Here’s how the stats actually break down for the first 10 calendar days of the spring semester — for this year, last year, and 2005. (Note that spring 2005 marked the beginning of the digest system, due to the overwhelming volume on the list at that point. Also, note that days 6 and 7 are Saturday and Sunday.)
Yellow cells give individual daily message counts. Blue cells present weeklong totals for all four 7-day spans in the dataset. Green cells indicate the total number of messages sent in the analyzed time period. Maximum values for any particular time period are emphasized in dark red type.
There’s lots of room for quibbling over sampling frames. That’s why I’m providing several weekly totals over the 10-day period in which we have data for this year: to demonstrate that I’m not trying to influence the data in any way.
Some results are interesting. In particular, last year’s totals were noticeably higher at the beginning of the period analyzed, while this year’s totals were higher toward the end of the period. Since this is data for the beginning of the semester, that may indicate that the first few days of this semester may have been abnormally low usage days. Despite that possibility, this year had more messages in all 7-day spans except the first, and is only one message behind last year in the totals for the semester so far.
It is also important to notice the dramatically higher traffic levels across the board in comparison to my freshman spring 3 years ago. This is especially important to keep in mind, given that we switched to the digested format 3 years ago.