An observation:

After taking the Nal. Drink first drink slowly and wait awhile (30-45 min) before another drink. Note how you feel. Just being aware of this helps to control amount ingested.

What occurs is a dramatically decreased interest in drinking any more, as it is not giving a pleasurable feeling. If it does, it is perverted or skewed. That also makes drinking or wanting to drink more an unpleasant prospect. If you drink fast, or don't wait, you don't give this a chance to occur.

Over time, this persists, and becomes a little more evident with each drinking episode. Because drinking doesn't bring the pleasure it used to, the desire to drink more, and eventually even in the first place, dies away.

Let's just say, it is the best thing that could be hoped for, and it works.

Regarding the null hypothesis...I always reminded myself of the rats for whom Nal worked 100% of the time in an average of 5 doses. It's a very repeatable experiment. Of course, with humans, it's much easier to override the Naltrexone using our reason, emotions, etc...