Generalizations are often based on experiences. And often times individual experience varries. In this case, our individual experiences with fellow metalheads. You say most metalheads are violent and prone to stupidity. I say that the opposite is true. Why? Because my experiences have been positive whereas yours obviously have not been so positive.
Your case could be an exception, as I know metalheads from Canada, the USA, Belarus, Turkey, etc, that have experienced very similar negative behaviour from the other metalheads they've seen at shows or have in general met, this is based on their own local observations.
A true and valid generalization is achieved when the sum of experiences of a number of people sheds a, well, general result. It is mostly a scientific process that throws some numbers, and based on those numbers a conclusion can be found.
To reach a valid generalization, it is needed to take into consideration a number of sources to reach a conclusion, otherwise it's just a half-baked approach to the idea in question. Again, this doesn't deny that there will be cases like yours where there are exceptions, but they are not representative of the general idea.
Still, I think you are picking single ideas and my main point is not getting across. The above is an example that only regards the
nature of a generalization and how it is given credibility, it could very well be the other way around (most metalheads actually not being ignorant and disrespectful), and then the whole idea would change from a generalization through observation to a case of misinformation.
A statement that says "Jews attend to their religious beliefs" is correct, as it is the general rule, even though there are those who do not. This is a generalization based on observation throughout the globe and coming from many sources. In this case, the individual experiences do not matter when isolated, as someone who has met Jews that do not attend to their religion might argue that it isn't true that in general Jews do so, just as it wouldn't be true if someone else said that all Jews around the world do. This kind of localized data does not serve a purpose for the wider generalization because it is a) taking the exceptional cause as the general rule, and b) excluding the exceptions completely.
In the end, I want to make clear that my point is to show that generalizations do serve a valid purpose, even though we as individuals will have a different perception about certain things based on our own experiences, they do not change as a whole. To quote what I said earlier: "Generalizations exist for a reason, and it's up to the individual to know when to use a wiser judgement to find an exceptional case." But I can't stress enough that the generalization does not change because of this, it is what it is.
Just to make my point even clearer: You need to go through a dark alley at night, you are alone, and you can spot two individuals that are approaching, they are dressed as the 'typical' thugs. What do you do? Do you take your chances by thinking the idea of "dark alley + alone at night + people dressed as thugs = bad" is just a stupid generalization?