Statistics and physics

[Roulettebeater]

Hey guys

I was in the last days busy in building up a piece of software which perform all these complicated calculus/simulations you have mentioned, the software is basically a sort of code which runs with a DB and provide the so called "hot number", all what we need is sufficient data (for every run, the software needs to be inputed with the last 4 numbers which came out) and then it will run some equations/simulations and provide the prediction.

everything looks fine with the prototype, but sometimes the result shows up 60 or 120 degree away from the predicted number, this phenomen is apparently because the metal deflector... i have no idea how to solve it.

maybe we can form a taskforce (kingmag, ego, steve and i) to make this go beyond our immagination?

[cht]

VB is not something you learn from a theorising stance, you're wasting your time.  Hypothesising this and that is not the way to go.

You can only learn stuff by practice with actual SEEING with your eyes the wheel, the ball and all that goes on in it. It takes time, observation, the details where you build up your knowledge and experience with practice.

It's like football, you can't learn to stop, dribble, lob or volley a football watching tapes in a classroom. You learn it by playing football on a field. The same with VB, it's all practical.

[Roulettebeater]

VB is not something you learn from a theorising stance, you're wasting your time.

You can only learn stuff by practice with actual SEEING with your eyes the wheel, the ball and all that goes on in it. It takes time, observation, the details where you build up your knowledge and experience with practice.

It's like football, you can't learn to stop, dribble, lob or volley a football watching tapes in a classroom. You learn it by playing football on a field. The same with VB, it's all practical.

VB is cool, but we want to automate it, we want to make it more sustainable.. we simply want it to make us money

[denzie]

@cht. .. What are the right conditions?

[cht]

@cht. .. What are the right conditions?

Do you want a hard, medium or soft spin?
Do you want the rotor speed to be fast, medium or slow?
Do you want the ball to strike the deflector or no?
Do you want bounce more or less ?
Do you want short, medium or long scatter?

The list goes on. The right conditions are the ones that you want so that your predictions are accurate.

Notice I played only 8spins out of 30, those were the right conditions.

[denzie]

The thing is...if I read how ego talking about vb it shows how less I know about it....But. .....

It might sound stupid but after few rounds of the ball I mostly know if it's gonna be a predictable spin. Just by looking at the visible speed and definitely how the dealer spin it.

Still I can't hit as much as you.  None the less. .... Still enough to profit. 

[Roulettebeater]

The thing is...if I read how ego talking about vb it shows how less I know about it....But. .....

It might sound stupid but after few rounds of the ball I mostly know if it's gonna be a predictable spin. Just by looking at the visible speed and definitely how the dealer spin it.

Still I can't hit as much as you.  None the less. .... Still enough to profit. 

I have modelled all what ego mentioned in a piece of software.

[denzie]

I have modelled all what ego mentioned in a piece of software.

What exactly did you create?  Can you be specific pls ? Much appreciated 

[Roulettebeater]

What exactly did you create?  Can you be specific pls ? Much appreciated

Denzie, i am a software engineer, my expertise is big data and data analysis.
i have a huge database of data mainly spins results, i have used these data with help of some physics equations to estimate the speed of ball and rotor, altogether modelled in a software which require only the last 4 spins to give me the hotzone on which the ball might fall.
the software is basically like a computer, it uses same equations but use the most recent spins to estimate the speed and capture the hotzone.
if you have specific question ask!

[denzie]

if you have specific question ask!

A few 
So you have a database with spins. Alright. How you calculate the speed of ball and rotor from that ? Coz basically you only have the outcome and gaps between the outcome. Did you add something like the average rotations the ball makes ? Or other stuff?

Im curious 

[Roulettebeater]

A few 
So you have a database with spins. Alright. How you calculate the speed of ball and rotor from that ? Coz basically you only have the outcome and gaps between the outcome. Did you add something like the average rotations the ball makes ? Or other stuff?

Im curious 

i use newton's motion theories

[denzie]

How's the results of testing so far ?

[ego]

 You need practical ways to measuring the ball speed and i will be honest i only know two ways that might work with todays wheels and conditions
I know many more ways but they are old and predict 6543 laps before drop and are not working with todays conditions

In theory this might work predicting ball at beginning of spin - but is not pure visual ballistics where you estimate the ball later during spin - so this become more a wheel signature way of estimating the ball speed where most of the pins samples of the same category has the same value

Let say you have the reference deflector based upon what kind of drop zone you have - then you wait for the dealer to throw the ball
After ball pass your reference deflector the second time you read the number below the deflector when ball is over it
This is your dealers release number

Now you follow that number and wait until the rotor has made one full rotation and the number is below the deflector again
During this time you count how many ball laps the ball made - for example four laps

Now the ball will not be over the reference number at the exact same spot at the reference deflector
This might happen five pockets from the reference deflector when ball is over reference number

Then you get a pattern for that particular ball spin and rotor length
That would be four ball laps plus five pockets - so next time you get four ball laps and five pockets pattern you will assume you get the same ball speed and rotor speed

But you can only assume you will get similar ball speeds and rotor speeds combinations for each pattern
So assume you collect ten spins with the same pattern - then your assumption is that most of them will become the same spin

Now to get distance from Point A (visual read) you have to notice the number below the reference deflector when ball is over your reference number to get the correct distance to Point B (drop point)

I test this with video spins and getting plus/minus one lap - so this is just theory

Here come the good part - assume that average rotor speeds and average ball speed create four ball laps during one full rotor rotation.
Then if ball/rotor is slower or faster the average you will get more or less laps during one full rotor rotation

So you can easy with eye observation track slow, medium, fast ball/rotor combinations
Now to the one part that i did not have time testing and is also theory

You can do all this with metronome in you mobile phone vibrating every 0.5 sec - then you watch the release number during one vibration and count how many vibration it take until rotor made one full rotation and ball made certain amount of laps until ball and release number meet again (when ball is over release number) lets say your get 21 vibrations then each time you get 21 vibrations you will have the exact same ball speed and rotor speed and the measuring will be very precise.

What you would do is to take the visual with the number below the reference deflector - then from that particular number divide the wheel into three sectors and see in wish one the number fall into - for each spin with the same vibration count or pattern you will get a new visual read number and divide the wheel into three different sectors then the past time - but distance remain the same from point a to point b - so when you see that some hits hit more frequent in one sector you place you bets there

Cheers

[Roulettebeater]

You need practical ways to measuring the ball speed and i will be honest i only know two ways that might work with todays wheels and conditions
I know many more ways but they are old and predict 6543 laps before drop and are not working with todays conditions

In theory this might work predicting ball at beginning of spin - but is not pure visual ballistics where you estimate the ball later during spin - so this become more a wheel signature way of estimating the ball speed where most of the pins samples of the same category has the same value

Let say you have the reference deflector based upon what kind of drop zone you have - then you wait for the dealer to throw the ball
After ball pass your reference deflector the second time you read the number below the deflector when ball is over it
This is your dealers release number

Now you follow that number and wait until the rotor has made one full rotation and the number is below the deflector again
During this time you count how many ball laps the ball made - for example four laps

Now the ball will not be over the reference number at the exact same spot at the reference deflector
This might happen five pockets from the reference deflector when ball is over reference number

Then you get a pattern for that particular ball spin and rotor length
That would be four ball laps plus five pockets - so next time you get four ball laps and five pockets pattern you will assume you get the same ball speed and rotor speed

But you can only assume you will get similar ball speeds and rotor speeds combinations for each pattern
So assume you collect ten spins with the same pattern - then your assumption is that most of them will become the same spin

Now to get distance from Point A (visual read) you have to notice the number below the reference deflector when ball is over your reference number to get the correct distance to Point B (drop point)

I test this with video spins and getting plus/minus one lap - so this is just theory

Here come the good part - assume that average rotor speeds and average ball speed create four ball laps during one full rotor rotation.
Then if ball/rotor is slower or faster the average you will get more or less laps during one full rotor rotation

So you can easy with eye observation track slow, medium, fast ball/rotor combinations
Now to the one part that i did not have time testing and is also theory

You can do all this with metronome in you mobile phone vibrating every 0.5 sec - then you watch the release number during one vibration and count how many vibration it take until rotor made one full rotation and ball made certain amount of laps until ball and release number meet again (when ball is over release number) lets say your get 21 vibrations then each time you get 21 vibrations you will have the exact same ball speed and rotor speed and the measuring will be very precise.

What you would do is to take the visual with the number below the reference deflector - then from that particular number divide the wheel into three sectors and see in wish one the number fall into - for each spin with the same vibration count or pattern you will get a new visual read number and divide the wheel into three different sectors then the past time - but distance remain the same from point a to point b - so when you see that some hits hit more frequent in one sector you place you bets there

Cheers

Thx, but impractical as it is too much complicated.
I have another approach

[ego]

You talk fuzzy with no example just like a TROLL
I think your knowledge is crap as you can not prove you point knowing anything about visual ballistics.
Is fun to teach when you know they know nothing and i know it all

Here is another way - set the metronome at 1 sec and watch where the ball is on the ball track each time the metronome vibrate
At the beginning you will see the ball at different places on the ball track

When ball two vibration meet and ball is in the same spot at the ball track - then ball has made exactly 1 sec
At this moment you will have 9 sec until ball hit dominant deflector

This way is not working with todays wheels as it predict to late during spin and you will have No More Bets calling
But i have news for you - you can set the metronome at:

1.0
0.9
0.8
0.7
0.6
0.5

And predict before No More Bets - i have test 0.5 and is hard to visually see when ball divide being faster and slower then 0.5 sec - but it can be done with practice - 0.8 is more easy and predict more early then 1.0

Maybe you have a hard time to understand this and find it to complicated - no worry i will explain again

You set the thumper to 0.8 sec and watch the ball position on the ball track each time the metronome vibrate
First ball will be at different places - but when ball is once faster then 0.8 and once slower then 0.8 moment - then it looks like ball is meeting on the ball track and are at the same spot during two vibration from the metronome.

Cheers