Out of Pod Experience

1. ) The answer would end up being something like greater then 1 but no greater then fifty (although it's probably less then that).

2.) I think you're asking what makes the fifty/fifty statistical split happen with a coin flip? I don't know, it just seems to be the way that existence is. We can chuck that question away into the closet along with all of the other questions that we can't answer, like "why does matter have inertia and where does it come from" and "something about magnets".



... I'll read Akita's post tomorrow, it's getting late for two full posts of text

That is completely backwards. True randomness means EXACTLY that all possible variations will be met. If you never hit one, with infinite tries, that would be the EXACT proof that it was NOT random. Expecting to see pattners over and over is EXACTLY what true random is. Pick any combination of heads/tails or 0/1 you want, even if it's a trillion heads in a row. Statistics can give you the odds of that happening and it WILL happen according to those odds if the individual coin flips are truely random. I don't know how you have this so entirely backwards in your head. That's what random means.

No you can't predict exactly when the pattern will show up. If it's a one in ten chance it could be on the first, or on the fifth, or the tenth, or not show up for a hundred tries, but (Law of Large Numbers already linked for you) the more times you flip the coin the closer to the odds you will get.






No binary code is not neccesarily any particular ratio of 1's and 0's. I never said it was. I said the data set would always grow closer and closer to perfect 50/50 as a whole. Remember the entire block of binary code is not 'infinity' it's just a string of coin flip results. The ratio of the set would be balanced by other results outside the binary code result.

@ Akita T


Try and look at it this way then. If you flip a coin 10 times you can expect 5 head flips on average, if you do it 1,000 times you are looking roughly 500, so on and so forth. This grows indefinitely, and no matter how big the number of flips gets, you are always looking at roughly a 50/50 split, each time.


Here is the thing though... as the number of flips grows, the distribution of heads to tails remains the same throughout. That is the problem with math telling us that you can flip heads 10,000 times if you kept flipping forever. The purely human invented theory of probability says yes, where as the realityof the universe say that heads and tails will always be more or less evenly distributed.



So you can get long sets of heads, and long sets of tails, but at some point you are going to peak in terms of the set length. That peak is probably something like <100, give or take fifty to one hundred, if you have infinite flips. I cannot say for sure, but if the coin is fair, no amount of flips will alter the tendency of relatively even distribution throughout. That is what math kids don't seem to be calculating when they submit mathematical proofs, infinite time will not change the natural pattern of distribution, of a fair coin, over any timeline of any size. And before you say "well can't you roll 10,000 heads and then have it followed by 10,000 tails" and that still be your 50:50 ratio? The answer is no, because that would defy the natural pattern observed in all coin tosses. There is a tendency for all lesser sets of 10, 50, 100, 1,000, 5,000 and so on and so forth, to have a more or less perfectly even split of heads and tails. It does not matter how many flips you have, if you cut a section out of the timeline 5,000 units will have a 50:50 distribution, and 100 units of that will have a 50:50 distribution as well. That is the pattern, and that pattern will reoccur forever. Kind of like how a fractal image can reoccur forever (again referencing a metaphor not to be taken literally)




No amount of time changes the pattern.
No statistical law alters the equally as important observable law of observable 50:50 distribution in (lets call them) data sets





Because your math proof is not factoring forces like these into the equation, it is thus fundamentally flawed and is giving you a false positive that could never be proven in reality. Ironically, because your theory can never be proven true or disproved as false, it is a very "safe theory". It does however violate what we can observe in the here and now regarding coin tosses. That should be enough for people to say "hey, this is just a theory" but since math is the new religion, those people end up saying "this must be true because my theoretical math says so" and observable reality be damned. Well IMO that's bad science, and Homie don't play that.

No... you're just selectively putting limits on the data set instead of infinity.

You flip coins (0/1, heads/tails) an infinite number of times you will have EXACTLY 50/50.
But that doesn't mean an endless parade of heads/tails/heads/tails/heads/tails, you will have streaks.

If your binary code is say... four thousand characters long... You check out how many possible different combinations there are of heads/tails in 4,000 coins. That number of different sets of possible results is X. Your chance to pull that exact string of code with four thousand flips is 1 in X. Even if you pick a combination like four thousand flips that are ALL heads, you chance to have that happen is 1/X (X possible outcomes, one toss.) just as surely as the chance to get heads on one coin toss is 1/2 (two possible outcomes, one toss)

The INFINITE data set as a whole will be still 50/50 however because for every time you get four thousand heads in a row, there will also be a set of four thousand tails in a row balancing it out somewhere else in the infinite data set. See: Law of Large Numbers.










This is NOT TRUE and I don't know why you keep repeating it in different words.

There are NOT an infinite number of things that are 'not Hamlet' there are only so many characters that can occupy any given spot and so there is not an infinite number of combinations. You have some specific chance of having that combination (Hamlet) come up, which is 1/X, with X being the total number of possible combinations of that number of characters.

If you put out X sets of that size, one will be Hamlet. True randomness means you can't predict which of the X sets will be Hamlet, and that Hamlet may not necessarily happen in every set of X, but the law of large numbers shows that the more times you bang out X sets, the closer to the exact ratio you get. Thus infinite tries means it will come up EXACTLY 1 per X times even if not neatly in order. Zero times in infinite tries is impossible.

No... there are not...

There are 2 keys on the type writer.
Hamlet is 2 characters long.
There are four possible combinations.

11
12
22
21

You chance of putting out Hamlet is 1/4.

Now adjust the numbers for the number of keys and how many characters long Hamlet is. That is X. Your chance of banging out hamlet every time you hit the keys randomly that many times is 1/X tries. Given infinite tries, it will happen in EXACTLY that ratio (law of large numbers.)

Your original post says infinite tries. Yes... if you don't have an infinite number of tries it is not inevitable, but that's not the question you asked.

A: YES

Your code is a 'non-infinite text'. (X characters long)
You have a specific chance for that to occur per try.
You have an infinite number of tries.

Within those infinite tries, it will occur in the exact ratio of it's chance.
Law of Large Numbers.

/thread

You cannot hope to have a real discussion about Chaos Theory if you don't understand how flipping coins works. You are attempting to run a marathon before you can crawl, and Akita T is daddy putting you back in your crib.

I will now cease.

If you actually read the above posts you would have seen that I never said otherwise. You will get head streaks to a certain size sure, but in practice you won't ever see 10,000 in a row. You would be hard pressed to see one hundred. it is foolish to instantly concluded that since you can sometimes get 10 heads in a row, you can also get 10,000. It does not work that way regardless of what a ultra-simplified equation on paper says.