TurboGenius When does it lose ?

@ Turbo, yes I did see the post where you described how you played, and I watched the video, but didn't see the post where you posted the profit (actually, I still can't see it in this thread).

But this isn't the number of bets, right? Reason I'm asking is that in order to do a hypothesis test I need the number of bets made and also how many numbers you bet per spin (which varied, but an average will be ok). Thanks.

Edit: Actually I think you mentioned somewhere that you were betting every spin. TBH It's not entirely clear how you were playing from watching the video. You started off with one chip on each no. in a dozen, but you only showed the one dozen.

 

MJ, it's not difficult to rationalise this strategy has no edge. It wins until random throw a Mike Tyson left hook to take back all and more, koed.

Do you even consider how many people over the last 200years have tested sleepers and they somehow missed this method?

 

Thats why you never play the last number, dare I'd say even last 2 numbers. You can run your test again but this time run it until you have 1 left, whats the max it will show you.

 

If it's so hard that it's not discovered by anyone in the last past 200 years, including scientists, mathematicians, programmers, etc., I think a few hints won't kill anyone. At least we find out if there is any way to select better bets for better hit rates. I've explored any possible way to select bets, and I'm putting aside the whole concept of bet selection. Besides temporary lucky sessions, I couldn't score any better hit rate than 2.70% per number in the long run.

I kinda enjoyed what TG shared, because it was what I was testing around 5 years ago. I got a bit more confident in my progress. At least I know what doesn't work, without even getting into testings again. I know a million of them. However, it is frustrating at the same time to give up on a very fun journey.

 

Since I know what the results look like (dispersed as fuck, not reliable to play based on), I can't find enough motivation to investigate this concept again. I might be wrong, but I'm not, lol.

 

These quotes of you can be interpreted in a million ways. How could these sentences possibly lead to a conclusion?

You probably are referring to the randomness of random trials. In Roulette, most of the time the LotT is applied to 37-spin cycles. The ones that don't comply with the so-called law, are few, but enough to bust any RTM player.
I'm reading your quotes in reverse:

No, the game cannot have a natural weakness, because no matter how untold the secret is, it will ruin the game eventually, and 200 years is enough for this event. The game is there. We are still cat-fighting here

No, because of busts caused by the cycles not complying with the LotT.

Do you mean that some kind of work is done to keep the game unpredictable? That is the definition of Entropy Maximization. I don't know whether it's necessary to be done by the security experts or the randomness is sufficient for this matter.

Betting against predictability... that makes the game predictable twice! Again, the entropy cannot decrease in an isolated system with work being done to maximize Entropy. You asked fewer questions to find the correct answers = Entropy decreased... [Confused]

 

I like this from Bombus on another thread:

It's a fruitless task trying to find a mathematical justification for systems, because there just aren't any. How can there be? But it's a good idea to go deeply into the math because then you'll understand why systems can't work. I mean real math, not the hand-waving that Turbo, TwoUp and others do. Then if you still want to use systems, go ahead. At least then you'll be using them in the full knowledge that they have no mathematical support, and you'll be aware of what's likely to happen.

 

We share a lot of agreements, but I disagree with your statement about math being useless in system design.

There is a paradox in utilizing progressions: By default, 1/37 hit rate is not enough to win, no matter how much you are betting each spin. The reason is when you categorize all the bets you have placed (or you might place) in a session, the house edge is applied to all of them. On flat bets, bets of 2 units, bets of 10 units, etc. But I believe timing makes a difference. Placing 2 units to win some of the losses back, and not to keep doing that, makes a difference. Also, knowing the limits of variance is useful. Playing based on a system that expects 1/37 (SYSTEM A) is a confirmed loser because you can easily witness an awful session of 0/250, or 1/213. But what if you design a system that expects a very very lower hit rate, like 1/80 (SYSTEM B)?

When our hit rate is 0/240, SYSTEM A has lost 6 levels of progression and playing the 7th, but SYSTEM B is 3 sessions behind and going to start its 4th level.

I think we can win the game by expecting less than what it typically offers, or offers in the long run. This is all based on strategic approaches to Roulette, like progressions and stuff. It's a whole other world when it comes to bet selection.

 

Here's the thing.

EVERYONE understand roulette outcomes as hotties, sleepers and normal.

EVERYONE found nothing. Zilch Nada.

There must be a change.
A change of the fundamental principle and/or DEFINITION.

This is where I first stumbled upon entropy.
Shannon entropy.
There's the entropy graph.

Then there's birthday paradox math.
And the probability graph.

I'm sure the characteristic of both these graphs is pretty clear.

I tested it with hotties, sleepers and normal.

It does not show this same graph.

Hmmm.....

Where have i gone wrong?
What am I missing?

Lets relook at birthday paradox math again.
23 people, 365 birthdates and the probability for a match.
That's it, so simple.

What about roulette outcomes?

"x" spins, 37 pockets and the probability for a match.
Why wont it work???

I must be missing something.
What's my mistake???

Well, the rest of the story, I found the silly miss-[take].
Wrote it on forum too.
Read CAREFULLY.

You ask for hint.

Here's the most important one.
It's not even new.
I already wrote it on forum.

Change the DEFINITION.
(Same definition same result.)

Get rid of hotties, sleepers and normal.

Think instead of saying I know this and that.
It must be what EVERYONE don't think about.
And its a simple change in definition(classification).

 

Not going to happen, too many other things will happen first.

A number sleeping for 757 spins already exceeds a 1 in a billion event.

In a 2005 paper published in Nature, as part of their investigation into global catastrophic risks, MIT physicist Max Tegmark and Oxford philosopher Nick Bostrom calculate the natural risks of the destruction of the Earth at less than 1x10^-9 per year from all natural (i.e. non-anthropogenic) events, including a transition to a lower vacuum state.

That lower vaccum state is the possibility of space time itself collapsing and taking everything with it.

So given a number sleeping for 757 spins has a similar probability as for the end of the world so we shouldn't worry too much.

A number sleeping for 200k spins is less than 1x10^-317. Basically a decimal point then 317 zeros.

The smallest possible measurement in the universe is the Planck length and that's 1.616255×10^-35.

So 317 zeros vs 35 zeros for the smallest possible measurement in the universe.

So let's work out how long we would have to wait to see this event of a number sleeping for 200k spins.

Well there are less than 10,000 casinos in the world and I'll assume every casino has 100 wheels (which they don't) just to give ample headroom and 45 spins per hour.
I'll assume every wheel is spinning all the time, (again an overestimate). That results in less than 400 billion spins per year. Still a huge overestimate but let's go with it. 400 billion is less than 1x10^12. So we can take 12 off 317, and it would still take 1x10^305 years.

So what will happen in 1x10^305 years?

Well the universe is about 13.8 billion years old 1.3x10^10 years old. So let's take 9 off 305 to scale our timeframe to billions of years. So we only have 1x10^296 billion years to wait.

In about another 6 billion years the sun will explode, destroying the earth and become a tiny white dwarf then once the universe is trillion years old (1x10^12 years) the sun will have become a black dwarf, all stars in the universe have died and ceased fusion and no new stars are being formed in the universe. So we can take another 2 from 1x10^296 to get 1x10^294 billion years left to go.

Ok from here, once the universe is about a quadrillion years old 1x10^15 all solar systems in the universe cease to exist, having disintegrated or dispersed.

So we can take another 3 off 294 and only have 1x10^291 billion years to go.

Eventually galaxies cease to exist or have been consumed by black holes.

Then after a long while all the protons in the universe have decayed so we take 1x10^30 off, leaving 1x10^250 billion years. All matter created in the universe by stars is gone and we enter the black hole erea as that is the only kind of matter left in the universe.

The universe will be 1x10^100 years old after the black hole era ends when all black holes have evaporated, so we can take about 85 off our 291 to only have 1x10^206 billion years left.

We get about 1/3 of the way through the dark erea, the long decline into the heath death of the universe with maximum entropy/disorder.

The universe will be about 1x10^1000 years old at the heat death while we only needed 1x10^305 years when finally we would theoretically see a number not appear in 200k spins with every imaginary wheel made and yet to be made running all that time.

Let me repeat. Not going to happen, physically impossible to happen.

Again not going to happen.

Again not going to happen.

 

You got it wrong.

No number will ever sleep for 200k spins. That stat is for all the numbers to catch up to 2.70% (1/37) hit ratio at least once. It means in that specific test, I had one number that had a hit rate from zero to 2.699% until that last hit came in and rounded up the hit rate to 2.70%.

It wasn't about sleeping. It was about catching up to the 1/37 ratio overall. That specific number probably had around 5380 hits until that last hit.

 

Also, here is another side note: If I used my computer's RNG to simulate a test and a number slept for up to 757 spins, the probability of it happening to me is so small that can be ignored totally. But I didn't need a million galaxies to witness one, it took me 5 minutes of coding. The RNG of my computer made that happen. Any other RNG can reproduce this result, but the probability suggests ignoring it. Actually, you can act as it has never happened, lol.

We are on the same page, but we needed to adjust our wordings.

 

A number sleeping for 1000 spins is happening all the time. I've run simulations with millions of spins and these sleepers will pop up from time to time (on average) exactly every x amount of spins. (and again exactly what the math says it should)

So lets say I picked these long sleepers (just for sake of argument) is it better to bet on them than to pick a random number?

 

Let me interfere here, please. I would say a number slept for the last 1000 spins doesn't have a higher probability of hitting, but has a higher probability of repeating with less distance than 37 spins. I think @Luckyfella agrees with me on this.

 

Isnt that the same thing? Betting a random number for 37 spins or betting the (1000s) sleeper for 37 spins? So you would say that the sleeper will repeat sooner in a cycle than the random number?

 

You can't possibly valuate to any degree a formula for coincidence. A hottest number can change to a coldest number at any time. It can also go back to continuing to be a hottest number for a while longer. Now when these hottest numbers change other numbers become the desired targets of any hot number hunter system. Forget the long term. You must find a way to come off a cooling number, take the loss, or find a better acting number. It's 100% bet selection. If it is part of a progression of any kind then you are bracketed in. You must find a winner before it is too late. It's do or die gambling. I'm experimenting with a common Marti with single hot number hunting. I've already gone past my $3,000 point in winnings and know that I can go a little over 120 spins without a hit before going bust back to zero start point. So my goal is to learn what a common interval is between hits. How hard is it to select a number displaying excellent activity? Why stay on a number past 37 spins if other numbers are waking up? It's a guess. Any rule that works in most cases still has that sequence of death lurking somewhere. It's just a game to see if you can solve it.

 

For something to repeat, you first need it to hit. A metaphoric example:

If you tell me that number 9 has been sleeping for 1000 spins, and after that, you start witnessing the next 200 spins, and then you tell me I just saw a weird shit happening on my table, bro! A number popped up 15 times in that 200 spins, what number do you think it was bro ?

I have 37 numbers to choose from. I'd answer: I dunna, 9?

 

The math says it's a 1 in a billion event for 757 spins so it's not likely to be witnessed in your lifetime of play, in fact if you played 24x7x365 days a year you would have had to started 500 years before Christ as it's 2,536 years of continuous play at 45 spins an hour and still may not even see it.

Might be an artefact of your RNG. They are not as reliable as people think, there are plenty of issues that could produce statistically invalid results.

Honestly I don't think chasing sleepers is the answer.

 

So you're answer would be that in the next x couple of cycles (37 spins) the sleeper will have a higher hit rate (probability) than the random number.

I like these kind of questions because it's tangible and it's something I can test. I'm also reading a lot on entropy and probability like @Luckyfella suggested. But it is really hard to grasp.