One of the sources of confusion initially was that the decline appeared to commence a little over a week before Linden Lab announced its new policies on bots and campers. We asked Linden Lab to check our figures and did some cross-checking of our own. The skew, it appears, was caused by lower-than-usual login-rates due to grid instability during that period.
The remainder of the decline is easily attributable to the bot/camping policies. While the temptation is there to attribute other causes to the decline, the data that is available just doesn't support that.
How many bots? That depends exactly on how you define a bot. Some campers may be considered to be bots, while others may not.
Nevertheless, we think we're looking at in-excess of 15,000 bots, with some bot operators still to be tackled by the Lab. Previously, we made a fairly confident estimate at 10,000, and factoring in overall growth, that number now seems to have been about right.
There are a whole lot of interesting aspects to the interplay of people, bots, usage and engagement. Minimum concurrencies for example, have not really shifted, remaining around the 40,000 mark, while medians and maximums have slid. We've got a few hypotheses about why the minima should remain steady, but nothing really solid at this stage.
There may be some other interesting surprises in store in the statistics, yet. Stay tuned.
||Are you a part of the most widely-known collaborative virtual environment or keeping a close eye on it? Massively's Second Life coverage keeps you in the loop.