Roulette Prediction Devices and Techniques >> Second stage testing of FF device


Second stage testing of FF device

Roulette Prediction Devices and Techniques > Roulette prediction devices >

Message/Author

Survtech posted on Thursday, March 13, 2008 - 06:05 am

The second stage of FF testing is complete.

Stage 1: Test on wheel with drop-zone

I carried out this test personally.

The device performed well. When used against a wheel with a drop-zone and manageable scatter this device presents a significant threat in the hands of an experienced operator.

Stage 2: Tests on casino quality wheels under controlled (covert operation not necessary) conditions and filmed using high resolution professional equipment.

These tests were carried out by several competent and experienced people. I intentionally did not perform the tests on live random wheels because I had a preconceived opinion that the methodology and it's application were unsound and did not want to influence the outcome or testing procedure. I believe these tests were conducted fairly and by competent people with an understanding of the principles involved. The tests were carried out in a casino environment on current operational wheels but without the need to operate covertly, this situation, I believe, was close to ideal. Given that the testers did not have to act covertly and could place themselves in an ideal position to observe the wheel, I would expect the testers results to be considerably better than could be achieved in a real-world situation.

Tests were carried out on several different styles of wheel and a variety of ball were used.

Whilst the device produces a satisfactory percentage of predictions when applied against digital video footage of wheels which have been retired from casino service, the percentage of predictions against live wheels was consistently disappointing. I don't believe the low prediction rate was the fault of the testers, the manufacturer of the device also reports low prediction rates against modern wheels.

When the device did produce predictions the results were close to random. The testers were unable to obtain a statistically significant advantage on any of the test sessions.

Whilst there is much talk of the device, it's capabilities and it's continuing development, at this stage it is just that, talk. I do not know of anyone, including the manufacturer, who claims to use this device to gain an advantage away from the kitchen table.

It may well be that this device is capable of gaining an advantage against some modern random roulette wheels but there is no evidence of this to date, and we, with the best will and incentive, were unable to produce an edge.

Time will tell if there is a viable threat to modern well-maintained roulette wheels but at this stage the FF device is not it.

We will perform one more round of tests.

Survtech posted on Thursday, March 27, 2008 - 08:42 am

I have received the report form the testers of the roulette prediction device known as FF.

The report is complete and professional; tests were carried out in a similar manner to those carried out by the National Weights and Measures Testing Laboratory under instructions from the British Gambling Commission.

The British tests were carried out on a new Saturn wheel with CNC machined rotor, supplied by the manufacturer.

The Australian tests were carried out on wheels currently in use on the casino floor.

The complete report is copyright to the testers but they have indicated that they will conditionally make it available on request.

The report is very detailed, covering all aspects of design, construction, ease of use, setup, deployment and practical considerations for covert use in a casino environment.

The report contains all statistical details, including wheel overlay graphs of prediction and strike point, standard deviation and scatter analyses.

As in the case with the NWMTL report, this report will not be made available to the public, however, the following details summarise the testers findings:

Predictions were obtained 51% of the time.

Strike point was predicted approximately 20% of the time.

Scatter after offset was average over an eight-slot envelope 42% of the time.

The device has not shown that it has sufficient, if any, predictive capabilities to overcome the other factors, which govern the final result after successful strike-point prediction against a wheel with random ball exit-point characteristics.