OK, let's try this again.

If I count up all the mag 6+ quakes in the Pacific area between 1973 and 2006 I find there are 3153 of them. There are 12418.5 days in 34 years and 12418.5/3153 gives 3.9 days between quakes if we assume a uniform distribution. This is what I meant by "average", that you might expect to see more quakes in the 4 to 5 day columns than in the lower or higher values.

Now the actual count of days between quakes has 0 as the highest, dropping rapidly with the largest interval at 46 days, producing the reverse-J curve I was incorrectly calling exponential. It's actually a poisson curve with average near zero.

The question is "WHY"? Why are there more quakes on the same day, even when they are 1000 or more kilometers apart.

Roger