The current crop of theories most systems are based on...
1. Frequency based... how often a number has or has not appeared, encompasses hot/cold/due/warm/dead, also includes skip based systems. The frequency can be taken overall for each number, or by position in a multi number game, cash34ologist's Hot Sheet is the prime example of this... skips form the basis of most Tic-Tac-Toe methods as well
2. aggregate based gross filtering... includes hi/lo, odd/even, decades (positional), v-tracs... essentially taking the results and simplifying them into a smaller set of selections... narrowing down 0,1,2,3,4,5,6,7,8,9 to o(dd) and e(ven) n pick 3 is one example. you could also include sums here, as there is no direct path to translate a predicted gross filter pick into a single play... another way to think of these is like draw profiling...
3. dvent based... this would be any system that uses draw dvents (like a certain pair coming up) to guide future selections... the best example is the pair cross-off calendars in the pick 3/4 forums where you cross off each pair dvent as it happens within a certain time frame and concentrate on the remaining ones.
4. random based... guru's lottosynch and even playing your state's Quick Picks... drawing numbers from hats or bingo machines, throwing darts... etc...
5. draw history observation based... would include positional ranging and announcer systems, also cover work done in classification and attempted prediction by neural networks.
6. rundowns/workouts ... a mainstay of the pick3 arsenal, yet based purely on coincidence and maybe a bit of observation. the easiest example is the 111 system for pick 3... add one to each digit until you return to the original number (which is the last drawing).. no carrying, it's isolated by position... this system could not pick a double or a triple if the drawing before had 3 unique digits... lotto king/goose and his confusing system was essentially just a rundown workout based on thr remainders of divisions... you could think of these as transforms of the last winning numbers, which is where they draw their source of entropy.
7. statistics and probability based... JADELottery's powerball and megamillions posts are the best example of this. Forecasting, looking at medians, modes, averages and any other statistic is covered here also. statistics are static... that's half of the problem...
8. personally significant numbers... either in their original form or processed with some step before resulting in a play. they could be numbers representing birthdays/ages/etc... numbers given by a "psychic", dream numbers...
there may be others, but I believe that covers the bases for all current systems... keeping in mind that any system may incorporate more than just one of the above...
my main question is... what's next?