When I see people being yelled at for their non-standard theories in internet it mostly boils down to them ignoring (let's call it) the structure of a scientifi

When I see people being yelled at for their non-standard theories in internet it mostly boils down to them ignoring (let's call it) the structure of a scientific knowledge, when they just pick from a textbooks statements they like, reject other statements they do not like, and then try to build something on top of a resulting mess. Scientific knowledge has a structure, similar to a structure of a math theory where axioms and definitions are used to prove theorems which in their turn are used to prove even more statements and so on. You can trace a path from any statement to axioms and definitions. But empirical sciences have structure more complex than that of math. They rely on empirical evidence which is limited by abilities of instruments, and they are subject to change when instruments become better. Sometimes they reject some hypotheses not due to contradicting evidence, but because there were several competing hypotheses and one of them had more supporting evidence. If you start asking questions why physics thinks that X is true, and ask a lot of them not accepting any statement without asking such question then you'll get a lot of answers (conditionally on a benevolent physicist willing to answer them all), and you'll get eventually the structure of physics. But then comes some freak and claims he proved Einstein wrong, while it is clear to physicists that he doesn't understand the structure of the theory. To get a simple example I can resort to a calculus and... Imagine a student that when asked to prove that sin(x)/x -> 1 when x->0, "proved" it with a help of L'Hôpital's rule by taking derivatives of nominator and denominator. L'Hôpital's rule doesn't forbid it explicitly, and it may seem to be ok, but the trouble is the derivative of sin is proved with reference to a limit of sin(x)/x when x->0. So it becomes a circular reasoning and if you see how student does it you know that the student doesn't understand a thing about calculus, his attempts to pretend to know calculus are pathetic. Similar mistakes one can make in any discipline, though in them the structure or reasoning behind the theory is more complex, in math you can rely on pure logic, you need no Occam's Razor or something like. In social sciences you need to know all the history of your chosen domain of knowledge, how it evolved and why so. In physics it is simpler but still you need to understand at least some decisions made in the past (or maybe all of them, I'm not a physicist, so I'm just speculating). One can propose any insane theory contradicting to Einstein, Newton or anyone else and not to be yelled at, but you need to state your starting position carefully with regard to a structure of a scientific knowledge. Like if you want to derive statement sin(x)/x->1 when x->0 by resorting to L'Hôpital's rule then you need first to build your own calculus with blackjack and hookers from scratch in a way that make it to be a non-circular reasoning. And probably you will need to explain your motivation why are you doing that, because people will be asking about it a lot.