Ball bouncing scatter patterns

Advanced roulette play
1

What is the accumulated wisdom about the variance in prediction created by the ball bouncing (or by its “scatter pattern” to use other terminology I’ve seen around)?

I’d like to see serious scatter patterns from non-tilted wheels of kinds commonly used in casinos! It seems to me that the variance of the scatter pattern is an unavoidable component of any prediction device (which is chaotic enough to totally prevent any other modelling than statistical sampling): Given a perfect prediction of the point where the ball will hit the wheel (or a diamond or some such specific point), then what is the expected winning number and what is the variance around that number? In your practical experiences.

2

Rollo

The site below discusses scatter and shows a plot for one setup. I agree its a critical component to get a clear understanding of. If the pattaerns I’ve seen are at all typical the scatter definately follows a pattern.

I’m going to buy the Spin DVD once I’ve finised analysisng the free 14 spins (very little scatter). From that I will constructr some real world scatter data and post.

Regards

3

Using the scatter graph which gkd linked to above, tt would be tempting to bet on number 9 on the graphs x-axis since it has the highest frequences. However, that i likely mostly due to chance and will not continue. For reliability, one should choose a zone with consistent above average odds. Here are two examples:

30/50 or 60% of the winning numbers are concentrated to 12/37 of the wheel (between 6 and 19 on the graphs x-axis). This means that each of those numbers on average have 60%/12 = 5% probability to win. With 36 times the money, one gets 5% * 36 = 80% positive return on all bets there in the long run!

18/50 or 36% of the winning numbers were concentrated to 6/37 of the wheel (between 12 and 17 on the graphs x-axis). This means that each of those numbers on average have 36%/6 = 6% probability to win. With 36 times the money, one gets a whopping 6% * 36 = 116% positive return on all bets put there in the long run!

Note that I assume betting on only one single number at a time. Note that this means great risk, in spite of the great odds! In the first example one could loose all 50 times in a row if one bet on number 7 on the x-axis (one out of twelve). In the second example one could have won only once, netting a loss of 14 times the unit bet, by betting on number 14 on the x-axis (one out of six, right in the center of the big peak).

While I certainly would like a greater sample than just 50 spinns, I will use the scatter given by gdk to put up a risk management model where risk and profit can be traded off against each others by betting on several numbers simultaneously.

It would be nice if the scatter could be approximized with a parametric distribution like the normal distribution. However, it has often been stated that the distribution is multipolar with a minor peak opposite to the major peak, as is also suggested from the scatter graph in this example. The ball might well find it easier to jump over the center of the wheel to an opposite position, than to a number say 90 degrees away, so it can make sense.

However, once the prognosis is ready, there cannot remain much time before NMB. And then one must identify the neighboring numbers and bet on them. Betting on 10+ numbers might not be practically realistic (but on the other hand, it can’t be difficult to learn the sequence of numbers on the wheel by heart).

4

Its a very typical Huxley scatter build up. You can be pretty sure that the 2 main areas is around 8-9-10 and 26-27-28 probably moving slightly upwards to 9-10-11 and probably wider. You will have an edge in both areas. Use the typical 180 degrees split bets 1/4 - 6/9 - 16/19 - 32/33 etc. and add 1-2 neighbour chips to the bet in the main area.

I agree that it is probably a coincidence that 9 is so large.

What we dont know either, is the rotor speed. If the second main area at 26 is a result of a 2.0 rotor and the rest a product of a 4.0 rotor, then we are screwed anyway. The scatter for 2 such speeds could be from to entirely different wheels.

Scatters from the same wheels should be categorized in different speed categories and predicted as such. Not as a multiple bunch of spins in same categori, unless they actually are in the same categori.

Same goes for Scotts prediction wise: All rotor speeds has a specific reading spot. The differnce from each observation point is a combo of the different rotor position (for a faster rotor) and the slightly longer bounce. (For a faster rotor)

5

Few same kind of wheel, at least they look the same with same kind of ball.
If on each of them ball rotation takes 1000ms. Remaining time on each wheel is different.
Few different cases and result in remaining time changes from 7.5 sec to 12.5 sec.
Therefore if on one wheel at particular rotor speed something is valid that maybe completely different if applied on the other same looking wheel. Make it simple, it is not enough just to define when ball is at particular speed range, it is important to know remaining ball traveling time on particular wheel for defined speed. Ball deceleration is mostly defined by fraction but remaining time is defined by angle of area under the ball. If area is more vertical the time will be shorter.
Ball deceleration partly effects remaining time but it can not define it.

6
Scatters from the same wheels should be categorized in different speed categories and predicted as such.
Yes of course! I feel so stupid that I've missed this obvious and important factor. I'm developing a wild idea about using some hitech to predict roulette, inspired by sites like this one. I've made some process modelling on paper and I'm really surprised that I never thought about this aspect. What else have I missed?

Actually, this is very good news! Considering the speed of the wheel would make the scatter distribution more narrow than I hitherto had imagined possible.

One part of the scatter is about which diamond and where on the diamond the ball hits (Forester has made a great study of the obviously significant scatter differences between hi, lo and middle diamond hits, somewhere on this site). This deterimines in part how far the ball will jump and of course the position av speed of the wheel determines which number will be there then. ANother part of the scatter is generated by the repeated collision of the ball with the rotating wheel. Of course the amount of kinetik energy transfered depends on the speed of the wheel. So it seems clear that knowing the speed of the wheel could add lots of information (and heaps of money! :lol: )

7

To predict position where the ball will hit diamond is possible only on tilted wheel.
On leveled wheel it would be hard because there is 4 equally dominant diamonds and we never know where breaking point is.
If standard difference of length of ball rotation we divide by 4 we end up with time of about 50 ms. Time of ball rotation 1.000s and 1.050s should hit 2 different diamonds. But we cannot know where breaking point is. Maybe first diamond is 0.951 and next one 1.001 maybe it is 0.975s and 1.025s. On top of that we have inaccuracy of clocking.
With tilted wheel it is possible. Wide range of speeds will hit same diamond.
If wheel is tilted so that 70% of ball speeds belonging to one earlier rotation will hit same diamond then we have gap of 70% of 200ms to predict position. That is 140ms difference. To position our prediction in middle of that we need some adjustments.
It can be done on 2 ways.
To program device for particular wheel or that device gives information not only about prediction but about ball speed as well. So we play only particular speed range where we know that the ball will safely hit just above middle of diamond. If ball speed is slightly greater it will still hit same diamond but with slightly higher position. If it is lower it will be lower position. If we are not balanced properly weaker ball may not hit lover part but it may happen that it will hit top part one revolution earlier or it will go for another diamond which would result in different result then what it was predicted.
By my understanding Barnett’s prediction of very selective spins is probably the most correct way. For example, if only ¼ of ball speed is selected and if we adjust clocking point according to that or if we tune device for particular wheel and particular point then we will have 100% (-scatter) accurate prediction. If we are balanced in middle and accepting only 50 ms difference we will have (140-50)/2=45ms on each side gap for error which even with single spin clocking would be hard to make. By selecting only ¼ of prediction we also reduce difference in ball traveling time by ¼. Therefore even if we do not include adjustment for that we will end up with reasonably good result.
:slight_smile: :slight_smile:

8

Barnetts computer predicts exactly like traditional wheel watching just mecanical. Detect rotor speed and detect ball speed and exit when a certain ball speed is achieved.

9

With only one difference, computer can define ball / wheel speed within 30-40ms per rotation.
That would be worst case without any correction and single rotation clocking.
It is still good enough to define particular ball rotation for tilted wheel prediction.
Computer can specify which speeds of the ball at particular spot will be playable and ignore the others to gain higher advantage and almost certain ball hit at particular diamond spot. On top of that it may accurately calculate wheel traveling for period of remaining ball rotations.
VB player would struggle just to define right rotation.
You were claiming that you can detect wheel speed change of 100ms per rotation just by watching the wheel. I can not do it and anyway how many people can do it? 100-150 ms difference when wheel is used as reference would lead to improper ball rotation and have as result wrong prediction. Principle of computer is same but accuracy is not the same.
Even if we accept that experience VB player can most of the time define right ball rotation but he can not define particular part of that rotation as if accurate timing device is used. I am saying that Barnett’s solution is very simple and can produce high advantage without years of practicing.

If particular observed point ball that takes 1sec per rotation and ball that takes 1.2 sec will travel same distance. VB player maybe ok. But what if only ball time range from 1.050s to 1.120s would hit same diamond and valid spot on diamond. From 1.000 to 1.050 the ball will just touch top part of diamond or miss it and go for another one or for another full rotation. And from 1.120 to 1.200 the ball will be caught by diamond but 1 rotation earlier or hit horizontal one and turn in to spinner or stop to early. To have advantage here we need high accuracy which I think VB player cannot have.

Wheel speed change, what if wheel changes from 3 sec to 3.1? Even if some players can detect that and somehow act according to it they can adjust final result and place chips few numbers earlier because of slower wheel. In this case that ball of 1.2 sec would match same pattern as ball of 1 sec with faster wheel. In previous example, which would be most common to find in casino that would be completely different result?
Even if wheel is strongly tilted ball of 1 s and ball of 1.2 sec would have same distance traveled but it will not be same time of traveling and that is additional problem.

By using only selected part of spin Barnett can avoid it. I do use it in calculation and it can be difference up to 1 sec in ball traveling time. 1 sec on 2.5 sec wheel is 14.8 numbers difference in prediction. Computer can define ball speed by using few timings it can be better then 10ms per rotation of error. With such accurately defined ball speed another option is available. Based on position where the ball will hit diamond scatter can be better predicted which is very important with new balls that casino is using.
There is another part in VB which I do not like.
If you are looking that from1 point wheel makes 1 rotation and ball makes let’s say 3 rotations. What if ball makes 3 rotations +1 number? Probably you would want to play and that spin because to wait for exact match may take logon time. That 1 number difference in same cases could mean same rotation but it can be different.
As you know my E2 system observes ball. It is very possible that with all movement on the wheel I pick number under the ball wrong by 1. In my case it is 3 numbers error on final result on tilted wheel and VB it can be wrong ball revolution.

If we compare it to computer and selective spins it is to much and not so effective.
If computer player selects only 50ms window he can play about every forth spin but he will have very high accuracy and definitely he will not define wrong revolution… If he positions his clocking that valid spins are in middle of diamond he will have every spin well predicted. And it does not require much of brain to play or to design that kind of system.

10

Exactly what Barnetts computer does. Defines correct revoloution and selects only the spins that the VB players define as “Nailed It”. They also operate with “Too Fast” and “Too Slow” and “Weird” some choose only to play “Nailed It” others adjust for the others exept weird which really could go any way.

Its not a struggle to define a particular revoloution. Not for a VB player, nor for Barnetts computer.

The strike on the diamond can lead usually to 2 main areas depending on the strike type on the diamond. `Of course there can be a lot of strike types but usually you would only play a wheel if it has no more than 2 main areas.

I can detect wheel speeds change down to 0.05 - 0.10 sec. and im not particular skilled. I am probably in the middle. Its all about a specific tecknik you have to learn. I guess, once you know you would say: Damn why didn`t i think of that.

If you detected the ball speed “Nailed it” in the correct revoloution, then the only difference in the outcome will be the difference in rotor speed from 3.0 to 3.1. Nothing else.

Im off to Nice, France now, maybe i will drop by at some point during the next 12 days. I don`t know.

11
If you are looking that from1 point wheel makes 1 rotation and ball makes let’s say 3 rotations. What if ball makes 3 rotations +1 number? Probably you would want to play and that spin because to wait for exact match may take logon time. That 1 number difference in same cases could mean same rotation but it can be different.

I don`t know what you mean. That particular scenario has nothing to do with VB.

If we are to discuss errors, then one thing that has puzzeled me from the very beginning on your visual system is:

You multiply angles when you are about to read the number. If that calculation takes 1 sec. on a 3.0 rotor you are 12 numbers off. If you are not consistent, you might do it on 0.5 sec. the next time and 0.75 sec. the third time. On top of that you need to see the difference from a 75 and a 80 degree angle very fast and read the number below at the same time.

Thats multitasking at the finest i must say.

In VB you read with the same consistency at the same time frame at a known observation spot every time. You don`t have to search for the spot.

12

If you can have good feeling for rotor and if you know how to handle change than you probably can achieve same results without danger of using device. Actual ball position you can use to know if spin is valid or not. I have feeling for ball but not for rotor. With device I exactly know when will be indicated targeted revolution. Well if I could develop that skill I do not see reason why someone could not do the same but with wheel speed detection. First time when I come across tilted wheel I probably did apply similar way, but I used timer to help me to notice wheel speed change.

“You multiply angles when you are about to read the number. If that calculation takes 1 sec. on a 3.0 rotor you are 12 numbers off. If you are not consistent, you might do it on 0.5 sec. the next time and 0.75 sec. the third time. On top of that you need to see the difference from a 75 and a 80 degree angle very fast and read the number below at the same time.”

Now I do not understand you. Time how long I calculate has nothing to do with result.
Simply if you started at zero and ball after expired time is at 15, add 2 times more that same distance. If angle is 75 or 80 it does not make such great difference. First position you see so you are not wrong, now on top of that if you add 2x75 =150 or 2x80 =160 is only 1 number difference. If that is the only problem I would never look for more.