In professional play I often use 3 sets ad other overlapping techniques but I completed my study on 30 independent sets of 100 spins.
Whether I play one set or three the rate of earning is + 0.723 ( average) per spin per set.
The reason my bet wins is that I use 'Cluster Analysis' identifying clustering tendencies into 4 groups of 9 numbers ( with one number excluded).
I was trying to establish what is the 'edge' in my favor.
I am totally sincere.
I live in New Zealand.
No doubt you live in New Zealand and you are sincere. I envy you living there :o)
However, unless you have found an unbalanced wheel, your system is of no consequence. It does not give you an edge and it goes no way towards defeating the house edge. I smile when you call yourself a professional player. You have won over a few spins. that is all.
How you can arrive at some of the statements in your response beats me, rather like standard responses to annoying cliche trivia questions. But my question as I state is sincere.
Never mind, I hugely respect you for your administrative work and your team's fabulous website.
I certainly have no intention of over-elaborating on my work or details of my work which was originally produced in collaboration with a professional statistician all those years ago in Sydney. I do not think you would smile at my stating his profession. I stated my original question on the basis of one bet per spin (level/ flat staking) where I target 9 numbers on a 6 spin cycle or until hit, and stop at that stage. Play resumes after a suitable statistically proven pause and trigger to enable a second 'recovery phase' where all or part of the initial loss is recovered.
Depending on whether hit on first spin attempt after target pre-selection or not, outcomes are...
0r if fail -54 units and pause
Recovery Phase consists of 3 independent strikes on the re-appearing target so that at best 3x +27 can be achieved, but generally the re-tracement loss damage is repaired. This helps net returns obviously but the effectiveness of my method lies in the ability to recognise clustering and takes advantage of the change of state between cluster and non cluster.
Also I play up to 5 sets independently ( 5 attacks such as Wheel, Finales, Streets etc).
The flat staking was done for ease of calculation purposes for the test. This test repeats test I conducted 5 and 10 years ago which showed similar results but I have now added further refinements through experience.
In live play I target very short cycles only and use a short progression for maximum efficiency. I pause and stop frequently. The flat stake test ploughed on 100 spins regardless, so my usual play is much more efficient.
Nevertheless the results of my 30 session test on 5 x 100 spins shows a best result in Finales Set at around + 1.5 units per spin. Other sets were less effective but all positive. The average of the 5 sets was +0.723 units per spin/ set.
My technique is completely outside of conventional roulette play and uses Cluster Analysis Theory from my work in Statistics.
My notes above will be meaningless if you are not prepared to accept that there are very different ways of looking at roulette data and in fact any randomly generated streams of outcomes, bring short cycle order to apparent randomness.
Thanks for your responses and I hope this reply has some interest for you and readers.