Can someone help with building a physics simulation? Using ml or neural network is also fine option.
What you mean :
Simulations are usually made with programs. If they are done well, they can be a near to ideal match that we can have on physical roulette. What is a physics simulation for you?
You see i mainly try to work with computer vision i just have interest in that. My goal till now was to get as accurate as possible values “speed, possition, etc”. But now that i think i have them, what do i do with them? How do i use them? And thats why i want to build the so called “Physics Simulation” to be able to use those values in someway.
Doesnt really work. Every dealer is different, every air pressure is different. Every tilt of wheel is different, even every ball replaced is little bit different. The key to success in roulette is to know how to adjust the bettings according to the above using common sense and a little of of help from data.
I hate hearing that something doesnt work. whatever factors there are if i need il change it every 2 min. i do not accept “cant be done”
It can’t be done with your program but it can be done with my human brain.
Not to be rude but
Computers predict meteorites path for the next 200 years with just 2 photos per day, but your human brain is the only thing in this univers that can predict the roulette? Why do you think you are the only one that has a right way of playing? All i had seen in this forum from you is just you demotivating people. You are not benefiting from that and the people in this forum are not benefiting from that. Why dont you say “yeah try your idea, give us results mby later on we can help with our idea” and etc?
But you do not say what the difference is between your “Physics Simulation” and common computer simulation …
In my understanding, computer simulation is more than fully enough if it is done well and gives results near to real results. It’s just that most of what I’ve seen done by others is not done at the right level and the problems come from there.
Often those who write programs for simulators do not know the game itself and the peculiarities of different roulettes very well…
In the game, the basis is to know exactly what happened and why. That is, to have the latest information. And that’s hard to do.
The programme based on the video has probably done it easily, what can be counted, but there are data parameters that are beyond the control of simple counting.
And to enter parameters that can be counted, a video-based programme is not necessary…
in my understanding common computer simulation is based on a lot of data and we are trying to match the data to get prediction. By physics simulation i expect using using all the needed factors to make as accurate as possible prediction. I just finished making as accurate as possible ball and wheel speed and im barely getting into thinking about actual physics. Also i dont mind sharing what i have till now so if someone have different idea of getting prediction than mine i can help with automatisation and computer vision
It all depends on the algorithm - too much data can also be harmful.
Simulation and prediction are different.
What do you mean?
How do you accurately measure speeds?
You have to realize that the forecast doesn’t depend so much on the accuracy of the velocity measurement. Or more simply, having more accurate velocities doesn’t guarantee a more accurate forecast.
Especially if is roulette with RRS.
If your computer can do what i can do, that thats great! However, if you cannot do it yourself, how can you create such programme?
If your program works, i am sure casino or wheel manufacturers coul have nailed it and roulette wont be beatable. It is just too hard for AI.
A simple change of dealer could fuck data up lol. A male dealer with a strong spin suddenly change to a female dealer with a weak and sloppy spin lol
Im not waiting for the ball or green to reach a specific diamond. Im starting to get speeds and possition the moment the ball is launched. Like the exact frame its released. I think " i might be wrong" but the initial acceleration of the ball might be the key. Again i might be wrong, up till now iv searched for perfection and im just starting to work with the data im gathering.
I doubt if something is possible to do this way, must be too big disperssion…
Here are no “suddenly change” 
.
Player see that and are ready for that…
Of course! It is easy to see as the dealer spin the ball and immediately i start to bet.
While i was less experienced, i asked the dealer to spin good so i can win. I can tell he is an experienced dealer and is a friendly one and know some stuff about roulette. Then he said i should bet good instead of asking him to spin good.
Good that all that is left in the past …