Using SIMILE Timeplot for my personal, daily spam statistics. The data is refreshed every day. The gray graph in the background is the sum of all spam per day, less false positives. The gray slope is the total amount of the sum of all spam since I started tracking, which, at present, is around 80,000.
Interesting points: June 10, 2005, the only day on the graph I received no spam whatsoever. Thanks to the event on April 8, 2005, spam levels didn't reach their prior rate until more than a year later. But now we're volume-wise worse than at the start; just a week before this timeplot was published, August 1, 2007, saw the most spam delivered in a single day so far.
This is an average with a week's moving window. The platonic ideal involves the gray and dark blue graph matching up perfectly. The most recent SpamAssassin upgrade and the resetting of the threshold score seem to have moved much of the last year far closer towards that goal.