The whole concept of these lottery numbers is that the numbers themselves are nothing more than a marker... a 5 has no higher or lower value than a 1.

We know that a pick 3 game has the numbers 0 to 9 available in each of the 3 machines, so 10x10x10 = 1,000 possible combos for which they pay out about half ($500 for $1 straight in PA). They could have easily used letters... imagine a pick 3 with A-Z as the markings... 26x26x26 = 17,576 possible combos... using the same logic, the payout would be $8,788.

The point being that the numbers themselves do NOT have mathematical properties, though we keep trying to search in that space for a solution.

When I made the move from number manipulation to frequency statistics, it painted a picture that was different from what I looked at before, however the results so far have not indicated any better way to generate a pick, because the basic truth is that the highest frequency followers are not always picked. They will be picked most of the time, but generally not at the same time...

So far, I have been able to generate a list of the distribution of all "numbers" that follow the last draw... this runs in seconds and can process all games. Then, I did the same for the angles between draws... again being able to get a distribution list for ALL follow up angles based on the last 2 draws... if the "numbers" do not have numeric properties, why would I imagine they would have physical properties? They, in reality, do not...

Today I will be reworking the input files to include both number followers AND angle followers. I am honestly unsure what good this will do, but I like to keep sharp with coding so why not...

How many times will I do this? How many ideas that deep down I know will not work will I try? Starting to think that I am just a moron with some coding skills that refuses to admit the impossible nature of trying to predict "random" results... or I am just an avid hobbyist...

Either way I seem to have a darker outlook on ever making headway in this endeavor.