I think a big part of the problem is educating on how the system works. Let me try and simplify.
If you trust someone, you can leave a rating reflecting that.
If you don't trust someone, you can decline to leave them a rating.
If someone has engaged in untrustworthy behavior, you can leave them a rating reflecting that.
*Here's where it gets tricky.
If you don't agree with someone else's ratings, you can exclude them from your trust network.
If you agree with someone's trust ratings, you can include them in your trust network.
What you shouldn't do is...
Leave someone negative ratings because you don't like their trust ratings.
Leave someone a positive rating to counter someone else's rating.
Here are my thoughts:
A negative rating means that you don't trust the person and by extension, their feedback. It doesn't make sense when you trust what the person says (but also don't really).
In reverse, someone can be trustworthy but you may choose to disregard what they say, resulting in an exclusion.
For me, leaving negatives to another individual requires any of the following (though not only restricted to):
a) Extremely shady behavior
b) Outright scams
c) Behavior that induces harm to other persons, the forum or objects of possession
d) Any action that casts a reasonable amount of doubt upon its moral standards or legitimacy (i.e. account sales)
In the case of Vod, I countered their rating because I disagreed with it. However, that doesn't mean that I don't trust Vod. It also doesn't mean that I don't trust their ratings. Rather, it was a single piece of feedback that I wanted object against.
The act of exclusion will affect all the user's feedback, not just the one that someone disagrees with. I don't think this is the way to go.
The act of giving the user a negative will not change the feedback and thus that also is not the way to change ratings that one disagrees with.
Counter-ratings make sense in this regard. Don't you think so?