It's hard to tell, because the graphs that he posts aren't the right kind for looking at growth. For folks who don't need the primer, skip to the "===========" below.
First, a bit of background for those who don't understand "log charts". When we want to study growth, it's important to use "log charts", where the vertical axis is printed using a logarighmic scale. In a lot chart, the distance from 10 to 100 is the same as from 100 to 1000. Why? Let's look at an example.
Say I invested in something that increases steadily in value by some percentage each year -- let's say 10%. This is what we'd call "steady 10% growth". Using a log chart, this would look like a perfectly straight line slanted upwards from left to right. If we look at it using a normal (linear) chart, it would curve upwards, and it would look like all the most significant growth was recent. (This is true in dollar terms, but not percent terms -- and when we are talking about growth, the percent matters more to the future, than the dollar terms!)
Let's look at another example, where I invest $100, and each year the value increases by $100. The first year, I double my money, yay!. The second year, I only get a 50% increas. The third year, it drops to a 33% increase. The percentage drops each year. Eventually, I get to the point where the "growth" each year is insignificant. This is not the kind of growth we want to see in a business or an ecomony, it's dooumed to failure!
Using a linear chart (like Zee's graphs), the latter plot looks like a nice straight line growing from left to right. It paints a rosy picture, when the real outlook for the future (if things continue this way) is grim. Using a log chart, we see great growth the first year, but in successive years, the trend loses its steam and curves to the right, eventually nearly flatlining with no visible growth.
People who are interested in the truth about growth (rather than in appearances) use log charts. Admittedly, linear charts are also useful, but they're secondary when the important thing is growth.
OK, now, here's how SL metrics look when plotted using log charts:
================================
http://learjeff.net/SL/stats-2008-11.htm
================================
click on the tabs at the bottom to see the different charts. (Note: the Lindex Rate chart is linear, not logarithmic, because the rate isn't growing -- it's staying the same, which is what we want it to do.)
So, what do these charts tell us? I'm interested in your interpretations. Here are mine.
First, I think that (since the Lindex rate is stable) the Lindex supply is a pretty darn good indicator of the overall economy. And frankly, it looks pretty good, if you don't dwell too much on the downward curve near the end. Overall, it's remarkably steady. No doubt the downward trend at the end will continue a bit, for reasons Zee Linden mentions in his blog. But if the overall trend returns, perhaps abated a bit, things look good for SL's economy.
Things may not look quite so good for LL, though. The next graph shows that Premium accounts not only stopped growing completely, they've fallen off considerably. Did this tred start when stipends were discontinued? In any case, if income from premium accounts isn't a major income source for LL, though, this isn't such bad news. And as Zee mentions, LL is looking for other ways to encourage premium membership.
In the next graph, we see $L transactions between residents. No surprise, there's a big bump that disappears after the gambling ban. As I mentioned above, this doesn't correlate to a big change in the linden supply (or as we'll see later, in Lindex transactions. My interpretation is that with gambling, money changes hands a lot, often with the net result of the average player leaving the table with almost as many lindens as they had before they started, and the difference going to the house. (That's true by definition, actually.) My point being, gaming inflates user-to-user transaction figures meaninglessly. So, with gaming permitted, this graph is not very useful. Ban gaming, and it becomes useful again. Wow, I finally found a benefit to the gaming ban!
The next graph shows that, other than a spike in 2006, the Lindex exchange rate is pretty stable. This is a good thing.
Finally, we see the amount of money exchanged on the Lindex. Well, this one does take a dip after the gambling ban. Clearly, the ban had a big impact; the growth slows considerably after it. To some extent, this can also be caused by the phenomenon mentioned above: if players routinely cash into lindens before playing and back out after quitting, that would cause inflated Lindex transactions. My guess is that it's a factor but not nearly as big as it is for user-to-user transactions, and that most players don't chash out their winnings quickly. Only the steady winners would cash out steadily. To do otherwise would pad LL's pockets with exchange fees.
But clearly, Lindex exchange growth is not strong, by historical perspectives.
In the past, I've also charted user hours logarithmically. I'm hoping to do it again, but unfortunately the data is not available (permission error, which I hope is simply a mistake).
Bottom line: I don't think the picture is quite as rosy as Zee paints it in broad strokes, but I do believe that his more detailed statements are valid. I'm optmistic about SL and expect growth to pick back up if the overall economy permits (bit IF there ....)
What's your interpretation?
Very true. I think the whole of SL is still recovering from that huge burst of growth - kind of absorbing all the new ideas and changes that the surge of new users has generated.
