Roulette Prediction Devices and Techniques: Roulette prediction devices: Further testing of roulette prediction device Tuesday, May 12, 2009 - 03:39 pm: Last year I purchased a roulette prediction computer from MZ Electronics with a view to testing and reporting on it’s effectiveness. I carried out some tests on a wheel with a very good drop-zone; the results were excellent. I then handed the device over to a client casino for further testing on a new wheel, which had not yet been commissioned.
A synopsis of the testers findings can be found here Last week I purchased another device from MZ Electronics. This device is far more sophisticated than the first device in its operation and operator feedback. This device has a synthesized voice output, comprehensive menu system and Bluetooth communication. The device was tested on a modern Huxley wheel, which was displaying a solid drop-zone. The device was set to play a drop-zone wheel and calibrated for time-to-drop. The device provided extremely accurate predictions.
The device output only one error (would not produce a prediction) over several dozen trials. This is vastly better than the previous tests. This device would provide a very significant advantage to its user against a wheel with a drop-zone and manageable scatter. We then adjusted the level of the wheel in an attempt to remove the drop-zone and create a random ball fall-off point. The device was then set to play a so-called “level” wheel and the time-to-drop was set. The device provided predictions on every spin without error. We observed that when the device said “now” (this is the point at which the computer predicts the ball will fall from the rim) the predicted number was directly under the ball. On some occasions the ball travelled one more revolution after the device said “now”. Had the ball not travelled one further revolution the prediction would have been accurate.
The device was providing predictions nearly twelve seconds from the end of the spin. We will carry out further tests but I strongly suspect that the drop-off revolution will be predicted with greater accuracy if the device is set to perform calculations one or two revolutions later. It should be noted that we were only interested in the predictive accuracy of the device (which was very, very good) not the end result, which is influenced by the behavior of the ball after it has entered the rotor. This is a well-made professional device. I understand from its manufacturer that it has sophisticated error correction algorithms designed to smooth out operator clocking errors and prevent the unit from providing a prediction if the parameters are out of specification.
This device is capable of providing its operator with a significant advantage under varying conditions unless consideration is given to the ball behavior