The probability would be based on long term statistics, such as 30 years, as opposed to one week. In addition, Alan's method takes into account aftershocks to large events I believe.

Your right, stars would be subjective and I don't even want to go there. Anybody who is interested in evaluating the predictions can form their own opinion in their own way, so I think we should leave subjectivity out of it. I think the goal with the status table is to just present the facts, and thats it, nothing more, nothing less.