universe in which they are not present.

Thank you for the clarification. I had mentioned two specific such concepts and thought you were referring to them.

So when I first mentioned assigning probabilities to these various ideas, it was with the idea of distributing probabilities among a hodge-podge assortment of more or less well-formulated ideas.

That would explain your criticism that it was not a well-defined sample space.

You subsequently changed it to considering each idea individually, as though there is only one possible idea and only one distinction to be made in the set of all possible ideas.

I was defining the sample space I was referring to, not changing anything but attempting to clarify what we were discussing.

I said that in order to attempt (vainly, as it turned out) to illustrate to you where you had gone wrong.

It didn't work because that was not an error I was making, I simply felt it was a ludicrous analogy.

If you are talking about using a uniform distribution in order to form a probability distribution, you fill that set with values that are all equally probable. If you don't know the probability for each value, then you fill it with all possible occurrences/examples. For the lightning example, you fill it full of all those times when I am not struck by lightning in addition to adding the one (or maybe more, I guess) time that I am. You don't simply fill it with one example of each. If you wish to understand the probability that the universe is a conscious entity, you fill it with all the examples where it is not a conscious entity in addition to the examples where it is. You don't simply pick one example of each.

A binomial sample space is created by placing all possible outcomes into one of two categories. Probabilities are computed by looking as all the different ways each outcome can occur. Thus the relative frequency of each outcome is assessed and a probability assigned based on that. You're correct about all that. But to move beyond the 50/50 assessment, you have to have some idea of the possible configurations of a universe like ours and which would result in it having consciousness and which would not. Do you feel you have such knowledge and can make such an assessment with greater accuracy for that reason? I would be interested to hear your thoughts on that.

I am already familiar with those ideas. Please don't hide behind the use of technical terms.

I thought you might be, which is why I didn't regale you with explanations of what it is and why I choose to use it in this situation. I'm not hiding behind anything, I'm attempting to explain my point of view and why I find it rational rather than silly.

Because I recognize that you have incorrectly filled your uniform distribution.

My choices are not incorrect; they are simply different than yours. I see no reason given for why your choices are so superior that you term mine 'very silly'.

I wanted to avoid making the distinction the result of faulty reasoning.

Linda

Understandable, as doing so would be an example of faulty reasoning. Making different choices for unknown probabilities in the situations we are discussing is hardly an example of faulty reasoning. The minimax approach is not the only choice in evaluating such estimates of probabilities, but it is a reasonable one.

So, do you find those differences sufficient to merit having different terms, agnostic and atheist, for the beliefs?

I may not be able to respond again until after Christmas. So I'd like to take this opportunity to wish you a merry Christmas.

 

[fls]

A binomial sample space is created by placing all possible outcomes into one of two categories. Probabilities are computed by looking as all the different ways each outcome can occur. Thus the relative frequency of each outcome is assessed and a probability assigned based on that. You're correct about all that. But to move beyond the 50/50 assessment, you have to have some idea of the possible configurations of a universe like ours and which would result in it having consciousness and which would not.

I have bolded your error. There is nothing special about 50/50 that makes it the default assessment.

Do you feel you have such knowledge and can make such an assessment with greater accuracy for that reason? I would be interested to hear your thoughts on that.

My initial question was rhetorical for just that reason.

I thought you might be, which is why I didn't regale you with explanations of what it is and why I choose to use it in this situation. I'm not hiding behind anything, I'm attempting to explain my point of view and why I find it rational rather than silly.
My choices are not incorrect; they are simply different than yours. I see no reason given for why your choices are so superior that you term mine 'very silly'.

Your choices are unjustifiable. Regardless of what technical jargon you wish to hide behind, all you are doing is assuming that which you are trying to prove. You wish to justify the rationality of your choice by making all choices equivalent. And you do this by assuming that all the choices are equivalent. It's faulty reasoning regardless of whether or not I choose to call you on it.

Understandable, as doing so would be an example of faulty reasoning. Making different choices for unknown probabilities in the situations we are discussing is hardly an example of faulty reasoning.

The fault is not in making different choices. The fault is in thinking that you've made a rational choice.

The minimax approach is not the only choice in evaluating such estimates of probabilities, but it is a reasonable one.

The minimax approach is not something that allows you to evaluate estimates in the absence of information, though. It still requires assumptions on your part.

Starting with an assumption and running it through a machine does not magically legitimize it.

So, do you find those differences sufficient to merit having different terms, agnostic and atheist, for the beliefs?

I may not be able to respond again until after Christmas. So I'd like to take this opportunity to wish you a merry Christmas.

Merry Christmas to you to.

Linda

 

[Beth]

Linda,

I like to take this opportunity to let you know how much I have enjoyed our conversation. It's always a pleasure to discuss these matters with someone of your intelligence and thoughtfulness.

A binomial sample space is created by placing all possible outcomes into one of two categories. Probabilities are computed by looking as all the different ways each outcome can occur. Thus the relative frequency of each outcome is assessed and a probability assigned based on that. You're correct about all that. But to move beyond the 50/50 assessment, you have to have some idea of the possible configurations of a universe like ours and which would result in it having consciousness and which would not.

I have bolded your error. There is nothing special about 50/50 that makes it the default assessment.

Actually, yes is it special. It is the choice that minimizes the maximum possible error for a binomial distribution. That is what I meant when I brought up the minimax theorem, which you claimed to be familiar with. As such, it is a fairly popular choice for situations without evidence to otherwise distinguish between the possibilities. It is not the only criteria someone might use, so I don't claim it to be the ONLY rational choice. By what criteria do you find the 50/50 assessment so wrong as to consider it irrational and unjustifiable?

Do you feel you have such knowledge and can make such an assessment with greater accuracy for that reason? I would be interested to hear your thoughts on that.

My initial question was rhetorical for just that reason.

So, you don’t have any idea how to set the probabilities, but the choice of equal weights is an irrational one? I don’t follow this reasoning at all.

I thought you might be, which is why I didn't regale you with explanations of what it is and why I choose to use it in this situation. I'm not hiding behind anything, I'm attempting to explain my point of view and why I find it rational rather than silly.
My choices are not incorrect; they are simply different than yours. I see no reason given for why your choices are so superior that you term mine 'very silly'.

Your choices are unjustifiable. Regardless of what technical jargon you wish to hide behind, all you are doing is assuming that which you are trying to prove. You wish to justify the rationality of your choice by making all choices equivalent. And you do this by assuming that all the choices are equivalent. It's faulty reasoning regardless of whether or not I choose to call you on it.

The ‘technical jargon’ you scorn is the justification of setting the probability of the two choices to be equal. You complain about ‘faulty reasoning’ on my part with an argument that makes assumptions about my motivations. Further, my motivations, whatever they may be, don’t affect the rationality of my argument. All in all, this isn’t a well reasoned or rational response to what I said. It’s an emotional rejection of my argument without any consideration of the substance of it, presumably because you don’t like the conclusion or possibly because you don't follow the technical details of the argument.

The fault is not in making different choices. The fault is in thinking that you've made a rational choice.

What knowledge or assumptions lead you to believe that my 50/50 choice is an irrational one? As near as I can tell, your argument that it is not a rational choice amounts to no more than “because I say it’s not”.

The minimax approach is not the only choice in evaluating such estimates of probabilities, but it is a reasonable one.

The minimax approach is not something that allows you to evaluate estimates in the absence of information, though.

You said you were familiar with this? It’s a criteria for selecting an estimate . It sets the criteria as the value that minimizes the maximum possible error.

It still requires assumptions on your part.

Of course. Everything requires assumptions. By employing such a method, the assumptions are known and stated up front.

What is irrational about them?

What assumptions are you making that lead you to believe that my 50/50 choice is an irrational one?

Starting with an assumption and running it through a machine does not magically legitimize it.

That’s not accurate. The ‘machine’ as you term it does not start with the assumption of a 50/50 assessment, but produces it as the best fit for the criteria selected. What is illegitimate about that?

So, do you find those differences sufficient to merit having different terms, agnostic and atheist, for the beliefs?

I’d still like to know your thoughts on this matter.

 

[fls]

Actually, yes is it special. It is the choice that minimizes the maximum possible error for a binomial distribution.

I suspected that is what you were referring to. The problem is that it is not relevant to the choice that you are making.

It is useful when you are doing something like figuring out the maximum sample size that limits your error to a certain interval when you don't know in advance what that proportion is (such as the proportion of ideas that are true). But it provides no guidelines as to what the underlying proportion is, other than assuming it can approximate a normal distribution. All you can conclude is something like, "the proportion of true ideas to false ideas is within 3% of the proportion of true ideas to false ideas in my sample". But it doesn't give you any information about the measured value for the sample parameter, and that is the only information that is relevant for this question. There are no natural limits for this value.

That is what I meant when I brought up the minimax theorem, which you claimed to be familiar with. As such, it is a fairly popular choice for situations without evidence to otherwise distinguish between the possibilities. It is not the only criteria someone might use, so I don't claim it to be the ONLY rational choice. By what criteria do you find the 50/50 assessment so wrong as to consider it irrational and unjustifiable?

Consider these questions...

What proportion of people will be struck by lightning this year?
What proportion of people will drink water today?
What proportion of particles in the universe consist of anti-matter?

Other than assisting you in calculating sample size, please provide an example of how you would act on an assumption of a 50/50 distribution.

So, you don’t have any idea how to set the probabilities, but the choice of equal weights is an irrational one? I don’t follow this reasoning at all.

You have picked a number between one and a million. I am simply recognizing that your choice is not anything other than arbitrary.

The ‘technical jargon’ you scorn is the justification of setting the probability of the two choices to be equal.

I realize that you think it is. But there is a difference between knowing how to use a formula and knowing where to apply that formula and what kind of answer it gives you.

You complain about ‘faulty reasoning’ on my part with an argument that makes assumptions about my motivations. Further, my motivations, whatever they may be, don’t affect the rationality of my argument. All in all, this isn’t a well reasoned or rational response to what I said. It’s an emotional rejection of my argument without any consideration of the substance of it, presumably because you don’t like the conclusion or possibly because you don't follow the technical details of the argument.

I don't know what your motivations are. I was simply pointing out that you were committing a logical fallacy. If you think I am not following some technical details, simply describe them in greater detail for me. All along, it has looked to me like you were talking about minimizing the maximum possible error for a binomial distribution rather than providing a reasonable estimate of the measurement underlying that error.

What knowledge or assumptions lead you to believe that my 50/50 choice is an irrational one? As near as I can tell, your argument that it is not a rational choice amounts to no more than “because I say it’s not”.

Because, as far as we know, any other choice is equally probable.

You said you were familiar with this? It’s a criteria for selecting an estimate . It sets the criteria as the value that minimizes the maximum possible error.

Of course. Everything requires assumptions. By employing such a method, the assumptions are known and stated up front.

What is irrational about them?

What assumptions are you making that lead you to believe that my 50/50 choice is an irrational one?

That’s not accurate. The ‘machine’ as you term it does not start with the assumption of a 50/50 assessment, but produces it as the best fit for the criteria selected. What is illegitimate about that?

As I pointed out above, your assumption applies to the calculation of the error, but it doesn't apply to the value of the sample parameter. And it is the latter which is relevant.

I’d still like to know your thoughts on this matter.

I will let you know once you have addressed the issue that your choice of a value is arbitrary, and that the outcome for any decision-making will be sensitive to that value.

Linda

 

[fls]

In support of the idea that assuming even odds leads you astray...

http://medicine.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pmed.0020124

Linda

 

[Beth]

I suspected that is what you were referring to. The problem is that it is not relevant to the choice that you are making.

It is useful when you are doing something like figuring out the maximum sample size that limits your error to a certain interval when you don't know in advance what that proportion is (such as the proportion of ideas that are true). But it provides no guidelines as to what the underlying proportion is, other than assuming it can approximate a normal distribution. All you can conclude is something like, "the proportion of true ideas to false ideas is within 3% of the proportion of true ideas to false ideas in my sample". But it doesn't give you any information about the measured value for the sample parameter, and that is the only information that is relevant for this question.

Samples don't have parameters, populations do. Statistics are computed from samples which lead to estimated values of the true population parameters.

There are no natural limits for this value.

Are we discussing the same value? I'm talking about the probability of the two outcomes of the sample spaces I defined earlier in this thread. There are natural limits for these values. Like all probabilities, they must lie between zero and one.

Consider these questions...

What proportion of people will be struck by lightning this year?

I don't know, but this I could look up and find out. It will be a very small proportion.

What proportion of people will drink water today?

This will be a number very close to 1.

What proportion of particles in the universe consist of anti-matter?

A unknown parameter that has various values depending upon which physics theories you favor.

Other than assisting you in calculating sample size, please provide an example of how you would act on an assumption of a 50/50 distribution.

In the first two situations, it would be rather ridiculous to do so because we all have personal experiences that it is rare for people to not drink water daily or to be struck by lighting. On the other hand, an assumption of a 50/50 value for anti-matter particles is a value that could be placed in various physics formulas to develop models of the creation and subsequent development of universes like our own.

I realize that you think it is. But there is a difference between knowing how to use a formula and knowing where to apply that formula and what kind of answer it gives you.

Indeed there is. Is there any particular reason that you feel this choice of criteria (minimax) is inappropriate in this situation? Or is just because you don't like the outcome?

I don't know what your motivations are. I was simply pointing out that you were committing a logical fallacy. If you think I am not following some technical details, simply describe them in greater detail for me. All along, it has looked to me like you were talking about minimizing the maximum possible error for a binomial distribution rather than providing a reasonable estimate of the measurement underlying that error.

I'm not following you at all here. What measurement? There is no measurement I am aware of, there is only a hypothosis and a sample space defined.

Because, as far as we know, any other choice is equally probable.

Again, I don't follow you. If any other choice is equally probably, why is selecting the 50/50 one irrational? You seem to have selected a value of near certainty on one particular side. Why do you feel your choice is rational and mine is not if all are equally probable?

As I pointed out above, your assumption applies to the calculation of the error, but it doesn't apply to the value of the sample parameter. And it is the latter which is relevant.

What are you talking about?

You have picked a number between one and a million. I am simply recognizing that your choice is not anything other than arbitrary.

I will let you know once you have addressed the issue that your choice of a value is arbitrary, and that the outcome for any decision-making will be sensitive to that value.

Well, you're close. I picked a number between zero and one. I don't feel it was an arbitrary choice any more that you feel your choice in regard to that value is arbitrary. However, there are different definitions of arbitrary. If you will provide me with your definition, I will be happy to let you know if I agree that my choice is arbitrary if you will let me know whether or not you feel your choice is arbitrary according to that definition. I will be happy to acknowledge that the outcome for any decision-making will be sensitive to that value though.

 

[fls]

Samples don't have parameters, populations do. Statistics are computed from samples which lead to estimated values of the true population parameters.

Are we discussing the same value? I'm talking about the probability of the two outcomes of the sample spaces I defined earlier in this thread. There are natural limits for these values. Like all probabilities, they must lie between zero and one.

Change my statements to "but it doesn't give you any information about the measured value for the sample, and that is the only information that is relevant for this question" and "there are no natural limits for the odds for this value."

I don't know, but this I could look up and find out. It will be a very small proportion.

This will be a number very close to 1.

Okay, so how would your assumption that the probability for either is 50/50 be useful?

A unknown parameter that has various values depending upon which physics theories you favor. In the first two situations, it would be rather ridiculous to do so because we all have personal experiences that it is rare for people to not drink water daily or to be struck by lighting. On the other hand, an assumption of a 50/50 value for anti-matter particles is a value that could be placed in various physics formulas to develop models of the creation and subsequent development of universes like our own.

Yes, you could model various universes using various proportions and observe the results. What specific advantage does the choice of 50/50 give to you (above and beyond an additional data point)?

Indeed there is. Is there any particular reason that you feel this choice of criteria (minimax) is inappropriate in this situation? Or is just because you don't like the outcome?

As I have stated several times now, my criticism is that your choice is arbitrary. The principle of minimax may be useful in order to minimize the maximum error, but it provides no guidelines as to the value you are applying the error to. You have picked a number merely for convenience; there is no rational justification for your number over any other number.

This is your original statment:

"I consider certain definitions of god, such as the non-intervening creator god and the pantheistic god as the totallity of the universe which has a consciousness of it's own to be completely indeterminable for us and as such, lacking other, more conclusive information, I rate the possibilities of existance and non-existance of those gods as equal. I gather you do not, so I think that would be a difference between our beliefs."

You essentially state that you base your beliefs on a number pulled from thin air. You seem committed to this idea, so it appears to me that the difference between you and me is that I am uncomfortable basing my opinion on stuff I've made up while you are comfortable doing so.

I'm not following you at all here. What measurement? There is no measurement I am aware of, there is only a hypothosis and a sample space defined.

The possibility that there is a non-intervening creator god or that the universe has a consciousness.

Again, I don't follow you. If any other choice is equally probably, why is selecting the 50/50 one irrational? You seem to have selected a value of near certainty on one particular side. Why do you feel your choice is rational and mine is not if all are equally probable?

I haven't chosen a value.

What are you talking about?

The possibility that any one (or more) of the god ideas you have mentioned are true.

Well, you're close. I picked a number between zero and one. I don't feel it was an arbitrary choice any more that you feel your choice in regard to that value is arbitrary. However, there are different definitions of arbitrary. If you will provide me with your definition, I will be happy to let you know if I agree that my choice is arbitrary if you will let me know whether or not you feel your choice is arbitrary according to that definition.

Chosen, not according to logic or an underlying principle, but due to whim or caprice. I would choose to pick a number based on logic or an underlying principle, instead.

I will be happy to acknowledge that the outcome for any decision-making will be sensitive to that value though.

We are in agreement that your justification for 'agnosticism' is sensitive to the manner in which you choose a value for the possibility for any particular god. Unfortunately, this suggests that the difference between agnosticism and atheism is that agnostics' beliefs are arbitrary, rather than rational.

Linda

 

[Beth]

Change my statements to "but it doesn't give you any information about the measured value for the sample, and that is the only information that is relevant for this question" and "there are no natural limits for the odds for this value."

I'm sorry, but your statement still isn't making any sense to me. What sample? What measured value?

Okay, so how would your assumption that the probability for either is 50/50 be useful?

I wouldn't make such an assumption in those cases because it wouldn't be useful. Why do you consider such examples appropriate analogies?

Yes, you could model various universes using various proportions and observe the results. What specific advantage does the choice of 50/50 give to you (above and beyond an additional data point)?

In such models, a point estimate is needed to run the simulations. Minimizing the maximum error is a reasonable criteria to select the value used.

As I have stated several times now, my criticism is that your choice is arbitrary. The principle of minimax may be useful in order to minimize the maximum error, but it provides no guidelines as to the value you are applying the error to. You have picked a number merely for convenience; there is no rational justification for your number over any other number.

I’m sorry you don’t understand/agree with this application of the minimax principle, but that doesn’t mean my choice is arbitrary. As I said before, the 50/50 number is a result of applying the minimax principle to the situation at hand. Claiming it isn’t appropriate (without providing any sort of reason as to why that would be) doesn’t make the fact that I used it an arbitrary or whimsical choice.

Again, I don't follow you. If any other choice is equally probably, why is selecting the 50/50 one irrational? You seem to have selected a value of near certainty on one particular side. Why do you feel your choice is rational and mine is not if all are equally probable?

I haven't chosen a value.

I realize you haven’t chosen a single value, but it’s clear from your previous posts that you hold a strong opinion on what are and are not reasonable values. Clearly, you don't feel the entire range of [0, 1] is equally likely because if you did, you would not consider my point estimate of .5 to be silly and irrational. You consider yourself atheist, which indicates that you place that value on the low end of that interval.

My guess is that you give it an interval estimate of something like [0, .1] or perhaps [0, .01] or perhaps even less, even though you likely don’t have a specific upper endpoint in mind. Is this an accurate assessment of your feelings? If not, could you please provide me with an interval you do consider to accurately reflect your feelings on the matter?

You seem committed to this idea, so it appears to me that the difference between you and me is that I am uncomfortable basing my opinion on stuff I've made up while you are comfortable doing so.

Perhaps you are right. I am quite comfortable saying what I think is a reasonable estimate for the value of that number. Please keep in mind that I'm not claiming it's correct, only that it's a reasonable guess.

You, on the other hand, do not find such an estimate reasonable. From other things you've written, it appears that you feel it ought to be a significantly lower figure for the probability of a non-intervening creator god or a pantheistic god, but are not comfortable admitting that you have, in fact, made such an assessment.

Chosen, not according to logic or an underlying principle, but due to whim or caprice. I would choose to pick a number based on logic or an underlying principle, instead.

We are in agreement that your justification for 'agnosticism' is sensitive to the manner in which you choose a value for the possibility for any particular god. Unfortunately, this suggests that the difference between agnosticism and atheism is that agnostics' beliefs are arbitrary, rather than rational.

Linda

I’ve given you a mathematical justification for choosing a particular point estimate of 0.5 – the minimax principle. You, for some reason don’t consider this to be logical or rational. You don’t feel this is an appropriate application for that principle, but can give no reason for that. However, even if you disagree with me regarding the use of this principle, I don’t understand why you would consider using such criteria to be an arbitrary selection rather than a rational one. Perhaps you could explain to me how you feel someone might rationally arrive at an estimate, either a point or an interval estimate? How did you rationally arrive at your opinion?

BTW I do not consider it rational to claim that it isn't reasonable to have an estimated value for a probability while holding the opinion that one outcome is far more likely than the other.