Seems like General Cluster II was one of my better systems afterall. i tried it flatbet, with one neighbour with great success.

trigger: wait until two numbers hit next to eachother within the last 10 spins, bet those two numbers and 1 neighbour. 4 numbers bet for 9 spins. flatbet

if a new trigger appear, change bets to that one.

These systems for the roulette wheel by Ignatus look pretty good. Why are there no programmers motivated to do them? It would be good to see the figures in the long term for the systems.

Why begging? I myself will try to make the programming in basic256. This system by Ignatus has really motivated me. It will do as a good learning experience and do as a challenge because it is based on the wheel disc.

http://www.betsoftware.cc/wannawin/code-for-general-cluster-ii-by-ignatus/

Yes I agree.

It is simple and effective sometimes and is flat staking.

It is forgiving in that a loss game can be quickly recovered.

Key to improving 'efficiency', ie effectiveness, ie net gain, ie rate of earning, ie overcoming the negative expectation, is in finding the sweet spot where is the peak of the bell curve or decay curve in distribution of outcome hits.

Is it better to stop after first attempt, or to go to say 4 or 5 attempts, or to 9 attempts?

It is a parallel/ cousin to the Clump Theory which was of course first noted by Vic.

Contiguous wheel numbers are attractive targets and the clustering phenomenon I believe also can be captured ( when it is in a phase on cycle) with a tighter parameter than 10 spins. That seems loose and though the graph is good, more varied tests are needed. I suggest say 5 spins, 4 targets. Give it a shot. Also note the off cycles. What do you do when the wheel is just not co-operating. Find a stop loss parameter or overview trigger that can keep you out of the game / off that table during such a phase.

This is all about knowing the bet characteristics.

This is a profitable methodology when applied with discipline and knowledge. Research needs to be definitive and the RNG samples of millions of samples can be useful.

If anyone is interested in having ideas tested, and access to over 22 million spin samples for testing, please write PM to me and I can refer you to the source. For ethical reasons I am not publishing that link here directly.

I have my own live spin samples for my own private testing. It is prudent to source your test data from several different locations, sources, samples. Results can vary surprisingly. That is why the large number spread has definitive role.

My own private work about which I have written many times uses 4 to 6 attempts on 9 numbers and flat stakes on cluster analysis.

That can be compared in effectiveness to say 4 numbers on 9 attempts. I am sure you will find that most effective duration, through testing by trial and error will be 4 to 5 attempts. testing as in a laboratory environment needs to be systematic and logical. To an analytical mind like mine the test process is exciting. It is like a search, a hunt, a quest, and sometimes may be rewarding. Othertimes not. You learn to deal with disappointment as the finds more than make up for that time and loss.

Key will be to read the true nature of short cycle phenomena and find most effective stops and starts and editing seams so as to keep that graph generally upward trending. You are looking to dampen variance exposure.

I have written quite fully on Clump Theory and application on my Blog.

The ongoing research and exploration by Ignatus is good work.

Thank you !