You measure these things with surveys and interviews and design statics. AA claims to have success and relapse numbers, but I’d prefer independently run ones. Not everything scientific needs to be or can be a double blind trial.

If it’s measurably, it can be improved. Even if AA works, does it have a better success rate then quitting cold turkey? Even if works, are there things that can be changed to make it work even better?

You’re a very bizarre form of evil.

You sound like you’re in a cult.