WIN D... the 3 digit IS the barn, the penny's center would be the big jackpot games...
I think that the reason most systems fail is that the initial design goal is flawed or not focused enough.
MikeK... In the past, I have tried to use the forecasting functions in excel, but after several dozen draawings, it tends to align itself with the arithmetic mean of the numbers (average). Even in experimentation with how many draws to include, results that were positive were not repeatable. I have also tried an error-correction forecast system... take the forecast from the drawings (reverse solved for past draws) and subtract it from the actual results for that draw, this yields a table that now includes negative numbers, then, using that as source data, try to forecast the error correction number... forecasting in excel does not like negative numbers, so that was yet another failed theory (extracting next draw data from the statistics of the previous draws).
WSLS... the bag results would be purely coincidental ;-) when guru first half-explained the lottosynch theory, I thought it was the most unique new approach in a long time, so I coded up something in VB to play with the theory (as I understood it at the time). Initially, It was a simple loop that would use the RNG to pick numbers, then compare them to the draw history and write the corresponding seed value to a text file... after running a 2,000,000,000 count test it became apparent no single seed would do the job (this is for powerball's 5 white ball combo). I converted it to the pick 3, and found that it could repeatedly match any 3 draws in history, but when just inputting the last 3, the next projected draw was always wrong. Of course, VB's built-in RNG is no reliable source of entropy, but If it was, the results would NOT be repeatable. I have since dubbed the program "Coincidizer" because any positive result would be purely coincidence.
if there is to be a first broad stroke, I think it should be a way of differentiating a false trend from a genuine trend AND determining which statistics are meaningful. In the big picture though, any system that generates more than one pick for the next draw is doing nothing more than guessing, regardless of how educated that guess may be... but if the end result is just one number to play you end up saving alot of $$$, win or lose.