tedster - 7:16 pm on May 31, 2010 (gmt 0) [edited by: tedster at 8:02 pm (utc) on May 31, 2010]
When I recommended ungrouping the ranking losses, it was based on several factors:
1. Google tweaks their ranking factors in minor ways almost every day, and in bigger ways less often, but still a few times a month.
2. A big shift occurred right over the beginning of the month. It was reported here almost immediately by pontifex and soon by others, here and around the web.
3. Matt Cutts verified that a major algo change rolled out at that time (Apr 28-May 3) and was complete.
4. Another smaller number of reports came in around May 17 - that's two weeks later. And it was definitely a smaller group, though still very real reports. Those websites made it through the beginning of the month without losing traffic!
It seems clear to me that there were two different changes at Google. The second might have been a dial turn of some kind, but it wasn't the main event, the long-tail changes we called Mayday. If we want to do rigorous analysis of either group, then I don't think blurring them into one event makes sense.
[edited by: tedster at 8:02 pm (utc) on May 31, 2010]