As you all know, a couple of weeks back I implemented a measure to discourage bots on the site.
I've been refining the code and for now it works as expected (most of the time). I've fixed many many bugs.
Today I made yet another change, one that will affect download counts.
Previously, whenever anybody clicked a story's title, the click counted if the clicker was logged in and hasn't accessed it yet in the last 24 hours.
I made a change that many will be happy with (free members), but may give the impression to authors that they've lost readers.
As of now, clicking on a story's title doesn't always increment its counter. If it's a serial, no increment happens until the reader loads a chapter completely.
So, if they click and view the index only, the click does not count. If they click on a chapter and fail to click the 'click for the rest' button, then no click count. Only on loading the rest of the chapter that the counter is incremented. As in it's a real read. Not just somebody checking if the serial has been updated (not an actual read). So the way it works now is actual text accesses are counted, access to the index only for serials does not count anymore.
I deployed the change a couple of hours ago and as far as I can tell, the stats show a marked decrease in story download counts. Almost 50% down.
Before now, if somebody was using a bot, their downloads counted even if they failed to download the complete text, so I couldn't tell if the 'click to load' measure was effective or not. Now I can tell.