I would rather have it the other way around: take a system, computerize predictions and then computerize backtrack for 30 draws back. Run simulations on real lottery numbers and see what comes out. I did it for 2 systems and the results were very surprising.
For testing I used Pick 3 numbers for Wisconsin for November 2018 and earlier.
System # 1 - simple Vtrack.
Synopsis - take the latest draw number, convert each digit into Vtrack and play it. Vtrack at its simplest.
When converting to Vtrack triples always change into triples, doubles into triples or doubles and singles into doubles or singles. On average there is a lot of doubles.
Playing for triples is probably not the smartest strategy but doubles are a different ball game. For a double you need to match only 2 digits out of 10, 10 times easier than matching singles. For Vtracks this is reduced to only 2 of 5.
Statistically doubles constitute only 27% of Pick 3 winners. But many times, in short periods of time (10 - 14 draws), this percentage may go up to 50% and even 60%. You've got a match - your Vtracks give you doubles and doubles are being drawn. That's the #1 prerequisite for success with this method.
How do I know when doubles will starting bunching up? I don't. But I watch. A simple computation for detecting doubles trend: take the last 10-20 draws and compare with 10-20 previous ones. What's the current percentage for doubles and what's the current trend? For Wisconsin's last 10 draws they were at 60% and the trend +4.
Unfortunately, doubles do not consist of Vtrack digits only. So I have to monitor Vtrack digits performance as well. Statistically they are 50% of all numbers. But what's their current percentage and trend? At the end of November in Wisconsin they were at 70% for the last 10 draws. That's the second prerequisite.
In November both Vtracks and and doubles performed well. Should it then come as a shock that this very simple Vtrack method hit a straight - and it was a double? Obviously, when certain conditions are met the probability of hit is significantly increased. On the other side, using this method for playing singles appears an obvious waste of money.
Did I know these facts before the simulation? Not even a clue.
System #2 - Vtrack expanded.
Synopsis - take a Vtrack number (only digits 1 to 5) and expand it into regular digits. Then wheel the expanded digits into 8 combinations and play them. Then backtrack for 30 draws.
After dozens of simulations it appears this method is suited for playing pairs. Because of the nature of the expansion if you hit a pair you actually always hit 2 of them, no matter front, back or split. Compared with other targets (singles, doubles) pairs generally provided the best ROI, and most consistently for that matter (I measured ROI for all simulations).
Wisconsin example: Vtrack 123 (the simplest you can think of) produced the following combinations: 012, 017, 062, 067, 512, 517, 562, 567. These combos were played for the whole month. Nov 25 draw was 962. 2 back pairs hit: 062 & 562. Another hit occurred on Nov 11: 712 hit 012 & 512. Altogether there were 3 hits for pairs: 2 back and 1 split. For back pairs ROI was 83%.
Did I know this before the simulation? Again, not even a clue. These are totally unexpected findings.
That's why I would rather computerize the method first because in many cases the backtrack results may be an eye opener. Based on only these 2 examples I start suspecting that many systems (if not all) do have their specifics regarding conditions to be met and realistic outcomes they can produce. You wanna win you better know what your system truly requires to win and what you can win with it.
Honestly, I cannot imagine playing any lottery using pen-and-paper system. No offence intended but it's like trying to plot spacecraft trajectory - using abacus.
Aquarius