Out of Pod Experience

So what does the gamblers fallacy mean to you Akita?




Because I think that it means that gamblers presume that after 10 heads in a row tails are bound to come up soon (tm).

In the next toss, or then next 2 or 3 tosses and yes it is common sense.
That is the point.


Thank you now we can move on although you did not explain how my example "had nothing to do with it". But I will let it go because I know how painful admitting that you are wrong can be for some people

The point is that "gambler's fallacy" was NOT relevant in ANY way in your example.

Those first 9 heads tosses did NOT make the 10th toss more likely to come up tails, it was a 50:50 shot.
The 10th head did not make the 11th toss more likely to come up tails, it was still a 50:50 shot.
And so on and so forth.

Your gambler did NOT rely on the assumption that any of the subsequent tosses would be more likely to be tails rather than heads.
THAT would have been gambler's fallacy.

Instead, your gambler relied only on the fact that he expects a tail to come up EVENTUALLY, and used his bets to hedge against the very good chance of losing many subsequent bets.
That's the martingale betting system. And it does not rely on the chances of a toss becoming different from 50:50.
In fact, it relies on the precise fact that the chances REMAIN 50:50 regardless of past tosses. That's the exact opposite of gambler's fallacy.

And how exactly would any of that have changed what he won ?

You don't need to wait to see a string of heads or tails to use that exact system to always win.
It's actually completely irrelevant whether the previous 20 tosses would have all been heads or all been tails.
He ALWAYS ends up with a gain of 100 whenever a tails comes up, REGARDLESS of how many heads get thrown up beforehand.
In fact, it even works if you're working with biased coins that come up heads more often than tails !!!

All you need is enough money to be able to keep doubling your bets and a house that will match those bets fully.
Casinos have limits to the bets precisely because of this, to combat the martingale betting system being used by people with huge bankrolls.

Except that you have failed to demonstrate anything.

The martingale betting system and the gambler's fallacy are completely separate from each-other.
You don't need to be a delusional sufferer of the gambler's fallacy to employ the martingale betting system to win.
And even if you know that the gambler's fallacy is wrong, you can still use the martingale betting system to win - in fact, you will win money faster because you won't have to wait for streaks to occur to start making bets.
Edit: of course, you can't use it in real life to actually win in either case because you don't have an infinite amount of money nor do casinos allow you unlimited bets, so you will eventually lose all you can invest on an unlucky streak.

Now try to make money using ONLY the gambler's fallacy - you are allowed to only place fixed amount bets.
Do you think your chances are better than even to gain money rather than lose if you only bet against streaks ?

http://books.google.ro/books?id=QEnjzV0rTvwC&pg=PR4&lpg=PR4&dq=978-0-618-62011-1&source=bl&ots=3pSUf9fzNY&sig=L0o2hPWU3QfoXrWSqFZWLe5OT7Y&hl=en&sa=X&ei=tXUeUJq9CYySswasnIGQCQ&redir_esc=y#v=onepage&q=monte%20carlo&f=false
==
http://tinyurl.com/c758dqa



Note that a roulette wheel is less than 50% likely to land on either red or black. French roulette has a single "house auto-wins" field out of 37 (a green 0), whereas USA has two (a green 0 and a green 00) out of 38.
A non-USA roulette has a 48.64(8648)% chance of landing on either red or black.

The chance to get 26 black in a row would be roughly 1:136,823,184.
That's roughly similar to the chance of a perfect coin to flip 27 heads in a row (1:134,217,728).

Let me repeat that...

The casino ended up making millions of francs.
One 1913 franc was worth roughly three 2007 euros.

That's just the most famous early case happening in THE SINGLE MOST FAMOUS CASINO IN THE WORLD with a wheel that was above suspicion of tampering where people made significant bets against the streak and the casino won big time.
It's not necessarily the longest streak ever recorded, and casinos should be reluctant to publicize such results out of their own volition due to fears of accusations of tampering, so you have to rely on data from just the people present there.

Streaks of 16 or higher of a colour on american-style roulettes (~47.3684 base chance, roughly 1:155660 for streak of 16 red/black) happen all the time and are witnessed by many people constantly.
Streaks of 20 to 25 happen less often in large casinos, but it's not unheard of at all, with employees occasionally talking about this or that table having spun that many reds or blacks in a row in the not so far past.

Roulette happens at a much slower pace, you could have as little as 40 spins per hour, with not many tables averaging over 50 spins per hour. And it's not open to the public 24/7 like other regular casino attractions, because the wheel is usually recalibrated and separators exchanged on a daily basis to avoid consistent biases which could give long-time observers any advantages.
Roulette is also far less popular than many other casino games (especially in the USA with their double green fields, very few casinos in the USA have european-sytle roulettes, and when they do, it's usually reserved for high-rollers only, with much higher minimum bids) and as such, the number of tables is not very high per casino to begin with.
For instance, Caesar's Palace in Las Vegas (one of the most famous and largest casinos in the world) is one of the very few Vegas casinos that offers european-style roulettes, and they only have 18 roulette tables total today.

Right back at you - what's so complicated to comprehend that believing the next coin will be more or less likely to end up heads or tails has nothing to do with making money when employing the martingale betting system ?

Look, here's a simple one.
http://www.random.org/cgi-bin/randbyte?nbytes=16384&format=b
That's 2^16=16384 random bits right there.

ON AVERAGE when loading of that page a good number of times, I expect to see very roughly one streak of 10 ones, three streaks of 9 ones, eight streaks of 8 ones, eighteen streaks of 7 ones, fourty-two streaks of 6 ones.
I also expect to see a streak of 11 ones every other page load, a streak of 12 ones every three loads or so, and a streak of 13 ones every seven loads or thereabouts.

How many do you think I'll actually get if not roughly that many ?


You did ? Weird. Because up until now it seemed like you totally DIDN'T.
Let's say that you did and it was all just a huge multi-page misunderstanding.


You need to be more clear about what those "additional hypotheses" might be.
Start off with how they are even necessary in the first place - as in, WHICH PARTS of statistics are "incomplete" and need additional details to paint the "whole picture".
Then immediately follow with a concise explanation of what those hypotheses are, what predictions do they make, and how we could test them.

So tell me what general relativity has any business in being injected into statistical theory, and how is that supposed to happen.

Because all I can see so far in your past posts are vague pseudoscientifical concepts of questionable practical utility even if they would be actually true.
Care to be a bit more specific and pile on the PRACTICAL / TESTABLE aspects of your hypotheses ?