This is a follow on and correction of a PAD that I did on 27 December 2014. ( http://www.pbase.com/akmc_in_au/image/158680274 ) Yes, this is again a screenshot, not a photograph. But it's on the same subject. (It was also replaced with a new one on 5 May 2017 to include the end of year figures for 2016.)
In that PAD I took a look at how the number of images posted on PBase had been trending over the preceding five years. Each image is assigned a serial number, which is one higher than the previous image. Although you can never get the exact first image of the year, you can get close enough. Where I made a mistake in the PAD was in trying to identify the last image of the year as well, which resulted in gaps. On this occasion I'm not going to bother with that; I'll just identify the first image of the year (which, for convenience, I take to be my PAD of 1 January or, for 2017, a more regular PADer who posted that day) and simply assume that the last image of the previous year was the one before that. Adopting the methodology changed the results slightly, but unfortunately not the trend.
In 2011 PBase had about 15% fewer postings than in 2010. In 2012, 18% fewer, in 2013, 19% fewer, in 2014 I had originally estimated about 22% fewer but under the new methodology it's closer to 24% fewer, and in 2015 the percentage drop eased back a little to 22.6%. 2016 doesn't bear thinking about. (When I replaced the image on 5 May I also did a calculation projection to the end of the year based on the number of images posted to date. If it's linear through the year (which it won't be; it'll go up and down) we'll come in at about 1.83 million images, down a bit over 707,000 making the drop a smidge under 28% (27.86%). That will mean a drop of over a quarter two years running. We shall see.)
Unfortunately, this has now become a case not of whether this loss can be sustained, but for how long it can be sustained.