Analyzing roulette spins and computer prediction


I took some time to analyze data, good results where computer predicts ball impact with rotor; I took all data in to XL then I sort it by accuracy with which ball hits rotor according to prediction.

I got this graph, so amount of pockets from prediction where the ball hits is my X coordinate.
Y is amount of pockets where the ball stops from prediction.
On the left the graph starts with spins where the ball hits -18 pockets from prediction. There is only few spins there as you can see the most of spins are -6 to +6, within 12 pockets of accuracy.

I am interesting to find out what is the difference on final results in between spins where the ball hits earlier compared to spins where it hits later.

If we look GREEN marked area , we can see the ball hits and most of the time jumps around 4 pockets or it jumps more to 12.

BLUE area is where the ball is more hitting lower parts of diamond. The ball hits few more pockets from predicted number, but it is coming towards rotor more vertically.
Increased percentage of spins the ball is not jumping it makes final results to overlap with Green area but if it starts rolling it makes ball to end up a bit further then Green spins. Also note that negative results in that area instead of -15 may be +21 etc.
RED area, looks the most unstable, it is probably where the ball hits very top of diamond.
It gives spinners, and various ball jumps.

JELLOW area is only 9 spins, the ball hits 10 to 17 pockets in front of prediction, so it ends normally. But most of spins are not good for us since the ball did drop to early.

It is what I call wheel quality, because of some distortion on ball track the ball did not end up according to measured parameters, it start dropping earlier, it gives good hit to diamond but it is one diamond to early.