Unlike most games, CCP makes the number of users in Eve readily available to the public. Not only does the figure appear on the login screen, but also is accessible via an api. Luckily for me, Chribba created Eve-Offline which tracks not only the server status for all Eve servers but also the concurrent user status. Eve-Offline is the source for all the stats used in this post.
|Graph 1 - Daily Peak Concurrency|
First, I took a look at the daily peak concurrent user marks. With the red line in the graph marking when the slow burn began, we see that in the 6 weeks before the latest anti-botting campaign the PCU reached over 50,000 users 10 times. After the campaign began, it reached over 50,000 twice. The first time was the Sunday of Fanfest when word of the campaign probably hadn't spread far (and was probably laughed at by those who heard) and on 10 April. Now, 10 April is an interesting day because if CCP really was banning a lot of accounts for botting, that day marks the end of the 14-day bans those caught botting would have received in the first round of bans. So the possibility exists that on the 10th a lot of botters logged back in and were hit with a 30-day ban.
|Graph 2 - PCU Sundays|
Focusing on the Sunday PCU numbers really shows that something drastic changed. In the six weeks before Fanfest, PCU exceeded 54,500 every Sunday, averaging 55,962 users. After Fanfest, PCU has not reached that mark and has failed to reach 50,000 four out of the six weeks, averaging 49,562 users. That is a reduction of 6,400 users, or 11.4%, from the 6 weeks before CCP Sreeg began the slow burn vs the 6 weeks after.
Does that mean that 11% of the people playing Eve were botters and that CCP Sreeg's efforts are responsible for all the reduction in pilot activity. Not at all.
|Graph 3 - 7 Day Rolling Average|
As the above graph shows, pilot activity was falling even before the War on Bots™ began in earnest. The above graph is a 7-day rolling average of the daily peak concurrency numbers. Although it appears that the number of pilots logging in was stabilizing at the 47,000 - 48,000 mark (the pink line), that represented a 3.8% decline in pilot participation on Tranquility in the 6 weeks leading up to Fanfest. Still, since Fanfest, the 7-day rolling average of PCU has declined by 6.1%. And other factors, like the anomaly nerf and dissatisfaction with Incarna could have led to pilots leaving Eve.
However, the spike in the middle of April leads me to believe that attacking botters was a significant factor in the decline of the PCU numbers. If Eve did not suffer the drastically lower CPU figure for 6 April due to the deployment of Incursion 1.4, the 7 day rolling average for 11-13 April most likely would have matched or even exceeded those from 6 weeks previously. The lower activity does not mean that accounts were banned. What is most likely given what I've read on the botter forums is that those botters who were not banned are more cautious on how they bot. The botters are now not running their bots 23/7, which may lead to CCP having a harder time detecting the bots.
I wish I had figures available for the Australia/New Zealand time zone because I think that would show botting activity better as fewer players are on during that time, but Eve-Offline does not provide those numbers in a reliable (i.e. at the same time every night) fashion. But that's okay, because Chribba's site provided a lot of information proving that activity is definitely going down on TQ. While I believe that CCP's War on Bots™ is a major factor, others can disagree. Until CCP Sreeg comes out with a dev blog, the debate will rage on.
Related Post: CCP's War On Bots: Peak Concurrent Users - 1 May 2011 to 5 June 2011