TurboGenius The Fallacy of Gambler's Fallacy - 2020

Certainly one of the more interesting topics in roulette is that of the
"Gambler's Fallacy".. ie. "Something is Due based on past spins"

Let's run off an example, because it so easy to prove that this "Fallacy" is no
fallacy at all but is instead a predictable event.

Using a 00 wheel -
After 114 spins we have 5 numbers that haven't appeared.
AS A GROUP - are they now expected to appear "above average", "catch up", are
they "due" ?

I can make a prediction (based on past spins) that this group of 5 as a group will
NOT have 0 shows during future 38 spin cycles - meaning I can predict a win based
on past spins alone playing only 5 numbers. How does it work out ?

Through the next 38 spins (1 complete cycle) we have
4 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
4 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
2 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
5 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
3 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
7 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
4 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
7 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
2 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
6 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
5 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
7 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
8 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
5 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
3 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
5 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
7 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
8 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
3 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
3 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
6 appearances of this group of 5

Through the next 38 spins (1 complete cycle) we have
1 appearance of this group of 5

838 played spins and not once was I wrong, every cycle of 38 spins
had a win on this set of 5 numbers, as compared to the qualification
spins at the beginning (114 spins) where this group did not appear once.

The amount of wins per cycle based on these predicted numbers :

1 win - 1 (below expected)
2 wins - 2 (below expected)
3 wins - 4 (below expected)
4 wins - 3 (below expected)
5 wins - 4 (expected)
6 wins - 2 (above expected)
7 wins - 4 (above expected)
8 wins - 2 (above expected)

Below expected = 10 cycles
At expected = 4 cycles
Above expected = 8 cycles
-- or better put --
Below expected = 10 cycles
At or Above expected = 12 cycles

Never a 0 win cycle, 100% accuracy in prediction.
This prediction and 22 continuous winning cycles are based only
on the past spins and the so-called fallacy.

 

Fascinating and plausible as usual. This is one of the better threads that you ever started.

So if you went 3 cycles to discover the 5 missing numbers then why would you not throw out the first cycle and replace it with each new cycle after it has occurred, thus finding a new 5 or so missing numbers after the new 114 spins? Just a curiosity point here. It looks like you used the same first 5 for all 22 cycle tests. That would imply that you could take any 5 numbers and get the same result if you could predict which 5 were not going to hit in the first 114 spins. We all know that you can't. You do realize that all you need is a progression that pays everything back by the end of 38 spins. That part of this is a no brainer. It's fascinating in this small sample that you never had a 38 spin cycle that never had a zero appearance of the 5 selected numbers.

 

I might power test this with a sim. It's so easy that any idiot can test it.

 

I wonder what would happen if you took the three hottest numbers from the past 114 spins and ran the same test? That too is a fallacy based on past spins.

 

Test run until all number appeared. (171 spins)
The results of the 3 hottest numbers :

Through the next 38 spins (1 complete cycle) we have
1 appearance of this group of 3

Through the next 38 spins (1 complete cycle) we have
1 appearance of this group of 3

Through the next 38 spins (1 complete cycle) we have
4 appearances of this group of 3

Through the next 38 spins (1 complete cycle) we have
5 appearances of this group of 3

Through the next 38 spins (1 complete cycle) we have
3 appearances of this group of 3

Through the next 38 spins (1 complete cycle) we have
3 appearances of this group of 3

Through the next 38 spins (1 complete cycle) we have
3 appearances of this group of 3

Through the next 38 spins (1 complete cycle) we have
2 appearances of this group of 3

Through the next 38 spins (1 complete cycle) we have
1 appearance of this group of 3

Through the next 38 spins (1 complete cycle) we have
2 appearances of this group of 3

Through the next 38 spins (1 complete cycle) we have
0 appearances of this group of 3 (end)

=========================================

After only 10 cycles we had a cycle without a single win.

0 wins - 1 (prediction failure)
1 win - 3 (below expected)
2 wins - 2 (below expected)
3 wins - 3 (expected)
4 wins - 1 (above expected)
5 wins - 1 (above expected)

Below expected = 6 cycles
At expected = 3 cycles
Above expected = 2 cycles
-- or better put --
Below expected = 6 cycles
At or Above expected = 5 cycles

Unlike the previous post - the results now favor a result below expected.
Also we have a failure with a 0 win cycle that ended the test.

 

Numbers are never due to catch back up. The expectation moving forward is for a number to hit no more/less frequently than expectation.
Consequently every number is expected to fall farther and farther below what is required to break even.

 

Thanks for checking. I expected it to be worse because it only used 3 numbers. But 5 hottest might be better.

 

Ha Ha, that's a fallacy too.

 

No, that's common sense and basic probability.

There are 38 pockets. The payoff is 35 to 1 (36 for 1)
Because the house payoff is short, the longer you play, the more you will lose.

 

Funny part is that in "The richest man of Babylon", ages ago, that guy allready understood that if probabilities are not in your favour, it is worthless to put time and money in it... Still, ages later, a lot of people haven't understood that... It confirms one of the theory about the evolution of the mass that is very weak.

 

That is factual but to a point.
According to the math, you could win flat betting over 20,000+ spins even with the house edge
being there, even with the imbalance of payout to odds of winning.
This has been tested and proven and I can do so again if you'd like.
Now that would be 66 casino visits with 300 spins each to reach 20,000 spins - considerably "long term".
And the player can certainly negate the small house edge by being on the hottest number flat betting.
The house edge will though come into play after a very very very long time and at such time even
the "hottest" number wouldn't be able to produce enough wins to defeat it - BUT... that is in an
incredibly long time, and there's nothing from stopping the player from "resetting" and beginning again
with the exact same results for the exact same number of spins, hence "never losing".
The house edge and imbalance in payout is held up as the monster we can't win against when in
reality it is nothing.

 

You're talking about survivorship bias where you only taking an extreme example to validate your theory but in real and more likely you'll lose your pantie.

But you should look at one thing, i could also resist as you do by playing the same number of number randomly as Jerome soft shows, from that point what's the use to play your hot system.

Everything is a gauss curve finally where you take only extreme positive example to validate your theory, besides what it says is that it is just resisting, if you spent x hours to win peanuts at the end there is absolutely not interest to do that..

 

Flat betting $25.00 on a single number for 20,000 spins :

+880 units....
I technically could lose 880 spins and STILL be much better than the house expected "edge" dictates.




The house edge isn't hard to defeat, however like I said above - if you don't use a method or system
or strategy that works - over enough spins even the hottest number can't win against the tiny bites
that the house edge has.
You can however just start a new session... 20,000 spins more....880 units more....
there's no secret "rule" that this isn't possible, just because the casino pays less than the chance of
a number appearing.

 

An $8000 draw down? Seriously???

After 20,000 spins some numbers will still be above break even. So cherry picking one number when you already know the best numbers is nothing significant. What about the other 37 numbers???

The house edge is very real, and the progression that you've posted above isn't going to even make a dent in it.

 

Turbo,

Hot numbers won't even make a dent in the house edge. The only way to win with hot numbers is if those hot numbers are biased due to physical problems with the wheel. Again, bias numbers may overcome the house edge, but random hot numbers will not.

 

You could bet also during one million spins and say the same thing, is that you proof you got? Because you could also say that you can win on hot colour, hot dozen, column by using the same demonstration...

 

There is no progression, it is flat betting.
My post is clear.
This isn't about a bias wheel situation.

Yes, seriously. A drawdown of any size is irrelevant to testing when the end result is positive.
For example - someone invested $20,000 in a stock that went up and down and down and down and up for
30 years and now sits with a value of $200,000. The drawdown isn't relevant to the final outcome.
Everyone knows this.

As I said, no single number being bet on flat for that many spins could win unless you changed where you played based on a
predefined condition (a system). The house edge eats away slowly so that after multiple tens of thousands of spins -
you wouldn't be able to stay positive. A system or method of picking/predicting where to bet resolves this.

And again, this isn't the other section - posts not relevant to the discussion will be moved.

 

The topic is about Gambler's Fallacy.
Thanks for reading.

 

I agree that regression to the mean seems to occur. But gamblers fallacy is actually a statement about a particular event. If you claim that the next 38 pins MUST contain that group of 5 then that would be a fallacy. But if you make a general statement that the group of 5 would tend toward the mean expected value then that would be correct.

If a dozen sleeps for 10 spins it does not mean it MUST appear in the next 10 spins. It might sleep for another 20 spins.

You can not use past events to predict the outcome of independent spins.

You know all this turbo, or are you claiming you KNOW which numbers MUST appear next?

 

Anti fallacy statement from me: " Every number is due every spin".
Well, it is a bit tough to understand with roulette may be, let us talk of a coin flip. In a fair coin, both sides are equally likely but randomness will not let HTHTHTHTHTHT happen mostly and usually, they will come in all possible fashions. Since both sides are equally likely in all flips, if we get a sequence like HHHHHHHHT, it is safer to presume that now Tail should behave relatively better than earlier since it was too harsh earlier.
This is no fallacy but a statistical reality, called Regression towards Mean. To make it fallacious, one should predict perfect hit rates or better than average hit rate or compensatory wins. Both concepts may look going parallel but they are'nt.
Probability of 5 unhit number in 114 spins of roulette is almost at the brink of virtual limits. As per the RTM principles, it could be safely assumed that at least one of the 5 of these numbers should show or most likely show in the next 38 spins cycle.
Ironically, even if one of the 5 left numbers are guaranteed to show within nexts 38 spins, we can not guarantee any profit out of it, playing with any sane progression possible.