@orathaic "...and since you are assigned probabilities without reason, i can legitimately claim you are being unreasonable in your arguement."
I have no idea what that means. Who said I assigned probabilities without reason? I am following the correct practice of probability theory.
"'...If I know nothing but that I'm getting a sequence of H's and T's, and nothing else at all, then I can't infer anything about future probability from past probability'
"that's the point, if you know nothing at all then you know nothing at all. however, if you make an assumption, then you can make a prediction."
Well, fine, but if your assumption is based on nothing, your prediction is worthless.
"Now every time your prediction is correct then you have reinforced your reasons to think that it was correct in the first place. Of course it is still an assumption, but without making an assumption you can predict nothing, not a 50/50 chance of the next coin, nothing. As you said, you know nothing."
No you haven't! That's a massive fallacy. Things don't become rational because they chance to be true. They are either justified or not beforehand.
Moreover, again per Hume's argument, no amount of success can lend even a whit of support to the hypothesis that the future will continue to be ordered, and therefore, not a whit of support to any prediction that it will be.
(We're talking predictions here. Predictions that come true are useless by then -- they're past predictions. We're looking for reasons to give them credence beforehand).
"i said: "Our best estimate for a chaotic universe would still be to assume a 'regular' universe, but with some perturbation. "
you said: 'Not at all.'
but you're wrong, because ANY other model/estimate gives us 1/∞ chance of being accurate, which is 0."
Yes, that's precisely my point. Including the regular one.
"in your example, ALL models are effectively wrong, but that wasn't one in which there was no regularity, it was one where space was always 4-d and mass was always well-defined. You can't assume such a countabley infinite set of universe exist if there is no regularity. "
Umm, sure I can. I mean, yeah, I don't know that 4

and mass will stay around, but that only adds MORE possible universes, making the probability of order even worse. If you increase the denominator, you decrease the probability. And none of the non-4

universes would be ordered (since I think we'd both agree, dimension changing would count as a breakdown of all observed order), so that would decrease the probability of order. You might say I was calculating an upper bound. Of course, since it turned out to be zero, I was also calculating the actual probability.
"The fact that we have discussed two irregular universes, and one regular one, doesn't mean that the number of irregular ones (or probability of them existing) is twice that of the regular ones."
Well, no, but it means that it's MORE. You keep saying (infra) that there are all these possible regular universes? How do you figure? REGULAR universes are highly constrained. They have to keep having the sun come up, electrons behave a certain way, etc., etc. The parameters for possible regular universes are very few compared with those for irregular ones. They're basically determined completely (whatever the laws of physics have been so far, they have to stay the same, and that determines the universe -- modulo only QM randomness).
"There could be uncountably infinitely many regular universes aswell."
Again... how do you figure? We already know what the regular one looks like.
"Just assume that you are trying to count the number of regular universes and the Pi is the only difference between each regular universe, then start counting them at Pi = 0, and go up for all Pi a Real Number."
I have no idea what this means.
But, assuming we did have uncountably many regular universes, there are still, yes, ways of talking about the probability of an uncountable set in another uncountable set, and the regular universes would still have probability zero. For example, there are uncountably many continuous functions on the interval [0,1], and uncountably many functions on [0,1], but the continuous functions are nowhere dense in the set of functions, meaning roughly that they have probability zero of being chosen at random. So it is in every situation where one gives some kind of orderedness property among a total set of items.
"I could similarly question does the christian world view actually provide justification for believing that God will always be all-loving. So what is your point?"
Sure of course. That's part of the definition of the Christian world view, and in that world view, we are also able to know such things, because God (who is all-Truthful) told us and made us able to know.
"it does not beg the question. I admit i can't know (sceptic) I now have to decide how to live my life.
I have two choices, make an assumption and live my life, make no assumption and have no way to determine how i should make decisions.
if my aim is to determine how i should live, then the arguement from usefulness is all that is needed. "
But again, it's NOT USEFUL unless it's true. There are PLENTY of algorithms you could use that would be useful if they were true. I promise that if you believed that saying "Hocus Pocus" would put a billion dollars in your bank account right now, and never again, you'd say it, even if it contradicted the idea of order. You just don't because you don't believe it's true.
"because if i make this assumption (that is without any assumption of regularity) then it has to compete with all other iRegular assumptions for prediction. "
Right.... and so does the prediction of regularity.
"We are human and not capable of making any assumptions we want. You seem to be assuming some ideal thinking machine with infinite states."
No it's true, we do have to assume regularity. That doesn't go to whether it's rational to do so, though, or whether knowledge is possible.
Knowledge is not possible in an atheistic world view precisely because it is all just arbitrary assumption. An instinct doesn't make knowledge.
"There are many models we can make. The model - perhaps best described as a belief system or world view - can be used to help guide our actions, IF it is capable of making accurate predictions."
Sure, but we can never know if it will be for another day.
"Can you repeat your position?
As i understand it
1) the assumption of regularity is at odds with the world-view of a materialist.
2) it is a belief system (contrary to what someone else said)"
I would say, for 1), that regularity is completely unknowable to an atheist, so there can be no knowledge or reasoning, only assumption. And there is no way to distinguish assumptions that has anything to do with their actual truth.
You say your assumption is useful. OK, fine. I disagree that it is, if it's probably false, but even so, useful doesn't mean true, so you have no principled way to criticize anything anybody believes for the same reason. Maybe somebody else believes science is evil and should be shut down, and they find that belief useful to them. You have no way to criticize this. It's the same epistemology you use. So ultimately there is just no knowledge at all.