Roulette physics. JAA Cross Referencing...

[Steve]

Although, if statistical anomaly is present, that would surely be reflected in some numbers  winning more frequently than others for various spin directions.

No that would be bias. But we still segregate directions, which isn't essential, but combining directions can nullify patterns/accuracy.

SDs outside expected certain results were for the available number of spins. And we could make that dynamic by continuing to collect data as we played.

We could use SD but we don't. We use something similar. Yes data is continually collected. There's a phone version that does that covertly at the table, and a web version where data is taken outside the casino, entered into the software, then betting charts taken back in.

Although I can see the advantage of having that done for you, as it would be a long grind manually. You'd still have to collect the data either way though.

Most players don't have the patience or stamina for data collection. If the wheel is suitable, an edge usually starts very early and peaks at about 300 spins per direction. But still it isnt smart to start betting until you have 50+ spins at least.

Conditions are optimal only with particular wheels and conditions. You cant beat every wheel in all conditions. And like with other AP, if you make mistakes you can get a negative expectation stronger than the HE (double-edge sword).

Also, is it possible to explain what you mean by cross referencing.

It basically means comparing dynamic variables to others, and modelling the relationships. Some data that doesn't fit the model is discarded, because it doesn't represent typical data. It is applied in various areas of science to predict the outcome when variables change. It is modelling cause and effect. I dont want to give specific examples of how it is applied in roulette. It is just one way of beating roulette, but not the most direct.

[Firefox]

Thanks for the explanations.

[The General]

As the prediction is made before ball release it must be trying to exploit some kind of bias in the rotor or maybe rotor and stator combined. However modern rotors are very accurately made. The one I have is a solid machined block of aluminium made to very fine tolerances. It's a work of art.

JAA isn't really related to bias, but you touched on the subject above.  Many wheels aren't well made.  Where they run into problems is with the assembly of the components.  For example the pocket compartment,  pocket inserts, the fret ring and in some cases when frets are individually seated.  The machining tolerance maybe high, but the assembly is sometimes piss poor.  Something that I've exploited countless times are pocket compartments that aren't evenly seated, that rise and fall and that aren't always in full contact with the rotor floor.  Sometimes it creates a loose energy absorbing section or number on the wheel.  Loose inserts are golden as well.  Casinos can be responsible for the loose lobes on the wheel when they attempt to rotate pocket compartments and do a poor job of resetting the component. 

In order to really grasp the causes of bias and the weakness of various wheel designs you need to take them apart and research them.

[Steve]

Almost every used wheel I've inspected has loose plastic pocket inserts. These are wheels that were in this condition in the casino, and they had obviously been like that for some time before being replaced. You can hear when the ball hits the plastic there's a pop instead of ping sound.

Of course wheels get replaced when they're "totally screwed", but they are still on the floor "screwed" for some time before they're "totally screwed". The wheel I have is used but was nearly new when I got it. The only reason I got it in good condition was because of outside scratches.

Anyone who says biased wheels dont exist really doesnt have any idea. But still bias is not my preferred method.

[The General]

I've always been into both VB and Bias.
With bias, I don't have to be the player and can remain discrete.
With VB, I have to be the player.

[Firefox]

In order to really grasp the causes of bias and the weakness of various wheel designs you need to take them apart and research them.

On my wheel there are no pocket floors or frets. It can't be taken apart. Everything is machined out of solid aluminium. Very little to go wrong. The numbers ring I think can be rotated if you take the turret off and have the right tool. I'll post some photos sometime.

I can see a little more about JAA now when only a minimum of 50 spins are needed to assess, that wouldn't be enough to assess on the basis of bias (I believe 300 to 500 are preferred).

But if software makes a prediction, it has to be either on the grounds of frequency or association with some anomaly. For example 15 comes up most, or if it doesn't come up most it comes up more after some numbers than others eg 24.

[Steve]

Everything is machined out of solid aluminium. Very little to go wrong.

Yes that's a much better design.

minimum of 50 spins are needed to assess, that wouldn't be enough to assess on the basis of bias (I believe 300 to 500 are preferred).

Correct its not enough to assess bias. But we dont bother with bias. Frequency of numbers is not a consideration unless the player has enough data for bias, and wants to look at that too.

[Firefox]

In fact mine is similar to the one below, but I will post some photos and videos of mine when I get the time.

It's a very low maintenance design and quite random. Still beatable if you know what to look for. A vast improvement for the casino from the wheels of old though.

I got mine from ebay when a casino folded. Cost about £900. I think new they were about £7000 though that was hearsay. Probably worth £1500 or more at today's 2nd hand prices. It will spin for 25+mins with a good push so rotor slow down during a spin is negligable.

[Proofreaders2000]

Most players don't have the patience or stamina for data collection. If the wheel is suitable, an edge usually starts very early and peaks at about 300 spins per direction. But still it isnt smart to start betting until you have 50+ spins at least.-Steve H

I do not see how it is feasable to collect data
for wheel bias with the close management scrutiny.

[Firefox]

I do not see how it is feasable to collect data
for wheel bias with the close management scrutiny.

It's not too hard. Just act like a normal player. I think Barnhart recommends a minimum of about 300 spins in his book Beating the Wheel.

You can mark off with little dots on the wheel card the casino supply to system players. Later transfer data to book or spreadsheet. If they have a marquee board it's even easier, you can move around between tables. Make random minimum bets on red or black every now and again it hardly costs anything. You can do it over two or three evenings.

Unfortunately, most of the ones I have done show little bias, or 1% or less. And if they move the numbers ring, and you can't spot it then back to square one.

We have a limited choice of casinos in the UK and they tend to be well maintained with modern wheels.

[Steve]

Data collection isnt a problem. It just takes time, and you do it at optimal times of day.

Beating most wheels one way or another isnt a problem either.

The biggest problem by far is avoiding detection during the profitable period. Its not difficult even for larger amounts, but its not something you can do at the same casino, day after day.

[Roulettebeater]

Steve
You make us laugh at you!
The guys here are struggling to beat the game, so far without success !
And you say that beating the game one way or the other!
I don’t speak about myself but I think many members find this a show off.

Anyway, let’s be positive, I was thinking of something that can make a boom ! And here it’s, has anybody thought or developed a a software or a system that consists of many subsystems, the subsystems evaluates many many systems in the background and the system on the top detects how the each sub system is performing and take relevant action by attacking the game .

[Steve]

The guys here are struggling to beat the game, so far without success !

Imagine trying to run on water. Then someone whispers to you, "hey, that wont work. Youll just sink. Try on land".

You dont listen, then try something you think is different: hopping. The result is the same.

The person gradually speaks louder, and you just try a different version, with the same result. Maybe try walking backwards on water.

Eventually the person trying to help becomes hated. Maybe even accused of a hidden agenda. Thats the position experienced members are in.

Once you understand what does and doesnt work, and why, you can try land. Its really not hard to overcome a small house edge. But you wont do it with an approach thats the same old repackaged shit.

[Steve]

has anybody thought or developed a a software or a system that consists of many subsystems, the subsystems evaluates many many systems in the background and the system on the top detects how the each sub system is performing and take relevant action by attacking the game .

Sounds like an AI type approach. Thats how i found statistical anomalies in roulette, and its what i now do in stocks and crypto with a bot now.

[Steve]

But if you do backtesting like that, you also need to understand the cause and effect, or you'll just end up curve fitting.

Curve fitting is basically designing a system to do well on set spins. Without modeling cause and effect, the system wont do well on different spins.