depending on the nature and amount of picks, just about all systems can be backtested to glean real results... but can just as easily be colored by hindsight...
for example... PA pick 3... I have fully backtested the following against the entire draw history (all the way back to 1977)... TicTacToe, Skips (longest outs), NBey6's system and a host of pencil and paper systems from the infamous 111 to the julian date system... no natural winners... with some tweaking winners could be found, but if you alter the systems they are biased to past draws and inevitably don't work anymore.
for the most part, the systems are based on either theories (not proven fact) or presentation of statistics... Some programs help if you wheel while some help to identify pairs etc... because systems are experimental, I prefer to design my own ,and because they are all theory-based experiments, I don't fool myself into believing they are anything more.... I would rather re-invent the wheel than pay someone for something no more or less successful than what I could do myself in excel/access/visual basic...
you need one ticket to be in it... unless you are a wheeling whiz, anymore than 1 is overkill...