It's not even in the same ballpark. You may have a point, if your "format" was off by one number only for a 4 out of 5 game, such as SC Palmetto Cash-5. Ie. predicting 12, 13, 21, 30, 32 with actual being 12, 14, 21, 30, 32. Even then, the odds of hitting the jackpot is much as 1 in 164, though could be smaller depending on the nature of the near miss. Would be worthwhile to play all those combinations. Again, that's assuming only being 1 number off.
Two or more numbers off (even if it's just a digit one way or the other), and the number of possible combinations skyrockets. At three off, it's even not close at all.
Using your examples of being off by 4 out of 5 or even all 5 (ie. "2?-2?-2?-5?-6?") for Powerball, one might as well blindly choose numbers on the play-slip or use quick-pick, since there's essentially no meaningful statistical difference at that point.
For Powerball and Mega Millions, being off by just one can be hundreds of times more likely. For example, Powerball Match-4 is approx 319 more likely than Matching 5. Matching 4 pays $100, Matching 5 pays $1 million. If one can be off by just one only, then it would pay to play every combination of the near-miss.
However, if the "format" is off by any more than that, the number of combinations skyrockets though one could in theory still profit. However, at 3 off (ie. 12, 19, 22, 42, 56 with actual being 12, 17, 22, 40, 57), playing every possible combination is both very costly and logistically near impossible (lotteries limit play; terminals are limited in how many requests can be served per second). None of these examples even take into account the Powerball number.
In short, the number space, even for a game like SC Palmetto Cash 5, is far bigger than many realize. The number space for Powerball is many times larger even for Match-5; several hundred times larger for jackpot. A combination may appear close, but still be way, way off.
It's great you're sharing number suggestions, but you started this thread titled "I am CONVINCED only way you WIN BIG if they let you win!". Your evidence so far doesn't seem convincing. Have you statistically analyzed number distributions from past draws? Run simulations? Back-tested your "format"? Would be helpful to do that before assuming the combinations coming out are suspect. Maybe they are, but the evidence you've put forward doesn't suggest that. Also, other players would likely have noticed, if something was seriously amiss.