Here is scatter for tilted wheel prediction.
Left side is distance from prediction to rotor hit
and on the right side is distance from prediction to final number.
Left side looks great, but how much real advantage will be it is dependent on how wide scatter is.
The test was done with 1/2 wheel clocking in accuracy 2 mode which would be about 8 sec. to the end.
Right side of graph has 2 pick points one created where the ball mostly stops instantly and the other one is when the ball bounces across pockets. It takes a lot from us but still there is nice advantage, and on some wheels it will be even better.
I managed to rewrite program to suit new microprocessor.
There are some changes, in options of system select.
Also security is added where the device can be completely disabled and protected.
Main change in program is on tilt 2 systems.
It uses to predict about 50% of spins, with this test it was 85%.
I could not stand device not predicting few times in a row therefore I did some changes to get more playable spins. It may result that small percentage of spins gets by 1 rotation wrongly predicted but I still like it more because total gain is greater then loses from possible mistakes.
On Iqe6 there is not much change except that new processor has additional 3 independent timers which could be started or stopped independently from main oscillator so I used them to achieve better time measurements.
On request of some extra accuracy setting is added, therefore prediction can be done even earlier. However I would only recommend it to be used on older kind of wheels where the ball decelerates faster. Because with Iqe6 time of prediction is close related to ball speed on older wheel even if we select earliest prediction it may be only 6 sec to end. Also with earlier prediction and time difference in between rotations is smaller therefore it will easier fit within systems specification.
There is new program on the chip as well, which I will explain later.
The chip is available from next week for $50AU which basically covers cost of chip and postage. Or make it $100 to get additional 2 switches and new chip.
What is the purpose of the new program? Is it for level or tilted wheels?
How do we go about ordering the new chip? Can we use Paypal to order from you?
Thanks,
Ptblltrader
FF was accurate as much as it is possible; there is not much improvement in that. Maybe only small improvements because of better timer usage.
Main change is that tilt system predicts more spins it was about 50% ,now it is more then 80%.
50% may look ok but it is not, same as black and red have chance of 50% but many times you can find results as RRRRRRR it could happened to system that when playing you do not get so often prediction and that could be very frustrating.
Bottom line is that iqe6 was progressing from E5 where parts of program were changed and added, with new program I already know what needs to be done so I would say it is done on better way.
Starting procedure is different so we can select system and change setting without touching device. Now FF program goes in loop until you select what you want. Accuracy has 4 levels, extra earlier level is added which should be used only on wheels where the ball decelerates faster. All that is needed is to switch on the system before play anything else can be done at any time.
Also security is added so system can be permanently locked, where it will not work as roulette computer. Do not worry it is had to accidentally lock it because you will have to handle strong pulses for few sec. It is if and only if you have to do it. Due to some things going around I start thinking more about security. Blocking system on time is very achievable because if in trouble you do not have to say, just a minute I have to enter code to delete program. All that is needed is to hold switch pressed as when resetting system, and handle few sec of warning strong zaps.
And finally something good for sooky babes as Mike.
New chip provides full zap strength adjustment by software.
It is preset to my comfort if you do not like it you can easy select one of 14 zap strength levels. It starts with weakest and gradually increases until you select one.
Paypal is good. ;D
And finally something good for sooky babes as Mike.
I would say super sensitive guys like me. ;D ;D ;D
lol i even do not know did i spell it right 
Yes that’s how you spell it!..em eye kay eee. ;D ;D ;D ;D ;D
This is what I just got on 50+ spins,
I did run test on Stefano’s first published video.
Used IQE6 accuracy 2, prediction is 2-3 sec earlier then on video.
Full rotor clocking.
I had about 10 spins not predicted so I repeat them to get result.
Sometimes it is hard on video spins.
Left column is prediction to rotor, and right one is prediction to final number.
Data prediction to rotor is only estimated, the best I could see without stoping video.
Forester
The prediction to final number is better than the prediction to rotor?
Is this because of the luck factor?
When you do prediction to final number what Offset did you use?
Are you using the same Offset for both spin directions or is Stefano’s video one direction only?
Mike.
Yes it looks ok; it may be with some degree of luck.
It is as it is I did not use any offset.
I only finished some changes and tried it more to check amount of spins predicted.
If scatter is not too wide as on this wheel FF may get results as that because hits to rotor (hit rotor with narrower angle and roll more) at positions around €“ 15 have higher chance to bounce longer then hits at position -9.
If ball is hitting higher parts of diamond it means that hit is earlier in time and the ball will make greater distance from diamond to rotor, therefore final result matches closer hits at lover part where ball almost stops instantly. I tried few spins trying to predict only to diamond as I did on previous graph about IQE6. It did not look so good even last time it was ok. So I gave up on that idea and left it based on this theory.
Ok this is additional spins on same wheel,
Some xxx’s were lost in my messy work.
But this is abutting it.
I had about 20% unpredicted spins that is what worries me the most.
Some spins I repeat 2, 3 times and there wasn’t prediction. It is probably switching in-between video frames. When I move position slightly I get prediction.
Also for some reason wheel slightly changes parameters.
After finishing all spins I went back to start it took me 4 spins to start getting prediction.
Simply the system was following wheel and gradually tuning, but on return to start, data did not match specification of the wheel from end of spins. When system is tuned it has very narrow limits. It took few spins to get data from spin within limits so the system can tune back.
I will try to find out if it can be better.
If I played 7 numbers it would return for every $1 played $1.74
If ABC plays with his 15 numbers sector it would return $1.43
But if we play unit per number on the end I will profit 771 and he will profit 963 and have less stress. Sure if I cover 7 numbers with 15 units (same amount) it will be much more.
Simply the system was following wheel and gradually tuning, but on return to start, data did not match specification of the wheel from end of spins. When system is tuned it has very narrow limits. It took few spins to get data from spin within limits so the system can tune back.
I will try to find out if it can be better.
Forester
I don’t think we want the FF to do this? In real life we want it follow any changes that may occur on a real wheel caused by things like environmental effects, temperature, air pressure and dealer effects like sweaty hands.
It doesn’t matter if the tuning changes the settings in a series of video spins - we don’t need it to be correct when you go back to the start of the video.
Mike
You are right,
Maybe I did not explain well.
I think changes were not so big and system should readjust easier.
Perhaps only defined limits are too tight.
The system will never be able to follow changes 100%, only up to the limits other way it would be to risky and cause more trouble then help. :-[
Ok this is latest what I have and I think it is final.
The first column is about 100 spins where rotor was fully clocked; next one is 1/2. I do not see much difference.
Last column are results combined. It is from prediction to point where the ball hits rotor.
Accuracy setting from 1-4 is 3. Where 4 would be the highest. Please click on image to enlarge it.
This is same spins but the graph shows consistency from prediction to winning numbers.
It is good as it gets.
From previous graph to this one we can’t do much.
This graph represents consistency of prediction in time.
It shows stable prediction during all time.
Please note that top pick points are not far from bottom part because of circular wheel.
[attachimg=1][attach=2]
And finally this graph showing prediction errors.
I did 2 times same spins and this is difference of predictions.
After 100 spins probably I created and more errors, but all together it looks ok.
About 80-90% of spins were predicted at first time.
Some spins I had to repeat few times to get prediction.
Reason for that is probably that they are for some reasons compleatly different in ball behavior or simply video frame switching at some clocking points created timing errors outside of system limits. During test I did alternate clocking points in between 2 V. diamonds.
At one of previous graphs where Mike told me it must be with lot of luck, I think I made mistake used wrong column, or something as that.
I am still trying to find why this glitches.
Results looks good. :
I do not want to push you.
You need time, we need patience, but please let me ask question;
When can I order new chip? ???
best,
abc
P.S.
One idea about security blocking system.
In my opinion before system will cancel itself extra confirmation is neded.
I had once situation … truly nice situation, right after profitable night in casino placed on the see shore, I went out to see sunrice on the sundy beech together with young charmy girl who I met few hours earlier…
Later on ( much later ) I noticed that my switch was locked by sund ! It was locked in on position !
Forester please add extra confirmation by extra klick, or kilcks combination right before system will be permanently locked.
There are situations in mens life that there are not enough blood in brain to think… FF shoud resist it also. :
Results looks good.
I do not want to push you.
You need time, we need patience, but please let me ask question;
When can I order new chip?
Some people ordered long time ago, and now I feel bad because I couldn’t deliver it on time. Nobody really complaining but I still feel bad and working hard on it. Testing and debugging is really time consuming.
I found and reason why gliches on last graph. The system shouldn’t tolerate that so next result should be even better.
Do not worry about locking system, hardly that it will happen by accident.
If it does there is way to unlock it, but I will help only people i know something about. The other ones will need to by new chip or convince me that they are genuine players.
If in trouble less you would want slow process of locking with confirmation.
We will see how it goes, maybe later on It can be option.
Man what did you do to poor girl. You must be pushing very hard if you managed to lock the switch. ;D ;D ;D ;D ;D ;D
Forester,
I pushed gently as always, and as usually was unusually. This is a matter of proper… wheel.
abc, 8)
Procedure of locking system is the same as broken switch in on position. Player can recognize this fare to late !
best
The reason I politely avoid it is small chance to happen and fact that I am already late.
It can happen only if switch gets locked at particular part of program.
It will not happen if switch is locked and you power up the system.
How about if I add that after switch is pressed and after warning signal ends you need to open the switch within 2 sec?
Be aware that with this if you forget to do it you will need additional 12-16 sec for another attempt to do it. The problem is that you have no chance to practice with this.