This... I do not have a better way to describe it. I notice it most prominently lately as I cycle the same combo for weeks at a time.

This is what I am trying to figure out! The Match 6 in PA is a great example of the churn... you can watch night after night and see the results come from everywhere all at once... hots/colds/neutrals/evens/odds/high and low... but they shift within a few dozen draws to show a different set of hots and colds.

Now we could chalk it up to randomness, but the churn is the only consistent part of the results. My classification experiment hoped to model some of this by mapping out the hots, colds and neutrals over a limited subset of draw history.

This churn, different groups of numbers cycling through in some "as yet to be deciphered" process are noticeable! Some weeks a 48 might come up 3 out of 7 draws, sometimes it is mixed that same week with something like a 21 or an 8, that also shows up 2 to 4 times in the same 7 draws, then BOOM... they vanish for awhile.

I am able to pull all sorts of stats for individual numbers, but have no concept of where to start trying to model this "churn"...

It is there from the power ball down to the pick 2..i have a feeling part of it will come down to sample size, which will have a different value for each game type (pick N, pick 5/60, pick 6/49 etc...)

I need a mathematical basis to describe this churn, only then can I begin to truly craft an algorithmic solution...