It is pretty funny (in a sad way) when your translation is scored pretty low by a word for word translation that absolutely makes no sense
I was wondering if the "real" scores are weighted based on this problem. For example, if a person submits a translation, and everyone who reviews it gives it a frown, should that person's scores for that particular translation hold much value? Probably not.
As for me, there have been a few times where I thought "dude, I have this translation perfect" and then I see one that makes mine read like a mechanical translation. In those cases I did vote the better translations higher...so I guess what I'm saying is, even though my translation wasn't the best, I still feel my high votes for the better ones were valuable.