Has anyone bought Jafco’s system?
What’s your opinion about the one?
I’ve read about the system on the forum, but I’m looking for the real user.
I know only one person but he claimed that ithe system doesnt work.
This doesn’t necessary means that it is a bad system maybe conditions under which he used the system were not the best.
Seller also claims to have roulette computer capable of predicting leveled roulette wheel.
From description he uses PDA, all I can say about it is “dream babe dream”
PDA is same as mobile phone with time measurements, it doesn’t have enough accuracy for leveled wheel predictions.
Then he writes this
“First, the clocker makes a profile (sample) of the current ball action, and the individual revolution times for each direction are logged with a view to matchmaking during playmode.”
That may be applicable for tilted wheel but never for leveled wheel predictions.
What he is saying is that he is entering ball rotations from one particular spin in to his computer.
For example
0.5
0.62
0.75
0.88
1.00
1.20
1.40
1.60
1.80
2.00
2.20
Then from that by measured ball speed to define how many rotations to the end are. PDA will measure ball time in not better then 20ms accuracy (add to that additional average 30 ms human error in clocking), also if he entered different spin ball rotation times he will not find same values.
However if from particular measured time he ads up remaining times of rotations he can calculate rotor movement with reasonable accuracy. About 10 pockets error but because the ball will not always decelerate with same parameters as he entered in to a computer reality will add additional ~10 pockets.
Then, since ball dropping point is not the same as on tilted wheel, it can be everywhere his computer doesn’t have even theoretical chance to define it correctly which practically adds additional 37 pockets of error. This may be slightly less because he may try to find positions in between ball rotations, but it is so inaccurate that combined with previous errors it is meaningless.
You may ask yourself how do I know all this.
It is simple, because I did it.
And I did it with much more accurate hardware that measures time in nanoseconds. It uses IQE6 error correction which brings accuracy clocked time to only few milliseconds. As a approach to tilted wheel prediction even with such accuracy in timing, where I was explaining his 10 + 10 pockets error my program was producing about 10 +5. 10 pockets error is not avoidable. So regardless how accurate we do it we will still have about 10 pockets deviation, simply because of wheel/ball imperfections. Additional 5 is from comparing measured time with recorded spins and defining remaining time for rotor calculation. Prediction where the ball will hit rotor with 15 pockets accuracy is not bad, after ball scatter is applied it still can give us something as 1:30 hit rate. But this is based on TILTED wheel prediction where the ball is dropping at constant place. Applying this approach on leveled wheel and in addition trying to define where the ball will drop is imposable with this approach, even if time measurements are 100% accurate.
The way he is doing it, as you describe it, is the way plenty of teams has done before him including Basieux. Its just not done only on 1 spin but you build up a library of many learning spins that way. The computer will then learn a fall off point (where the final click is done) and make a match calibration when you do new timings. Plenty of US teams did the same. Its the classical way of doing it. I know periferically a member of a team who made 150.000 Deutsche Mark with basis in that. Im not in a position to talk about possible errors that way, i just know that a lot of money has been made that way and that most errors is done in the rotor calibration. The errors in rotor calibration is actually causing more drift when its done with the computer than when its done in VB.
The argumentation against it seems to be the same as what you argue against Scotts system, so i will rest my case before im being accused of making money on Jafco, Scott, Basieux etc.
There are actually a funny incident on the practice CD where the dealer, probably “Jafco”, drops the pencil hat down in one of the pockets as the wheel is spinning. There are some 3 - 4 rollers on the CD. My own predictions comes pretty close to Jafcos, (he writes his predictions on the wheel) some of them better and some of them worse, some of them exactly the same. There are a few where Jafco is off +1 and im off -1 and the correct number is just in between us. But its not the hardest wheel i have seen to put it mildly, if i switch CD to “Teds” Kies wheel with one of the diffycult ball types, one quickly see the difference between a easy and a diffycult wheel.
The argumentation against it seems to be the same as what you argue against Scotts system, so I will rest my case before im being accused of making money on Jafco, Scott, Basieux etc.
LOL , did I accused you of anything.
Kelly I told you that I do not have anything against Scotts system.
Even I promote his work at my web site.
I only say his old material requires a lot from player, knowledge and skill, but it doesn’t give so much in return. My main complaints are fast rotor and rotor used as a reference.
In regards that computer approach.
ONLY AND ONLY ON TILTED WHEEL or taking very small advantage from slightly tilted wheel if nothing changes on the wheel until they collect enough of data. I hope they doing it a bit better then Stefano and using data without ball scatter. But still it is not very efficient way. Because drop point is not constant it will deviate but also it will start shifting by time. Because thy need a lot of data it will come to point where changes of ball behavior and drop point might be greater then what they collected.
“most errors is done in the rotor calibration”
Rotor calibration is something that FF doesn’t need, because of specific design.
The FF is more designed as copying function directly from the wheel whatever it is and calculating only changes.
But for them that is additional problem, and it is not the only one. They will have much more inaccuracy then that. Prediction roulette is as slowing 4 elusions with 4 variables where each variable can deviate by some percentage.
They use PDA which will not provide very accurate time measurements.
That affects all involved calculations
Adding rotations times to computer is very primitive inaccurate way, the system will never have accurate details for ball deceleration in any particular moment.
It weal lead to huge errors.
If computer can not accurately predict ball failing point there is no point in collecting any additional data from there for particular measured ball speeds in relationship to drop point or final result.
So if you telling me that he won on leveled wheel then he was lucky and probably did not tell you about losses or perhaps he did not know what he was playing.
On tilted wheel it is possible and it may lead to close prediction to TILT2 system.
It could be tilted or not, i dont know. Probably tilted. The logic behind is as old as in mid eighties where the first writings about it appeared. But played as late as in 2005. Like i said, i cant say much about possible errors, but i know that the logic works in reality because they pocketed the money all right. They also did it in the eighties and nineties. The computer calculates the “remaining run time” for the ball including fall off by entering at a calculated place related to the learning spins and clicks and adds the average scatter.
Adding rotations to computer then measure time to match one of rotations and add remaining rotations times to calculate rotor is reasonable approach on tilted wheel.
How much you can go wrong? One rotation, and if you are early in time playing slow rotor that wouldn’t be so important.
You can read about similar experiment I did at
http://rouletteplace.com/index.php/topic,568.0.html
But that approach has no chance to predict leveled wheel.
And there is no need to collect many spins except to analyze scatter.
Barnett is not silly man, he did a lot of research, but he start believing that leveled wheel can be predicted when I explained him basics of how FF works.
There were some suggestions to add to FF data analyses as you were describing. Based on FF’s capability to define ball parameters it could be possible.
With that I do not mean simple graph as Stefano is doing, but independent graphs for particular ball speed groups. Then let the FF shift prediction according to that.
That might produce better prediction where there are wheel defects or slight tilt.
But it will be only truth if nothing changes on the wheel. Collecting 1000 spins where data most likely will mismatch after 100 spins is not smart idea.
The computers measures “time of fall” and “position of fall”. Time of fall is expected to have some variation in it because of human errors, where position of fall is a vertical diamond. Time of fall is only half the equation.
The computers measures "time of fall" and "position of fall". Time of fall is expected to have some variation in it because of human errors, where position of fall is a vertical diamond.
Such computer can measure time of fall, but can not measure position.
As you have said.
Position of fail = vertical diamond.
So it is not measured but assumed. And that can be done only if there is dominant drop point (tilted wheel).
Only FF-IQE6 can define position where the ball will drop, on reasonable wheel with accuracy something as 1:12. If we take same conditions and tilted wheel it may be ~1:4. So it is 3 times less accurate.
Time of fall is expected to have some variation in it because of human errors,
It wouldn’t be only because of human errors. But many factors will affect prediction.
Clocking errors, time measurement, spin differences, errors in calculation, rotor errors… Etc.
Barnett’s approach which I like to call classical (based on VB) would be more accurate then that assuming that he positions time frame within right limits.
It would be more accurate because he has time of prediction define so everything after prediction except rotor changes becomes a constant.
The other method I may be better only on tilted wheel if we need prediction to be early as much as possible, and if wheel is with fast decelerating ball so we need to clock it at fast speeds. As Snowman said, do prediction as early as possible so rotations are shorter, therefore even if you miss by one rotation rotor change is small for that short period of time. I think it wouldn’t make difference for VB player because for him it would be same, since it would be harder to guess right rotation when the ball is faster so he may miss by 2 or 3. But computer can do it with reasonable accuracy.
No i meant what i wrote. That is what the computer measures. TOF and POF.
Position Off Fall is not an anticipated diamond but a diamond that is logged when it is hit and is related to Time Of Fall. Its then up to the clocker to measure if TOF and POF keeps relating to each other or conditions change. POF is a fixed thing because you can just look where it hits. Time of fall, is a result of a human who is clicking, with the errors that might bring. Hence the anticipation that TOF varies.
Time of fall, is a result of a human who is clicking, with the errors that might bring. Hence the anticipation that TOF varies.
Don’t you think if TOF is inaccurate that POF will be inaccurate as well?
If he has time of fall inaccurate there is no chance that position can be accurate.
Actually it will be more inaccurate then TOF.
If time of fall is wrong by one sec. on rotor 10p/s you will be wrong by 10 pockets.
But when calculating position of where the ball will fall you will be wrong by ~37 pockets. So in total it is 47 pockets error.
It is example to show you that position of fall is changing more by inaccurate time.
It is because ball at time of clocking is faster then rotor.
There is no point in adding data which diamond the ball did hit if it can not relate to specific ball speed within few ms of accuracy.
Anyway I believe trying to predict particular diamond as position of ball fall might be first logical solution that may come to someone’s mind, but it is not the smartest solution.
The FF is designed differently. It doesn’t care which diamond the ball will hit but which number will be there when the ball hits. Predicting to rotor position instead to diamond is more accurate. Because ball deflection from diamond can be different, if we predicting to rotor point we can take to our advantage ball hits to various positions on diamond which causing different ball deflections. Predicting to diamond is as defining digital state zero or one, Predicting to rotor is finding and levels in between zero and one. If we relate it to roulette it would be as prediction 0,9,18,27,36; and prediction as 1,2,3,…37.
hey, recently having been recommended by another player, i decided to try and make an appointment with john at jafco roulette.
I met john on friday at this south london base of opertaion where i recieved a full 3 hour demonstration of his visual roulette system.
At one point during the demonstration JOHN also showed me his dealer spin where i told him my favourite number (13) which he in the first try he got the neighbour numbers and in the third go he got it straight up!
We did two sessions of actual system play where he correctly predicted that the second sessions would be a stronger winning edge then the first. ( he has done research into the times to play )
He did 110 spins in total with a single neighbour bet and won 37 of them. he also had about 20 that lost by being 3 numbers away from the prediction, which is just 1 pocket away from the neigbours.
The ball was landing on 3 out of four diamonds and bouncing longer than on website videos. He explained how the ball goes through phases of long and short bounces, during different times of play, with the shorter ones being better for system play and also therefore more likely to be occuring at best during his video highlights.
The ball used for the demo was a john huxley/teflon and the wheel was a TCS/Scallop.
Also i want to add that he is not mark howe as people like stefano claims, he lives in london and mark howe is in sheffield very far away! and he is not a scam for sure, i have seen his research and its actually interesting. Im a die hard mathmatecian and i can tell you this that he has put alot of work into his calculations. HE is a very nice and gentle man, true to his words and he is deffinetly worth a try. You all know me as a natural man, and i have experience with stefano, mark howe, forester and now jafco, and deffinetly forester and jafco are the 2 best for to turn to help for. He welcomed me warmly at his base and im returning the favour by giving him the respect he deserves.
Yous.
Few months ago I wrote this.
I can't say it is scam because it will work under conditions as displayed. I've seen it few months ago. Applying basic dealer signature and small adjustment for rotor I have had better and earlier production without timer.
From his video on that wheel he doesn’t need prediction since he demonstrated that he can spin the ball to stop on number he desires. As Kelly said it isn’t very well choice to present the system. It looks more as basic DS, but from what I hear it is better then that.
The question is how well it will work if wheel was normal instead of ACC=330.
I assume it will still do something positive. If you want to get a fair assessment of how well system really works, you’ll have to try it on normal wheel with ACC 130-200 as you will find in casinos.
There was a recent discussion at GG forum about the system, involving
Kelly, Laurance Scott, Forester, geddinthere.
***He explained how the ball goes through phases of long and short bounces, during different times of play, with the shorter ones being better for system play and also therefore more likely to be occuring at best during his video highlights.
Nothing wrong to pick up a bit better results.
It would be good to know why ball sometimes bounces more and sometimes less.
Until we know that, for us it can be only added all together and considered as scatter.
It is well known(to those in the know) that Jafco is Mark Howe, maybe you are too?
He is using that base and payment receiving method to hide his id.
Well, as i said before, I HAVE BEEN TO HIS BASE OF OPERATION!! secondly HE SHOWED ME HIS PASSPORT, which said john, with his picture, so there you go.
When i emailed mark howe with my license and asked him to give me a demonstration he never got back to me.
and i agree forester, the system should be applied to different wheels, but on website as well he did a couple of spins on a cammegh classic, and he is got loads and loads of knowledge about various things affecting the outcome of roulette and we all admit that there are a few factors which do influence roulette, but he has a few extras, which may even put his system above all the others.
Yous definitely is not MH.
And MH’s real name is not MH.
i have talk to yous on web cam he is not mark howe!
Don’t forget MH has a number of associates with links to Canada & European countries.
Yous posts with regards to Jafco sound like they have come straight from the MH BS School.
I am suprised there isn’t a female here saying good things about it.
A contact of mine offered to pay money direct to the Jafco seller and he refused.
Obviously because he didn’t want his name to be known.