the system we have been using up to this point was much more complex, involving averages, standard deviations, log functions, and a complex web of interconnected variables... and it still had it's problems. The KISS simple solution would be to count every rating the same and take a straight average. Hopefully you can see that, in a fully anon environment like SR, this could be broken easily. Weighting as I have outlined is fairly easy to comprehend and makes it substantially harder for fake feedback to have much effect and gives weight in the right places imho. I urge everyone to have patience as we continue to flesh out this new system. At each step I get lots of angry people acting like the current state is permanent, or they are reacting to what they think will happen, not what is actually happening, or aren't reading the discussion. There is much more to come that will make all of this fit together nicely, including additional help for vendors to judge buyers. I'm not trying to stifle discussion, but I think we can take the emotional edge off of it and get back to discussing ideas and having fun with that. At the end of the day, if all of this really is a huge mistake, it will take a few hours at most to roll everything back to exactly the way it was. I do think we are going in the right direction though.