I think you are misunderstanding both scales. The OVR ratings are just a byproduct of the way that the attributes are attained. So trying to get to a set OVR is actually the opposite of what both systems at FBG have done. The equal interval sets the highest to 99 and lowest to 0 for any attribute, regardless of position. The true mean system adds a third point that coincides with the average for any attribute, which skews the regression toward the mean and away from the line drawn by the maximum and minimum points. The OVRs are a simple result of the values of the attributes.