A few months ago, I discovered a way that some people are manipulating Wikipedia by banning people that they don't like.
It's very simple.
Ban someone. Then whenever you don't like someone, you can say the person is a "sock of a banned user". Even if the checkuser results don't pan out, there are ways that they can use this ploy. One of the most common, but not the only way, is to point out imaginary similarities.
Once you get a few banned people, you can expand it so that soon you can ban about 1/3 of anyone you like.
The checkusers will even participate in the conspiracy. Presented with a banned user who already has many "socks of the banned user" because the manipulative group has managed to label several people, the checkuser will naturally try to be accomodating and want to stop what they think is just adding a name to a long list of names of that "bad person". I have even seen a checkuser where one user is stale and the other is the newly proposed sock. The checkuser then pronounced them "confirmed" or guilty. Well, if it's state, then there is no IP match. The most they could honestly say is "inconclusive, go on behavior".
Even if the manipulative group present contradictory evidence it still works. For example, if one person is a sock of a banned user and other person is a sock of the same banned user but these two people cannot be the same person for a variety of reasons, ArbCom will turn the other way and ignore it.
It doesn't even matter if the newly banned person is a productive editor.
How can this problem be solved and the cycle broken? All it takes is one honest administrator but there are none. It's like in Cuba, none of the generals or politicians are willing to stand up against Castro, probably because they don't want to be executed.
Can you think of a way to solve this problem? Or is the answer just to give up on Wikipedia or to start vandalizing?
I have not mentioned any names so you can just consider the facts without being influenced by names.