During development it is a must to design test processes, so we can be sure that individual parts of product will function properly. When everything is done and tested positive final test of functionality of product was required. Internet world of roulette is full of scam sellers and technically incompetent people. Of course that they don’t like me, for publishing test procedure, testing their product and publicly writing about it. Since then they do anything to discredit my work. Writing false information about me, about my system or even paying hackers to hack my site. Well, site is fixed and it works better then before therefore there wasn't much harm except that we lost few posts in forum section. Here I will explain process of testing roulette computers with better explanation.

Some computer owners used my test and applied it on different roulette computers. After disappointing results they complained to seller. Sellers as sellers, they took money and they can not fix product so they had to discredit my test. Oh noooo, nobody does that to me. Therefore I will explain test in detail and add few pictures for better understanding. If I find time I will even add some video files. That should be enough so even average Joe can understand and see by himself lack of knowledge in statements of whoever discredits my test.

I will comment on few new pages created by Stefano Hourmouzis. Stefano is guy who is smooth talker, selling few pages of paper system Genuine Winner for $2000 or roulette computer that starts with price of $5000. I’ve looked and his paper system and his computer and I honestly I cannot say which one is worst because none of them works at all.
Stefano is most dishonest guy that I ever met since I am publicly involved with roulette advantage play. Recently I had opportunity to exchange few emails with him. While Stefano’s computer is far from anything applicable to gain advantage on roulette I would have to admit that his marketing scam is close to perfection.

Stefano is also owner of http://www.roulettesystemreviews.com/ the site spouse to be genuine but he used it for years to discredit competitors specially Mark Howe. Recently he removed all links to my site probably believing that not many people will visit my page. If I continue explaining everything it will take pages of writing but short description about Stefano’s system and his character you can find at http://myrulet.com/forums/index.php?topic=175.0 or search the net for more.

I’ll better focus on testing and explain why test is good and important.
We use single spin for simple reason. Because we know that ball at any moment within that spin is leading to same result. When we know that for any ball speed combined with rotor calculation result is same, we can check what will happened if we measure at different moments ball / rotor speed and we can easy check will computer calculate it correctly because we already know what to expect as result.

On the left hand side we have picture of spin where the ball is above number 11, and on the right hand side is same spin but about 1 sec. later in time so the ball traveled in CW direction for 3/4 of spin and now it is located above number 30. If we measured ball speed at position when it was above number 11 and calculate it in combination with rotor movement the result MUST be same as if we do same when ball is located as on picture 2. Therefore if computer can accurately measure and calculate, it MUST produce same result no matter where we do clocking.
Stefano Hourmouzis comments on my test.
Testing on different diamonds is pointless because you only ever use one diamond for predictions, and you can tune the computer to rotate the "final" predicted number by any number of pockets. The tuning can actually be automatic with my computer, or manual. If you set to automatic, my computer will automatically select the most profitable calibration settings to use.

It is doesn’t matter will we use different diamond or not, this way we are testing computer. Computer does not know from where we took timing sample. All that computer would know is ball and wheel time per rotation therefore with this test we deliberately give to computer information with which computer should produce all the time same result.

Let’s say that on picture 1 the ball travels with speed of 1.000 sec/ rotation. On picture 2 the ball speed is 1.150 sec/ rotation, there is difference but final result is still the same because it is same spin and computer should produce same result. It means that ball of 1 sec above number 11 will end up on same number as ball of 1.150 sec when it is above number 30. Someone may get confused here because position of ball is not the same. Therefore lets look what will happened if one spin the ball clocked at top diamond is 1.150 sec. number under ball is 30 and rotor is same speed as it was with previous spin. If we put it all together we will end up with something as on picture 3.

This is same as picture 2 but rotated so it represents same clocking point as on picture 1 but rotor position and ball speeds are same as on picture 2. The question is will this ball end up same as ball from picture 2. Of course it will because it has same speed and same rotor position therefore and ball traveling time will be same. Whit this I am trying to explain why with testing, clocking on different diamond is important because we give to computer different ball speed and checking how good calculation will be. We can do it because we already have result to compare with.

Further more you can do test by clocking all the time at same position but 1 or 2 rotations earlier or later then previous spin. The computer still should produce same result.

Stefano's claim that his computer can automatically tune is total nonsense. He actually refers to selection of pick point created from few spins, but how can you get pick point if computer cannot correctly calculate. Every single spin computer will probably have different ball speed to calculate. When we do test as I describe we give to computer ball speeds for which we already know result, and if computer can not accurately measure it and calculate according to measured data any further adjustment by rotation as he explains is pointless. You simply cannot rotate result for every single spin. Anyway if you have to do it manually, every time different for each possible ball speed then what computer is doing?

Now lets look what Stefano saying about my test and what he suggests for testing. You can find all of his explanation at http://www.stefanohourmouzis.org/

Stefano is referring to my test as linear test. How it can be if test is done on roulette spin. Roulette spin does not have linear parameters. Therefore if results after test are correct they are definitely not results of some king imaginary linear test but results from parameters created on roulette wheel. If computer tested under those condition calculates correctly then calculation definitely is not linear.

Further more he says that in my test I am giving similar inputs therefore and outputs should be similar. Well I don’t see that at all. With my test I am deliberately doing opposite. By clocking ball at different location or different time moments I am giving to computer very different inputs. Infect input is different as much as it can be so we test computer on wider spectrum of ball speeds. At his page you can find video where he trying to demonstrate my test. Of course he used only one clocking position because he knows that his computer can not do it. Even with what he couldn’t get good results. He actually used tilted wheel prediction approach.

Here is Stefano's suggestion for testing.

TEST 1: By manually inputting data into the computer and seeing the actual output of equations. You can simulate human timing errors, timings with no errors at all, basically whatever you want. I'd consider it imperative for a computer to have manual data entry feature whereby you can enter whatever timing you like to see the actual output of equations. If a computer doesn't have this, then how on Earth could you possibly test the computer's algorithms? How would you know the programmers and mathematicians have done their job properly? Neither Mark's or Forester's computer have this absolutely essential feature. Why not? Such a feature is required to test the exact accuracy of the algorithms. What have they to hide? With my computer you can perfectly see the accuracy of the algorithms with this manual test feature, yet they have the audacity to make false claims that my computer's algorithms are inaccurate.

He suggests manually inputting data. How on earth with manually inputted data you can have any idea did computer even measure time correctly? For purpose of accurate measurement I use the fastest programming language and special functions of microprocessor which really can be controlled only in machine language. Therefore timing is very accurate. Stefano uses Nokia platform where he has no idea about processor commands. If we enter manually data in computer of course that we will get something as output. But result by itself is pointless since that result we have nothing to compare with.

I even doubt that he uses interrupt futures of processor when timing the ball. If this is not used timing will be inaccurate and will have variations in result based of how much processor is loaded. Additional source of trouble maybe switch bouncing during clocking which also needs to be properly handled. There are many other problems which are not tested at all with his suggestion. He probably is not aware of them or he simply does not know how to handle them. Whatever it is his test definitely is not covering it. I really do not see that anything is tested with his test except that when you enter something in you get something out.

I will not comment his frustrated statements about me. If you want to find out why he is doing it and how much it is incorrect it would not be hard if you do some search and find out how conflict really started.

I also run closed forum for all users of my system where there is much shared information. That includes a lot about Stefano's computer and his paper system. Information is not provided only by myself but by ex users of his computer as well.

Some people were lucky. They were planning to buy Stefano’s computer but after some searching they come across my page, bought my device received access to support forum and definitely made their mind about Stefano. It saved them thousands of dollars but most importantly they end up with much better device that really does accurate prediction.

I think Stefano's computer does not have prediction at all because it is probably based on his paper system theory which is really dealer’s signature where he estimates everything believing that average of estimations will produce pick point. With his mobile phone that he calls roulette computer he probably does same but selecting only spins which are more similar. On leveled wheel that will never produce good results because for accurate prediction much more is needed but most important is that computer can measure ball speed within few milliseconds of accuracy.

Stefano expect that every wheel has some degree of tilt which may help his prediction and produce result better then 1:35. That would also explain why his computer could not produce any accurate prediction based on clocked ball and rotor speed. Dealer signature will not work on leveled wheel or slightly tilted wheel in real environment. But on kitchen table with taking care of rotor and ball speed and with slight tilt it may produce small advantage specially if demonstrator uses reverse engineering and after test is done points to pick point. That would also explain why Stefano could not do just in time demo.

This test I will also recommend to every person who is trying to develop his own roulette computer. I hope you find this information helpful. Coming soon.....advantages and disadvantages of roulette computers...