I was glad to see your comment, this one and the previous. I basically agree with you. I can tell you that I have considered many of the inputs and outputs you listed, and perhaps others (do I have a formula with values assigned to each variable? of course not). Ultimately, I'm unconvinced there is a final model that can be applied to address all the issues. Therefore, in the interim while consideration is put into the details of such a possible (or impossible) model, action is necessary. My problem is with discrimination. If there are better alternatives to AA, I'd like to hear them. Unfortunately, nearly every one who argues against AA simply wants to eliminate it with nothing proposed as an alternative.