OK, look;

I predict a mag 6 quake from now to next week. You give me credit for a quake 2 months ago because of the minus part of the +- date range.

If you don't, I'll say you broke your own rules.

You also are allowing more latitude to someone who says a 6.0 2 month from now because he gets to use the whole date range whereas I only got to use the + part.

better?

But if the predictor has to specify the date range, no argument is possible. A 6.0 in June 2001 is clear; July is a miss, May is a miss.

Evaluation HAS to be precise; ranges such as you propose make it impossibly difficult because you have a HUGE range of possible answers to the seismicity question.

Roger