Out of Pod Experience

Before you can presume randomness you must first prove that it exists. Does true randomness exist? Well I don’t think people have come to a consensus on that fact yet. If there is no consensus then how can we come to a consensus that any outcome is truly random? If it is not truly random then what is actually happening? You see how this works?



Relativity of the observer is also a very important concept. Everyone keeps pointing to the odds of the coin at the moment it was tossed, but what about before it was tossed? What are the odds of generating a desirable sequence before the sequence is generated? It seems like a contradiction only when you do not include ideas involving relativity. The coin can have a 50:50 static probability but a string of tosses no longer has a static 50:50 probability. The string has a different probability of occurring relative to the observer and before the coin was ever tossed.

So far, so good. Nothing objectionable, but nothing new either.


Actually, as more flips are made, his outcome becomes
a) either immediately completely impossible altogether (if any of the flips are different to the flip he picked for that particular spot, he can no longer get the desired outcome regardless of future throws)
b) or increasingly likely otherwise (because he has to match less and less flips, as the ones so far already matched)
Small inaccuracy, but, eh, ok, let's call it an oversight not a mistake.


That's more philosophy than science, but ok, not that you really place an overwhelming amount of importance on it yourself either, so, eh, moving on.


I don't see anybody disagreeing that the odds are 1:2^12 against in the case of 12 tosses - not me, not you, not anybody else. Everybody perfectly agrees on that. There's nothing contentious about it. Moving on.


And what point are you actually trying to make exactly other than use a lot or words to say... well, what are you even trying to say anyway ?
What does any of that have to do with anything ? And especially what does that have to do with the issue at hand in a fashion not already explained just fine by established statistical theory (when properly understood) ?

Which is something you had roughly 1:294 odds AGAINST witnessing in the first 7 pages, and a few people might go for over 5k refreshes without ever witnessing it.
Yet you were lucky enough to stumble upon it.
Or, in the case of you starting to bet against the streak by the 10th spot and doubling your bets, a very, very unlucky break for you, ending with you going broke quite fast (unless you had at least 26,214,300 dollars available when you started betting 100 dollars in round 10, and would have been allowed to bet your post-round-25 remaining 13,107,200 dollars on round 26).

1. Considering the odds are 1:33,554,432 against it happening starting from the very first spot, that's nothing unusual at all.

2. Again, nothing unusual - you expect to see on average very roughly twice as many streaks of length X than streaks of length X+1, so you're not observing anything on top of what just statistics already says is likely to happen.

3. That's fairly easily explained by the fact that you expect to get 15 ones in a row more than a thousand times more often than 25 ones in a row, and to get 35 ones in a row more than a thousand times more often than 45 ones in a row, and so on and so forth.
It's only a matter of how much patience or computer cycles you can bear sparing for the pursuit of simulating anything like that.
Why bother wasting a thousand times more computer cycles to see a 35-streak on something you have already observed to happen pretty much as expected so far up to 25-streak and you have no reason to expect it would happen significantly differently later ? And if you get to 35 for some reason, why bother doing it more than a thousand times more just to get to 45-ish ? That would be one million times more than was already pretty satisfactory. Where exactly would you stop ?
It's not like you could keep a lot of supercomputer time busy with such trivial calculations. If you do want supercomputer time, you'd better come up with something more useful than doing an empirical verification of already well-established basics of statistics which don't even really NEED a practical test because inference will suffice just fine too.

Also, let me see your sources that claim not just 25-30 streaks of heads/tails, red/black or similar, but also 25-30 letters of Shakespearian text too. In case it's not clear enough, I've just called you a liar.
And I've called you a liar because while the upper estimate of 30 heads in a row is "only" very roughly 1:1,000,000,000 against, the lower estimate of 25 characters of Shakespeare in a row is something closer to 1:1,000,000,000,000,000,000,000,000,000 against (give or take roughly half an order of magnitude), and pretty much in the ballpark of the chance for 96 (plus-minus two) heads or tails in a row.


That's because the expected results for one page load of 2^14 bits are on average very roughly somewhere around one streak of 10 ones, three streaks of 9 ones, eight streaks of 8 ones, eighteen streaks of 7 ones, but only getting one streak of 11 ones roughly every 2 loads, one streak of 12 ones roughly every 7 loads and so on.
Which is pretty much what I actually predicted.
A much more relevant and important thing you should have counted would have been the exact total frequency of occurrence of streaks of various lengths.


ANY streak is impossible to predict when it actually happens, you can only say roughly on average what you'd get if you'd keep at it for a very long time.
Any individual throw is also impossible to predict, but in the long run you can say that roughly on average you'd probably get close to even percentages of heads and tails
That's the whole point about randomness in the first place. If you could actually predict it precisely, it would no longer be random. You can only make statements of chance to get something, not certainty of getting it at any precise moment.


And that "somewhat predictable way" is pretty much exactly what statistics actually describes. It needs no additional hypotheses. It already has everything it needs.

It's NOT less likely to happen at the beginning.
It's just as unlikely to happen starting at the very beginning or starting at any other particular place.

The sequence "011011010111" is just as equally unlikely to happen at the very beginning as "111111111111" or "101011010111" or any other sequence of length 12.
But one of them HAS to happen first either way.
You just don't know which.

Don't believe it ? TRY IT YOURSELF THEN.
Pick any particular combo of length 12 you consider "plausible" before going to the random bits page.
Refresh the page until that EXACT sequence (the one you picked before going there) actually shows up at the very start.
Not a sequence that looks like it, but that exact sequence, precisely the way you picked it BEFORE going to that page.
Now tell me how often that actually happened.
Here's a hint : it should be about as often as a row of twelve ones in a row at the very start - that is, horribly low chances of actually happening.

If you draw sufficient random bits however, EVENTUALLY, all of those sequences of 12 bits are LIKELY to happen, and since twelve ones in a row is just one of those many possible sequences, it's also obviously likely to eventually happen.
What's not surprising at all is that one SPECIFIC sequence you picked does NOT OFTEN happen in the VERY FIRST place it can possibly happen.

You keep repeating that as if it's some kind of victory, when it's totally meaningless, as nobody in this thread ever disputed it in any way, shape or form, and very few in other threads had the lack of attention to claim anything of the sort.
Statistics can't predict the future with absolute certainty, neither can you, and for that matter nobody else can either.
Nothing ever has exactly a 0% or exactly a 100% chance of happening in the future when randomness is involved, unless you want to go into "infinite tries" territory, and there it can get pretty iffy (and we don't need to go there because we're talking very finite things and pretty specific chances).

I never claimed I can predict the future with ABSOLUTE CERTAINTY.


Which I never did.


"MAY never happen" is not the same as "WILL never happen".

First, you bring literal infinity into the mix, by using the word "never" with regards to future events.
Second, you recognize that that specific something COULD happen.
Combine these two and what do you get ?
Yes, there is a chance that it may NOT happen up to ANY particular point in time, but that chance gets closer and closer to zero the more sets of tries you undergo.
So, in the end, while you could say that "it might never happen", that chance of it not happening has almost no chance whatsoever itself, which renders that phrasing as close to completely meaningless as you can possibly get.

In the end, there's only three meaningful things you can say: either "it can happen" (and add a chance of it happening if you like), or "it will never happen" (which implicitly means absolutely zero chance of it happening in any given trial) or "it will always happen" (which implicitly means 100% chance of it happening in every trial).
"May never happen" is not meaningful, it's meaningless.

He corrects people on the internet

facepalm...

I came to this thread expecting something smart and well thought out, and I found this.
I am not great at math but last I checked coins were cylindrical and 3 dimensional. No statistician is going to say, nay; insist that a coin toss in a vacuum is a 50/50. It's not because coins are not flat they are 3 dimensional.
It doesn't take an autistic polymath to see that your argument is at its foundation built on sand and mud.

Go home everybody, nothing to see here.

So thats it then? You are just going to ignore the fact that you are wrong and your argument is built on fallacy?

The only way a coin toss could be 50/50 is if it was in a vacuum, 2 dimensional and without any markings on either side to unbalance it. So even if you amended your original post to account for this lapse in reasoning you would still only be talking about hypothetical coins and not actual coins. What you think you are referencing is a "fair coin" but once again a "fair coin" is not something that can exist in reality.

Seriously, Get on my level bro. I know **** squat about math but I still know enough to realize that your a terrible troll at his best.