@Kelly and his views on roulete computers

Kelly says

The only one who has ever produced statistical sound and trust worthy proof of a computer test is Mike Barnett with the NWML test.

Kelly past few months you like to talk rubbish. Deliberately ignoring facts, and I really do not know why. Whatever I do you simply ignore.

You start talking about 2 pin game as America is discovered.
Just go back to video and watch it for 5 min, and you will see explanation about 2 pin game. It is not my problem if from video you couldn’t understand what I was talking abut.

The best roulette computer

For you no matte no matter how much something is explained, you have problem understanding it if it is outside of you frame of thoughts.

Yes Barnett’s computer is ok, did I ever say is not? He never claimed that it is for leveled wheel prediction.

But how can you say I is the only one if even Barnet tells you that FF can gain advantage on tilted wheel but and on leveled wheel as well. Not sure why you can’t add 2+2.
FF computer will perform the best possible in both cases, and in reasonable conditions it can produce huge advantage.

There is no much difference in functionality between his computer and FF , when we talking about tilted wheel prediction.

Except that FF is capable to measure time more accurately, it uses error correction in ball timing, it has automatic settings while Barnett’s computer has to learn from experiments so that time window can be properly adjusted.

But that’s not all, you are here long enough knowing that all people at forum are real, with your public statements you are ignoring their opinion and really discrediting this suite.

Through the stock market I got in contact with a guy in London who is willing and capable to pay up to 50.000 £ for a computer if he can have some evidence that it actually works top notch and that he can get some sort confidence that it will not be confiscated if he is caught with it. But as he says, the market seems pretty empty for his wishes.

***Really, interesting two requests.
50,000 is not required but 1,000 so is it a problem?

For the other request is up to him and his confidence, nobody can guaranty him that. But eve if it does get confiscated, if he used FFA with such money he still can buy another 49 units so problem is solved.

I won’t argue with him.

Well I would, since his statement and your approval of that is very unjustified.

My friend says: "If the roulette computers use algorithms to calculate where the ball lands, why do they then need to click 4 - 8 times on the computer and then relate to the outcome number?"

It is not your friend but you, since you also think that way.
Simply you need few ball rotations clocked so system can get something from it, and more accurate ball timing. FFA 2009 will predict exactly where the ball will hit if used on tilted wheel.

On leveled wheel it can’t be done and there is no point for me to explain it since anyway you wouldn’t be able to understand. If you would then you would already know it from one of previous posts where it was explained.

He has a point there, the computers relate the clicks it received from the ball revolutions and relates those clicks to the outcome number. Now that's a pretty basic mechanical way to sort out received data and I must agree, the algorithms for that, can't be that complicated.

Truth, if we are talking about basic tilted wheel prediction.
But even Fife’s TILT 2 program is not complicated, did I ever say it is.

With IQE6 and leveled wheel prediction is very different.
To reduce possible errors to minimum computer doesn’t calculate function for duration of all spin. Large part after prediction is considered as a constant. As such it doesn’t need calculation but only time length. If that part is not reasonably constant then there is no point of predicting. If it is a constant then there is no point of calculating it. That shouldn’t be hard to understand.

I will not write on VLS since three is no point.
I asked genuine questions about the test and nobody answered it.
Also I noticed that my questions were not welcomed.

You are making a lot of asumptions on things i have said nothing about.

I repeat; Barnett is the only one who has done a thorough test carried out by knowledged people and delivered with the nessecary data on Chi Square and standard deviation. Correct ?

I got no opinion on any other computers than that one and one more that i can`t go in detail with in public. But the only public shown test is from Barnett. Yours and Steves computer might work or might not. I will form my opinion when i see such a test or something similar from you both.

I cant believe you can get your blood pressure up that high just because i maintain a neutral stance until something good or bad is being verified. Those small spin samples dont cut it, there is a reason you need your spin data valued by both Chi Square and standard deviations calculations. For one thing, it puts the data in relationship to what could happen randomly.

The rest of your ramblings i wont even comment on. A lot of your stuff would probably make sense if you had it rewritten by someone whos first language was english. Even this post of yours, i had to dechiffre 3 times to get the meaning of. Jafco gave in good english with good illustrations the beginner a soloution for a modus operandi for a wheel with 2 or 3 pin strikes. Personally i try to predict the diamond but sometimes on diffycult wheels, cuts some slack and go for a “wider” soloution which is pretty much in the same department as Jafcos. But its second choice. But of course you won`t give him credit, that would be the first time you had given anyone credit for anything including Basieux, Scott, Thorpe, Kaisan etc. They are all either too basic, too old time, too wrong or whatever.

PS: I DID manage to dechiffre your VB2, which is basicly a description from Carlo on the Paroli foum some 4 years ago. You apply a specif timeframe anywhere in the spin and read the number below the ball. Correct ? If yes, then why didn`t you just say so. Its an old hat and work in a narrow wheel speed range and a narrow wheel type range, a little bit outside, or a ball that ratters, because of unevenness, the predictions vary way too much if applyed too far out.

Feel free to sue me if i have violated your integrity.

It wasn’t assumption but your text quoted. You only wrote that you are neutral, and follow it with nonsense and your friends not very logical requirements.

Where did you see Barnett’s data, I did not see it? And if I did not see how come I believe that computed can work?
It is simple, because I can understand whole concept and what he is talking about. Data doesn’t need to be presented in your desired format for which I anyway think is useless. And who is stopping you to convert it if you like to look at it that way.

If someone shows you 1:1 hit rate in format of simple graph for 100 spins you can only challenge if data is genuine, but you have no right to deny it because data is not represented with SD.

I still do not see why I would send the FF for anybody to test it.
I did not even send it to Barnett, he purchased it. As soon as I made FFA he wanted to buy it from me but I refused and stopped selling it completely except to members. This time he nicely asked me that he will buy it and if I don’t want to sell it directly he will get it from associate.
Because he is so much impressed with FFA is not surprise for me, because I know it from long time ago and he is not the first one with such opinion.

Common Kelly, on his comment “it is a deadly accurate device” you tried to be funny, instead of trying to understand what does it really mean.

Sure my English is not perfect, but in usual when people do not understand something they ask me to explain. In your case, no matter how many times I explain you still sitting on your ears.

There is nothing more to explain about 2 pin game, it is not part of my or Jefco system but it is an effect on tilted wheel, where FF can take full advantage of it while VB will benefit only 50 of it and remaining 50 will balance first 50% beck to start. Perhaps he did explain it on level understandable for your mind but he explained it wrongly and you even did not notice it. And again, I explained why but you were not hearing it or deliberately ignoring it. No I did not give credit to him, simple reason for that is because from what was represented around I believed it is something innovative. For that reason if I bought the system to see what it is, personally I would be disappointed. I have explained every single point why at http://rouletteplace.com/index.php/topic,580.0.html
If you believe that you can correct me or for any reason change my opinion then challenge any of my statements over there.
However because of some new people who found the system useful for them I did clearly write bout that with positives.

People you named obviously deserved credits; there is no need for me to say anything. I am only familiar with some of Scotts work therefore I could only comment that. About the others I only hear from you and I believe you that they do something good. So how can I give credit based on that.

PS: I DID manage to dechiffre your VB2, which is basicly a description from Carlo on the Paroli foum some 4 years ago. You apply a specif timeframe anywhere in the spin and read the number below the ball. Correct ? If yes, then why didn`t you just say so.

Yes and I did say so, but for few months you didn’t understand it. It works in reasonable wheel speed range, as I was explaining. If you read what I have wrote then you would see that what you commented is already explained over there.
Good thing is that FF with provided data can help user to define which reference time interval is required on particular wheel and FFZ can be set to deliver such time as a reference.

Feel free to sue me if i have violated your integrity.

:D, it may be fashionable for some but not for me.
Some people just do not have integrity.
When Stefano published my name on internet, straight after Barnett contacted me with my full name. But that wasn’t all, he also spit out location where I lived at Mermaid Watters. And now Stefano is talking about privacy. At least Barnett was reasonable; on my request he instantly removed it from document he published and apologized. But damage is done, there is no point winging about it.

Same Stefano is doing now to Bago, as far as I can remember Bago was always worried to use computers in casino, can you imagine how it is for him now, when he finally found out that FFA is much different then Mark or Stefano’s computer that he have owned. I received interesting comments about him from people who know all the story.
With meaning “FFA must be very good, if even stopped Bago from winging”.

I think you misunderstood. These are the data. I have never commented on yours computer at all, except that i have no opinion on the accuracy.

I dont care whos computer it is, but i think its Barnetts, but its this kind of test or something similar i will build an opinion on.

As for Jafco, i wont argue about details. because i dont play that way anyway and if i start a longer argument people might think im a shill for him or on his payroll or whatever, as some people did with with laurance. Maybe you had your expectations up too high, but actually you wrote that you had received it from someone else who couldn`t do anything with it. And complaining that it was an easy wheel. Like i said earlyer, if it was me i would have added a part 2 where i would have shown a diffycult wheel with a large drop zone and how i build the reading spot up from the bottom of spin 1. But i would still have kept part 1 with the easy wheel for education purpose.

Playing a 3 pin game usually gives more than 1 way of tackle it because there can be various soloutions that gives equal end results when all bets are added up. The best play will always be on accuracy.

Kelly I was working on similar job for years. Designing test methods for various products and systems.

This test is OK, but I can see and some misunderstanding, from tester but also from Barnett as well. I did not see details of remaining time; I did not see details of ball scatter. The ball scatter must be very narrow because through explanations of offset, there is a huge difference if data is observed only 5 pockets in distance. Like offset 5 = 15 hits while 5 pockets from there only 3 hits were located…etc. There is no information about rotor speed. Of course this is extreme but what if the rotor was one pocket per second?

Also there was no SD or chi-square representation except of calculating probability with which wheel may be tilted.This may note the best example for your point of arguing.

The offset adjustment wasn’t done on very appropriate way, shifting clocking position of rotor and ball by 10 pockets may lead to completely different results.

If one spin at 12 o’ clock he got ball with ball speed 1195 ms and got +10 pockets result.
By shifting clocking position 10 pockets the computer will identify different rotation and instead of 1245ms would get ball one rotation earlier to 1045ms. He should shift rotor clocking but keep ball clocking point at same position, or readjust time frame.

Since his time window was relatively wide it did not affect him much but definitely it did prediction for 10/37 amount of spins to be potently predicted in a different ball rotation. Most likely it was compensated by same amount from the other time frame end of the rotation by same amount. But it isn’t appropriate way and if for example widow frame was 100 ms it may produce huge differences.

I also know that Barnett is aware that 200ms is to wide time frame and in relity it would be reduced ~100ms and appropriately set somewhere in between 1000 and 1200ms based on experiment and tries. If it was done he may get 50% less predictions but more properly predicted results. So the test is only to give general idea will the system work on that particular wheel with unknown to reader parameters.

I do not believe it is justified from you and from Snowman to insist on such test from Stefano or from myself. I did not see you asking Laurance to do the same. If sales person shows you data on a paper is pointless. Also video spins are pointless as well. The only prove is in words what you understand and believe.

You shell know that Barnett with FFZ one year ago couldn’t get many predictions on leveled wheel. When he approached me second time regarding FFA I told him “man leave me alone I scammed you once why do you want to be scammed twice”. So what do you think made him to insist to purchase and the FFA?

Tests I described

Would reasonably indicate how computer operates. But as you can see neither Stefano or Ronjo took notice of highlights I explained at VLS (Stefano even got frustrated and bomb me with emails and threats, I can only imagine why).

If test is done with reasonable results then everything after that is only a matter of probability related to wheel on which computer is applied.

It is also very invalid for Ronjo to say that overall results are ok as well, since they both claim that all 100 spins have dominant point at one side of wheel. If you want to measure overall results then you need at least close to leveled wheel as it is explained in process of testing Barnett’s computer. So are they trying to prove that Stefano just made computer for tilted wheel prediction or what?

Yes I got Jefco’s system to look at from someone as you described but
I do not remember that I ever said he have had easy conditions when he was using it.

No the results could have been given a value by the testers, but the results are outlined so that it is pretty easy to do it yourself. The scattergrams of 40 - 50 spins that keeps appearing are by far not as informative as the test is.

My point is, whenever you or Stefano or Bago comes with a new scatter diagram and prediction diagram, you will all argue till the cows come home if the chart is random or not. If you made larger samples and made Chi Square calculations you would always be able to refer to those values. The one with the best value would obviously be the best and least random and would also give a bit of professional touch to it. Neither snow, laurance or myself play conditions where we dont know these values before we start to play. None of us is interested in putting money in a wheel we dont know the true nature of. In VB, it doesn`t mean we would always sit down and analyse any wheel before we start, some wheels will exibit a nature we have seen 20 times before and we would as such not be afraid to take it on with a predefined goal of possible winnings within a given period of time and with basis in the defined edge on that wheel.

You know very well laurance can`t provide data on results the player shall achieve himself. Buying a VB system is like buying the keys for a car. Buying a roulette computer, then you buy the car.

As for the Ronjo/Steve test, i got no opinion. As i see it, it is more for Steve to prove Ronjo exists rather than giving scientific worked out test of the hard and software. I think his intentions IS to give a good test, but what can i say. I need a little bit more than what comes out so far to build an opinion. But who am i to complain, i refused to take part in it, in the first place. The way i would have testet it, would probably also have costet me a trip to court, looking in retrospect.

Don`t make me the judge of anything, its just my personal opinion on data results. And i know im not alone on this point of view.

Kelly, why did I start looking prediction vs. position on rotor hit?

It is simple; if I can get that within few pockets accuracy for 50-100 spins when I play, I do not need anything else.

If I come tomorrow and all results are opposite, who cars if I can set computer within few spins and play with new settings.

If I add data for 2 days I get nothing, if I look and play them individually I get something.

The point to understand here is, if computer can predict accurately where the ball will hit on rotor (which is objective of prediction), of course that advantage is possible if ball scatter is not random. After that advantage on that particular wheel depends only on ball scatter, and it has nothing to do with computer.
If someone wants to convert all data I published about FF and convert it to probability, nobody is stopping him.

"Chi Square calculations you would always be able to refer to those values"

Do not say it twice; someone may start using it just for marketing purposes. :wink:

I know that you testing your system for 300 spins before then you play particular wheel. If I have to do that, casino would never see me.

"You know very well Laurence can't provide data on results the player shall achieve himself."

And you think I can. I do not see much difference; VB is also predicting point on rotor, and final result is affected by scatter, so there is no difference.

Reality is that potential buyers of his VB can’t see anything related to success.
No single prove, but when you talk about computers you need scientific prove.

Chi Square, is nonsense.
If I make 1000 spins with all data formats, that still is not prove that someone else on different wheel will have advantage. If I make video just as one above or display graph with hits to rotor, it is more valuable then 1000 spins probability charts.

Reason for that is that data is clear without randomness of ball scatter and easy can be used for estimation with different conditions.

Why I start looking and complaining about Stefano’s video?
By you, it may be valuable advantage, but for me when I see his prediction doesn’t match hits on rotor with needed accuracy, it is nothing.
If you want to believe or “be neutral” that his computer predicting and ball jumps it is your problem.
I can’t blame you for not taking part since I gave up on Stefano-Ronjo-Test as well.

Sure you are very neutral, especially when you talk about cross patterns or 3 pin game. Maybe word “careful” would suit you better. Except when is matter about myrulet.com then lately you show some signs of mild cynicism.

The point to understand here is, if computer can predict accurately where the ball will hit on rotor (which is objective of prediction),
THAT IS ALL ANY OF US NEED.. Obviously Kelly this is key to having an advantage.. if you know where the ball strikes the rotor and you also know from a few observations that the scatter zone is 10, 15, 20 or 25 numbers etc. (but not 37) then you have a tremendous mathematical advantage and it doesn't matter how many bets in the scatter zone you manage to get placed you will always beat the casino after a reasonable volume of spins. You are being paid 35:1 (on a single number) and you know that the scatter zone is considerably less than 36 then clearly you are going to triumph. You don't need Chi, Standard deviation or the mathematical abilities of a savant to give you comfort. I must say I am surprised at your comments given that you are definitely very experienced in this game. Claims made by others that they are getting X hits out of Y events is totally useless. They should be able to say categorically that they can hit their predicted strike point on the rotor plus or minus 2,3 or 4 numbers time every single time that their device 'provides a prediction'. If their device can't do that then all they have is an electronic paper-weight.


NB FFV does it very well.

This is going nowhere. You guys twist my words just as much as everybody else. I stated already 4 - 5 years ago that the point to predict was the strike number and not the outcome number to measure the value of the predicting algorithm. At that point everybody else was talking about predicting the outcome number. There are plenty of cross references besides strike/outcome number that you guys has not even mentioned yet. I wonder why. You havent found out about them or dont they fit the temperature in the kitchen ?

Obviously you have a problem with my opinions so you can close my profile if you don`t like it.

With VB you can`t know if the person is measuring the rotor speed correct or making other mistakes. So how should a third person be able to post data on the performance. Get real.

On the contrary should you be able to post some data on X wheel if the person is able to press a button and the error correction in the computer is good, different persons should get some sort of equal good results.

If you don`t like posting values to your scatter diagrams, by all means carry on not doing that. But in other circles its looked at as pretty amateurish.

But like i said, delete me.

stated already 4 - 5 years ago that the point to predict was the strike number and not the outcome number to measure the value of the predicting algorithm.

We do not know what you have said 4 years ago or did you change your mind.
We only commented what you have said in this thread.

With VB you can't know if the person is measuring the rotor speed correct or making other mistakes. So how should a third person be able to post data on the performance. Get real.

I thought this was about computer testing, and what to display/show as an indication of computers quality.

If you want statistics of overall performance of one of players, or to set control limits then of course you can use something with defined standard that has same meaning for all of you, but that is a different subject.

There are plenty of cross references besides strike/outcome number that you guys has not even mentioned yet. I wonder why. You haven`t found out about them or don`t they fit the temperature in the kitchen ?

HaHa, as you can see, actually you are arrogant and cynic. But only at this site. There isn’t any justification for that, since we only discussing how to test roulette computer.
The only reason you do it is because you know that I am nice and never complain.

Neutral-when not talking about VB and 3 pin game
Neutral- when knows that he can be abused
Arrogant- when can not justify what he said, this time you forgot to say(that is it I am done here)
Cynic-when … lol I do not know you tell me why did you write that?