We can design and even test E2 system without device. For that we would need video recording of spin, and kind of video editing program that can move picture in steps.
I use (Life View Suite)(“avidemux” is also good)
The best would be if we have spin of about 3 sec per rev. of wheel that we will play on. If system is defined based on video of some old wheels with fast deceleration of the ball would not be good to apply it on wheels that are used in most casinos. Design is very simple.
Find position of the spin about 18 sec before then ball drops down. Notice number under the ball and in steps go forward until ball makes full 3 rotations and comes again to same number. In same time you need to record how long it takes to do it. In usual program moves in steps of 40 ms. (25 frame pictures in sec, standard TV video recording or if it is in US it would be slightly different, they use 30 pictures / sec.).It should take somewhere about 2 sec in total.
Now we move about 4 sec. in front from 18 sec. So somewhere about 14 sec before then ball drops. And we do same. But this time we apply same time that took ball to cross all numbers 3 times. Now the ball in same time needs to cross all numbers 2 times. It means that ball is at 2/3 speed then when we did it first time. If ball crossed more than 2 wheel numbers, try to start it later. Maybe 13.8 or 13.5. You must adjust it that ball crosses exactly 2 times all numbers. If that happens at 13.5 sec then 18-13.5 = 4.5 sec. So we need 4.5 sec that ball slows down to 2/3 of speed when we started. There is fine adjustment to that but that can be done by experiment when we do testing. Why do we need it? The answer is simple. 1/3 of speed is missing or if we look it in numbers it means 1/3 of numbers is missing. So when we multiply it by 3 (as explained with my system) result will tell us when in time we started, but in numbers not in time. That is what we really want.
Lets look simple example. One spin in different time moments. If we start at 16 sec. Zero is under the ball. We finish at 11.5 and again zero is under the ball so the angle that we monitor is zero. Multiplied by 3 is again zero. If offset is excluded it tell us that ball will hit zero. Now what will happened if we start later then 16 sec. Let's say 16 plus that ball makes another 12 numbers. In that case number 13 would be under the ball when we start. After 4.5 sec it would not be again 13, because ball is going slightly slower. It would be 2/3 of those 12 numbers so it would be on number 17. Now we do have angle in between starting number 13 and ending number 17, it is 4 numbers. We can now apply process as explained in E2 system. Add that angle 2 times on top. So after adding it ones we are on number 4 and ones more we come to zero. Here we go; the system put us back to same reference point. And this law applies all the way from 19 sec to 9 sec before the end. But because we need 4.5 sec to predict we must exit at 9+4.5=13.5. We can take it as 14. So 19 - 14 = 5 sec. That is time frame that we must start. We need to have feeling for that. It is not small it is about 7 -8 ball rotations, and after some practicing it should not be a problem to recognize spin within that time. Other answers are logical. If ball did not stop on zero but on 12. Than from zero to 12 would be our offset.
How wheel change is includes in all of that?
Because we are monitoring the ball but over the wheel numbers, it is included. It is not 100% included that is why with more constant wheel system is more accurate. How much it is explained at my page
http://www.rouletteplace.com/advanced.htm. If everything is same with spin as in previous example. But wheel in our time frame is faster by 2 numbers. Then angle that we monitor would be also different by 2 numbers. So when we multiply it by 3 it would show total difference of 6 numbers to the end of spin. If ones we started at 14 s and another time at 18.5 we would have error in prediction slightly smaller than 2 numbers. That is why I come with E3 where I managed to reduce it by adding one extra switch click. For the end, I do not clock the ball by manually clicking switch, by my opinion it is imposable. So I observe it visually. I do clock the wheel it is much slower and it is possible because even if I make error by clocking the wheel, the system is not so much affected because it is only adjustment. Main part is done with visual observation.
Testing, use came video recording and apply the system. Use time as you defined. It should be somewhere in between 4-4.5 sec. Try to start in different times of the spin. It should always point to ref point- offset. If slightly losing accuracy readjust your time by 0.04 sec. steps. Then try on another spin with slightly different wheel speed. It should work same good. So you can perfectly test it at home but to be able to play it you would need some kind of device as I make. You can try to make similar and it should be reasonable good. I would recommend my device because there are few more small things that makes it perfect. I did a lot of testing and playing, so I can confidently say that I know how I build them and why it is that way. Wish you best luck in development.
