So, I had the idea from a YouTube video (probably TedX or something), where a guy used Spam Assassin to filter out profiles of girls he liked and girls he didn’t like.
Well, what about applying that same concept to Minecraft servers? Is there some way to take some of the traffic that Minecraft users send, and analyze it? I know with mail servers, if the spam points get high enough, it’s pretty definite that the mail is unwanted.
‘/ban’, ‘/mute’, ‘/jail’, ‘/is-a-jerk’ would be like ‘sa-learn spam’; and ‘/pardon’, ‘/unmute’, ‘/is-not-a-jerk’ would be like ‘sa-unlearn’. For ham, you take all the data you have about your server staff and turn it into ham and/or have ‘/is-a-friend’.
Another thing you could work on is having every player configure their own spam assassin settings and have a server-wide spam assassin for admin/owner use.
Then, the server admins/moderators can have a percentage possibility of troll, hackers, and griefers. Many people like using social engineering to mess with players (getting access to their claims) and admins (“hey, I’m from planet Minecraft! Give me op”). Then, when it comes to looking at staff applications, they can see what the likelihood is that they even want to look at the staff application.
It could also help people determine who they should ally with (Minecraft clans), it would cut out some player drama.
I’m not really sure how Spam Assassin is with databases and pulling global statistics of spam vs. ham, but I think you would want to make a custom database for Minecraft Assassin (or call it Sponge Assassin) (play on words there).