Using the NEIC search and PDE for 1973 thru 2000, I get a total of 643 events over a 28 year period. This translates to:

22.964 M6.8+ events per year, or
1 M6.8+ event every 15.894 days, or
roughly a 33% probability of this happening by chance

One thing that I always stress, though, is that these numbers are only averages. They do not reflect that these quakes do not occur exactly every 15.894 days. Maybe in my spare time (yeah right), I can come up with a chart showing the distribution of these quakes in intervals of how many days have elapsed between each occurrence. I've done this in the past with smaller magnitudes for So. Cal. and they looked interesting.

Dennis