This bears out in the program you wrote. If we subtract the combinations aaax, where there is an exaggeration of a’s, there will be a corresponding exaggeration of b’s.
This means, necessarily, that if you have 3 consecutive heads, and that is all you know, all you have flipped, your total distribution is an exaggeration of heads. The next flip is likely to balance that exageration out. In a single instance of 4 flips, the margin of error for that to occur is likely to be large, but in 1 million consecutive aaax combinations, we would expect to find pretty much x=b every time, with a very small margin of error.
Motor told me up front, there’s no evidence I could show him that would change his mind. At the time, I thought that was very, very fucking idiotic.
I’ve come to appreciate it now, after spending an evening writing programs to show you evidence, just to find out there’s no evidence that would change your mind either. At least Motor didn’t waste my time.
We could all learn something from the introspective ability of someone like motor. “I’m a brick wall and I’m proud of it”. Good for you motor, good for you.
Well if you hadn’t written a shitty program.
Put your money where your mouth is and show me what a good one looks like. Something that directly tests the claim you made. Can’t wait to review it.
That was my point.
If the logic can’t be seen - authority rarely works to convince.
No matter how much a dog respects his owner - he still can’t see color.
See what I mean?
Because it is impossible to get to - forever.
I haven’t seen that he is - but hope is eternal – 
Hold on, I’m cutting up my tobacco, I’ll be with you all in a second.
None of that is true.
An exaggeration of a’s does not imply an exaggeration of b’s later.
A.) humans can’t program randomness with code.
B.) because we know the code isn’t random, cheating is assumed on runs.
C.) the code is written to not allow runs (not totally random)
I gotta learn to write C++ to show you why you designed an experiment poorly?
Your programming by itself is excellent, there is no doubt of that, I think we were all suitably impressed.
What is wrong with it is the design.
If you are attempting to determine the probability of x being b in a distribution of aaax, and you include in that distribution a random set of a’s and b’s with an overabundance of b’s, that is a poorly designed experiment.
Origami is right about that part.
The problem is that principle doesn’t apply to his scenario - but he can’t see why not.
He is trying to guess about a single toss - yet his reasoning is about the average of many tosses (nothing to do with his “next toss” issue).
Origami, If you were capable of being convinced by evidence, you would be. You would at least be able to say “I’m not sure you’re right, but you’ve definitely put doubt into my position”.
I’m beyond caring what you think. If you’re immune to evidence, then who cares what you think?
Programming something up yourself would at least give you the opportunity to restore some integrity.
FJ -
If you program it like I said - precisely - I suspect he will see the problem (probably won’t understand why though).
This just sounds like a cop out, a bizarre goalpost move
How did you say?
@obsrvr
The probabilit of the next toss is affected by the distribution of results for previous tosses within the same set. One thing follows from the other.
Consider that you are not saying what is wrong about it, you are just stating that it is wrong.
That’s not the way it works flannel.
Randomness allowed for an infinite number of beings to hit the jackpot played by an infinite number of people…
An infinite number also lose.
Let’s get philosophical here for a moment, number theory is all I care about anymore…
Reality was hacked.
What are the odds that nobody hits an infinite number of heads forever?
0%
At least… if it weren’t random.
You are whining that my data set includes a bunch of random data, when the entire point of the experiment is to test what happens in a random coin flip. Of course it includes random data. It’s fucking random.