Further, it is definitely not about many times or multiple times, not even speaking about countless times. I don't deny the fact that you can win big time a few times (pardon the pun) but it is not like you (or whoever) are constantly winning insane amounts of cash all the time. So yes, it is possible to win 7 times (though it is highly suspicious with lotteries), but do you really think it proves that you can beat the odds all the time?
I think it proves that someone, can be beating the odds for a pretty long period of time, long enough to consider himself a pro in this business, and to be considered as such by others
Let's focus on what we know best, i.e. gambling
The point is, the longer you play, the more your losses will match the pattern defined by the house edge. You can win insane amounts a dozen times but only if the total count of your rolls is not much higher. If the number of your bets hits millions and keeps rising, there is less and less chance you are going to be that outlier you are talking about (unless you use martingale and earn dust)
The sad truth is that even if 10 billion people roll 10 billion times each, all of them will be losing because we are dealing here with one dynamic, i.e. the effect of the house edge, massively overriding another dynamic, i.e. a natural tendency of random events to have statistical outliers (what your point comes down to). In other words, you would need like 10 billion billion people rolling under 1 million times each to find that extremely lucky one