FF and VB2

MyRulet visual prediction -VB2
1

The idea is to add to FFZ VB2 with rotor clocking.
I am trying to get the best idea and this is someone’s input.

We always start when under ball is 0 and that is over 12 o’clock diamond.
You see that standard formula is not good in this case… Of course maybe here I take abit too big deceleration but that I done for simplicity, I researched very much variants and had the same problem.

So what is solution of that problem .If assume that before cadence is left about 13 sec then final number will be equal no mater if wheel is 4 sec or 6sec , in the picture final number is at 12 o’clock and it is 2 in both cases. And these cases fifers only that 6 sec rotor do about 2 rotations and 4 rotor do about 3 rotations . If we assume that ball speed is about 1 sec per rotation and its deceleration is about 160-170 ms then with 6 sec rotor ball and wheel together do about 1.17 rounds and with 4 sec rotor 1.25. If we take reference time for 6 sec rotor 1 sec then for 4 sec we must see after cadence the same number so reference time for 4 sec rotor must be such that rotor and ball together will do twice bigger way or about 1.17*2=2.34 2.34/1.25=1.87sec.
If we repeat all that from these 13 sec how we started in the picture we will get about the same number.
But what will be if we started 1 rotation earlier?

With 6 sec rotor we will finish 0.83 sec earlier and ball will do about 6 pockets longer way and rotor will be about 5 pockets less in way so we will see about the same number.
With 4 sec rotor and 1.87 sec cadence time we will finished also 0.83 sec earlier, ball will do about 7-8 pockets more than in previous example and rotor will be about 8 pockets less in way so again about the same number. If we start one rotation later error will be but again not so big.

2

You see that standard formula is not good in this case… Of course maybe here I take abit too big deceleration but that I done for simplicity, I researched very much variants and had the same problem.

Truth, the formula is to give you same prediction in different ball rotations and not to adjust for rotor speed changes since for rotor with VB2 we adjust manually. If your remaining time was shorter lets say 8-9 sec then the times may be ok and solution could be so simple. Applying formula and changing reference time according to rotor speed would in same time give us rotor adjustment and required linearity to be able to predict in few different ball rotations.
When playing VB2 6 sec rotor and using 1 sec reference time, if 4 sec rotor comes I would still use 1 sec time and add “remaining time” times “rotor speed change” in pockets.
6 sec rotor = 6.15p/s
4sec rotor = 9.25p/s
The difference is 3.1 pocket
3.1 x 12 sec remaining time = 37.2 pockets
So it is same. Problem is when prediction is at different moment, then for each sec difference it would be 3.1 sec error.

RC-VB2 objective is to cover rotor changes on the best way.
If I can understand right what you did 1.87s is the time that needs to be applied for a 4 sec rotor in your example. I am not sure how you come to that but it looks as a good value.

If 6 sec time uses 1 sec then for 4 sec rotor we need to add time in which the ball will cross 37 pockets (difference that rotor will make to the end of spin)
Since at 4 sec rotor ball of 1s/r in one sec. crosses 1.25 of the wheel then in additional 0.8 sec it will make adjustment for 37 pockets. Therefore 1s +0.8s=1.8 sec, should predict 4 sec rotor.

With earlier prediction yes it will work that way, some errors will be reduced, 6 sec rotor should be perfect and with 4 sec rotor there will be some errors. Instead of 0.83s time difference we should look about 3 sec. Since difference in between 1.8 and 1.5 (for linearity) is smaller than 1 and 1.5s the errors should be smaller as well.

I would say the system should be calculating as that but I think adding option of setting it for the most common rotor speed might be solution to increase linearity.
For example 3,4,5,6

Where if we select 4 they system uses VB2 formula defines time of 1.5sec (based on previous sample) then changes it based on rotor speed. Now if 4 sec rotor comes the system changes it to 0.7s.
If ball deceleration was greater some values might be out of proportion therefore some tests should be added. After playing at one wheel if other rotor time is selected, the player can learn to shift reference time starting point.

From clocked ball rotations we can get ACC and remaining time, needed in calculation.

Another idea is to cock the ball until it reaches particular speed then the system applies the time.
I do not believe it is a good idea. The truth is we might be more accurate in starting time but it is to much work. Clock rotor clock ball rotations in that case better use unlimited prediction.

Any ideas?

3

I wish I could have more of a technical input but it’s above my head. One would have to do a setup on each wheel one plays. Not a problem because it’s not like we going to find more than a single DD tilted wheel in a casino. If you do you are very lucky.

So what I would be looking for is to do a setup on the wheel using rotor speed and ball revs (deceleration). So the FF uses VB2 formula and works out offset time, that is setup right?

I do not think that it’s necessary to know exactly which rev to play in because as we know vb2 is flexible there but we could still do it without too much work…

To play:

  1. Forget about rotor timing first up. We do correct rev first or rather maybe 2 revs before correct “playable” rev. We do this by clicking everytime ball passes DD. Let’s say FF now reconizes we are two revs before ‘ultimate’ calc rev, let’s say 10 revs out. FF vibrates telling us to start calculating.

  2. So now we do rotor timing. Click on DD and click on DD. Setup has been done on this wheel so only rotor timing is a necessity. FF works out correct offset time and applies formula. The next zap or buzz number under ball is strike number.

Conclusion:

So I think I would have setup mode per wheel, then when playing a wheel, get correct rev, then rotor speed (could time half rotor like FF can do), FF automatically activates on second click (end of rotor timing) and calculates. etc. etc. etc.

Appologies if there is nothing interesting in there! :slight_smile:

4

I’ve made VB2 on FF with rotor clocking. (part of the program)

First set up, clock ball rotations.
Go back to menu select 4
The system calculates ball deceleration at point of 1 sec/r.
It also calculates remaining time form moment when the ball is 1 sec, it is used for rotor adjustment.
We clock rotor, then wait to start timer at desired wheel position.
After reference time we read number.

Also when the ball crosses same numbers again the system zaps again, just in case if we didn’t see it properly, but also to get an idea if we prediction around same moment during the spin. If The ball speed is 1170 ms when next zap happens it will perfectly match with first zap alternatively it will be close as much as our starting point. The idea is to start timer around 1-3 rotations before than the ball is 1 sec per rotation. Sure it can be and earlier but as you know, VB2 likes starting moment around same point.
As a main reference time the system uses calculation based on 5 sec rotor, then adjusts for changes.
With such approach there may be a problem if rotor speed is much slower. For example 10 sec rotor may reduce reference time a lot to make for rotor change adjustment. I added limit 0.75 sec min time. Common sense is required to play VB2FF. For such cases manual setting will be required.

First time I tested it, it looked ok but I believe I can do better just with a timer VB2.
Different rotor speeds just didn’t give right results, but I noticed that the reference time changes… increases with faster rotor, believing system functions as it should.
I found mistake in wrongly calculated remaining time to only ~4 sec where obviously it needs to be more. After correcting it the system predicts fine from 2.5 sec rotor to 6 sec rotor that I tested.

Not sure should timer be started on switch release or push. Somehow when rotor clocking is used I feel it might be better to change it to push.

5

I think it is good idea. Also it would be able to quit to menu without reseting power, when you use just a timer.

6

Forester, do not forget also on set up rotor timings. Here I mean : 1/2, 1/3 or even 1/4. At least it should have posibility to chose 1/2 of rotor timing.

7

With this approach when we clock rotor we can go back to menu same as if we play tilt2. The system wills also have and timer, so if I reverse switch function I might be able to make it to go to menu as well. Thanks for the suggestion 
It can clock 1/1 and ½ rotor , it uses systems rotor clocking. I do not believe ¼ would be any good.

8

2 green balls are 2 predictions that FFVB2 gives, they are inline.
First prediction happens when VB2 time elapses, then the system assumes the ball is 1.17s per rotation uses rotor speed as well, calculates the time when the ball will meet with the same number again, waits for that time and zaps again.
Two green balls are inline which means we had prediction of about 1 sec ball/r then next rotation was about 1.17s.
Blue ball with an arrow is second prediction, it passed the position of first prediction which means we predicted a bit earlier.
Red ball marked with an arrow is shorter, so the prediction was later.

This is about 3 ball rotations difference but the system would predict it equally …magic of VB2 ;D.
On most wheels we are better off when the ball is passing first time predicted number since with slower ball VB2 start losing linearity.
When we use constant timer as with VB2 without rotor calculation after applying the time we can always look and compare how much the ball travelled during that time, for example 1.5-2.5 rotations and accept it as valid.

FFVB2 changes time so double prediction may be handy. Surely who has a lot of VB experience he is confident in his judgement but after playing for longer the judgement may change. Small observation to see the difference in between 2 predictions can get us back on track.

It depends on a wheel.

VB2 is flexible, we do not have to predict most of the time when the ball is 1 sec/r. We can predict at any. For example we can predict earlier so the second prediction is 1 + ¼ of rotation form first one. However for the best results we need to keep staying in that range. On some wheels (without strong “the knee point”) we can predict later then one sec and still get good results.
System defines reference time automatically.
Until I set it many times and didn’t find mistake that system made however with stuff like that and all different roulette wheels parameters it might happen so common sense is required.

In addition after selecting FFVB2 I’ll make the system tell us basic reference time with zap pulses.
For example 2 sec reference time, will be represented with 20 pulses.
Next time if we play same wheel perhaps the system can give us 19 pulses.

FFVB2 is positioned at menu 4.

If we just click it we let the system to decide, but I will make it so the user can override system settings to any. For example press and hold the switch at menu 4 and the system counts zaps, we open the switch after 25 pulses and that will set us basic 2.5 sec which might be better if the rotor is mostly 3-4 sec per rotation. Theoretically it is always better to set it for a bit slower rotor speed then the most common rotor speed is. Such options let us have flexibility in setting FFVB2 to anything we find is good for us.

From set up spin the system also learns how long the ball will travel after 1 sec per rotation. We can’t change it but by my opinion it is not so important.
For example if the system defines it as 10 sec remaining time and we predicting 14 sec before the end of spin. VB2 covers the difference except of 4sec x change in rotor speed.
For example if the rotor changes from 8p/s to 13ps the difference is 5 pockets. Mind you that 5 pockets per sec difference is very significant. VB2 will create error of 14-10=4 , 13-8=5
5x4=20; about 20 pockets but because with increased rotor speed reference time is changing, it will increase , in that increased time (probably added around 1.5 sec for such rotor, to adjust for 5 pockets difference x 10 sec to the end is 50 pockets ) because we predict when the ball is faster it will also cross more pockets in that added time which closely matches 20 pockets difference ,reduce or cancel it completely.

Some people ask me is FFVB2 more accurate then tilt2?

If tilt2 is properly set it should be more accurate especially if it is set for later prediction.
If we compare it unlimited predictions it’s probably same.
With unlimited we clock the ball, clocking error may lead to prediction in different rotation.
With FFVB2 we do not clock ball so we do not have errors however, VB2 by itself is not so perfect.
FFVB2 is definitely less clocking so I would say it’s just another option.
Regarding tilt 2 program change, using rotor as an indicator will be removed, its not required since the system can have multiple predictions, double rotations clocking will be added.

After set up spin the system gives zaps to indicate ball speed so we can decide which spin to leave as good.
It is made on very simple way.
For example if the ball is 1345 ms.
The system will zap 3 times and again 4 times.
3 means 300ms , 4 means 40ms … 5ms is not required
So it doesn’t tell us seconds and last digit.
Since they system tells us time of 5th ball rotation from the end of spin most likely it will be one sec + something.
In the case if the ball is less than 1 sec per rotation of course we would know it since first zaps will be 9 or 8.
Zero is represented as longer zap.
For exampole
z z z z =400
zzzz =0
is 1400ms
even if zzzz we interpret as 1 it would be 1410.

I am trying to keep it simple. :stuck_out_tongue:

9

Couldn’t resist so I took unfinished program to test it under live conditions.
At home system worked fine because with my test unit I do not get zaps but beeps.
When I loaded program to my FFZ it worked ok, for start I adjusted zap strength and thought everything is ok. Soon I found that each time I go to set menu I lose zaps and couldn’t feel anything. The system wasn’t memorising zap strength properly and saving it. Since I changed that restore is also at menu 1 it was restoring wrong values.

Anyhow I adjusted zaps then instead of setup I went to play; in play mode after prediction the system will save results including zap strength. Luckily I insert it there for testing. I managed to make it work that way so I was able to use it.
Wheel slight tilt,
As in usual had problem with dealer and NMB. It was early after some time she starts calling it even more early. I was predicting more than 10 rev to go, very fast ball and rotor about 3-5 -4 sec.
During VB2 reference time the ball was making ~3.5 rotations. In usual I predict when it makes 2 rotations. On second zap after prediction the ball was making 9 pockets extra because it was much faster than when I should predict. I didn’t look for that much since I didn’t have time. Even predicting so earlier the dealer liked to stop me with NMB.
But everything went good, in less than 2 hr I was about 500 units in front.
I start playing with 7-8 units and progress to about 20 per spin. On top of that I was making some bets based on DS. By my free estimation about 150 units was won from DS (luck).

New dealer
Slow rotor 5-6 sec but short spins. So short that if I clock full rotor I predict 2-3 rotations before the end of spin, so I had to clock ½.
Of course I stopped playing DS, but even with FFVB2 I wasn’t winning. Everything was opposite. So after about 10 spins I changed where I play, but the ball wasn’t stopping there it went back to normal. So I changed again. I have had many good ball drops but the ball just wouldn’t stop at my sector. It was more related to ball jumps. I was reducing bet amount observing what is happening but I couldn’t get hits for many spins. Everything went wrong. I couldn’t target particular ball revolution but I was more to play at whatever time is left. Predicting 10-12 rev to go 4 sec rotor didn’t match well with predicting 4-6 rev to go 6 sec rotor.
Squeezed with time once by mistake I placed 4 units on 36 instead on 33, and 33 come. Had many missed by one pocket hits. Due to game as that even I was reducing bets I went down by 400 units. It was many lost spins against all odds. On the end come a dealer spinning hard I managed to get few hits but it was time when they decided to close the table. My target of winning 1000 units wasn’t achieved. About testing the system is hard to tell, surely it needs more time.

I have data from set up stored inside ff.
VB2 reference time was set to 2,195ms
Remaining time form 1 sec /r ball was 11,505ms
It looks as very well defined by the system.
From set up data I found that below one sec ball deceleration rapidly drops, such thing do effect vb2.
850ms then next rotation 900ms, only 50ms difference
While ball at one sec was defined as 110 ms and at 1.2 as 200ms.
Because there is not much linearity there on this wheel predictions really need to be within max. 3 rotations range.
I’ll work on program more see if anything else can be done and finish other parts.

10

Did some work on FFVB2, fixed problems with zaps, and some others and wanted to test it more.
Adjust zap strength, moved it to SET mode and proudly walked to the table.
Clock ball rotations but for some reason I couldn’t exit form set. Tried this and that but nothing worked. It looked it exits get start pulses but there was no menu to select. Tried this and that but nothing helped. Ah well there is always VB2.

This time I played $2.5 units. Start placing 10 units then 15,20,30,40,50 per spin. About 70% for VB and 30% as earlier bets. If an average was 30 units for 100 spins played it would be 3000 units placed to gain 500 units. 20% advantage. Not bad but when playing 40-50 units per spins few bad hits reduces bankroll quickly. After few lose in a row have to reduce beets. I do not know what was happening for some time the wheel starts behaving completely different. Spinners, wrong DD hit, if hits crazy jumps, on the end I stayed only with 120 units 4% 8) and I had to go. I was limited with time. I tried FFZ again and strangely it looks as it works. Ok I’ll stay 30 min longer just to see how it will look but the dealer changed. This one was spinning rotor fast; 2 sec per rotation. I observed 10 spins and had no idea where to play so I left. Previous spins were mostly 3 sec, even that was fast for my taste.

The FF program is not finished I am trying to use some parts but it looks as I clicked something wrong then anything can happen.

11

:o :o :o

12

I received picture, my new baby, coming home soon. 8)

On board clocking switch
Plug for external clocking switch, not 3.5 mm jack
Smaller battery 16mm
Power switch
Two holes, for whatever…
Made in China
Made one mistake for robots assembly by 0.3 mm but they fixed it.
All fits under my toe ;D

13

Wow, cool! :slight_smile:

14

You are always the best Forester ;D ;D ;D

15

Nice !

16

New FFZ compared to pen and coins. This is board only it will be wrapped in black rubber.

More details, the only mistake I can see the power switch is a bit too much inside.
Also I do not know to know how will clocking switch work with toe clocking without trying. Maybe it needs to be replaced with higher profile one.


How many ? ;D

Plenty, this people are amazing they even packed them, each one in separate antistatic bag.

Here you can also see golden plates for zap at the back side.
Since it is all across the board they are actually larger then on previous FFZ. Curved shape in middle is made so if we trying to set it with hand it makes better touch with fingers.
I assume using it with toe and applying to much force on board might damage the board but there is nothing we can do about it. Perhaps I may have some new sales or repairs.

Alternative is placing it on leg and use external switch, the unit is so small I can place it inside sock and go to casino in summer shorts.

This is the profile:

FFZ compared to coin

17

That looks really impressive, I would not put it under my toe though :-.
The power switch I would not move, slighly raised switches on small
mp3 players and the such like are prone to getting snagged and damaged,
having it set back from the edge gives more robust design. Its better to
activate switch with the hassle of a pen as opposed to going home
wondering where I put the soldering iron and flux before I can play. 8)

18

Took some time to test it.
Programed it, but no response.
Checked for signals nothing.
Found that one pin is actually wrong connected on the board.
On schematic it is ok but on board is not.

Shit, i made mistake when rewiring tracks last time before manufacturing board.
Instead of pin 4 I connected it to 3. Without pin 4 it can’t be programmed.
Pin 3 is not used, so I thought if I join them it might be ok.
Yep, i manage to program it but still it didn’t work.
Somehow I managed to cut pin 3 from pin 4, only 0.5 mm distance and BINGO it works.
Not easy task to do but still better then manually assembling units.

Pin 3 gives high on output so in software I set it to be as an input.
After that it works and all I need to do is touch 2 pins with soldering iron to fix the problem.
I soldered external switch plug and made a switch for that everything works perfectly.
New FFZ is born.
Perhaps this should be a prototype if it would be manufactured in high quantities but in my case main cost is to set up robots after that the cost is almost just parts per board.

devilish
I wanted to add small switch, previous FFZ was disconnecting power when plug was removed. I didn’t like it. If clocking by foot, main pressure is on battery holder and the clocking switch, two the highest components. 4 mm in total with the board ;).
The power switch is small but it has 7 soldering points and it also goes through the board with plastic pins to support it form moving.

Also this board goes inside plastic heat shrink tube to protect it. I am not sure if I will add some liquid plastic on top of components.
I tried clocking and holding it with hand, not that I like zaps on my fingers but it is reasonable.

The battery is CR1216, 3V

Have to complete new program and send happy emails to members.

19

C’mon, Forester. You know any comments are either personal choice
or preferance because your work is good stuff!

I have the old FFz and would have held off for this one if I knew it was
coming so soon, I maybe learning but I already know who to buy a
roulette computer from. ;D

20

I do not complain, just trying to explain.
I like commenting because it makes things better.

Sorry, by forum names I hardly know who is who and I am not obsessed in tracking people.
I mostly know people who come to MSN for chat or sending emails.
Regarding new FFZ we can talk when everything is done.