Although I have been recently obsessed with creating this classification system into a mobile app, that does not take away the fact that this is a data presentation tool, and still needs to be explored for interpretation.

The simple process of sorting the results by frequency took the emphasis off of the digits and put it on the frequency patterns directly. These tables should be saved and compared with each cycle. One such analysis tool could be to capture the classification table, then overlay the 21 cycle draws to determine if trends are present or absent. For the next cycle, though only playing 1 combo, I will record and track the performance of the 10 combos generated by reading across the table, and also the floor and ceiling of the neutrals, if not contained in those 10 combos. Simple as recording the combos and with each, a simple x in the next 21 cells across to indicate a hit. Paper play on a larger scale to help find the best performing area of the result grid. I do not have to wait for the GUI to start capturing data, and paper play pick 3 can be run regardless of which game is in the current cycle.

Once the steps are known, then a back test can be figured out. So there will still be plenty of experimentation ahead...