Last Wednesday I ran a long result with the neural network script I have been using and was exhilarated by the fact that a near date had been accurately predicted, and had the proper quantity of numbers in my spread sheet. Β As I prepared to announce said numbers, I noticed something was amiss. Β For Wednesday the 31st of October (All Hallows Eve) Β there were 6 main numbers and a bonus number. Β On closer inspection though, it seems, they were not 6 unique numbers. Β The number 22 was listed twice for that (then) future date. Since I do not guess the bonus ball as a matter of habit, I decided in the tradition of any great collage or cut-up artist I would take the results as an oracle to filter and decode. Β The easiest way... the bonus ball replaces one of the 22s, and voila... we have a number for Halloween. Β 21, 22, 27, 37, 41, 59 Β
I figure this way, maybe, we get the bonus ball, at least. Also... I like the pointy Β zig zag of the data plot.
