Hi Pixelosis, thanks for your comments. It is a dead thread now as far as we are concerned. So my last comment on it as a closure. You are correct that the intent was to create a buzz but not for downloads of the app as you have suggested. Doubt anyone is naive enough to think posting in a few developer forum will move the needle in downloads even if the target audience for the app is developers. The goal of maintaining this discussion for a while was to keep it visible long enough for people to stumble on it. We have achieved our purpose of getting visibility for crowd-sourcing the analysis and/or getting additional data points from other developers for their own set of apps as they instrument them. People have privately provided some insightful comments/analysis and suggestions on the data. Individual developers are justifiably averse to discussing specifics in public especially if it might appear unfavorable to Apple who control the fate of their apps. As you might know well, public forum threads like this devolve into conspiracy theories or thread-crapping with tangential or argumentative opinions just for the sake of being contrary. As I have mentioned in an earlier post, we have received a number of tips (some of them mentioned above on observations not all of which we can follow up on) and a few developers are involved in collecting data of their own. This data gathering takes time to get enough data for statistical significance. That is better than we had hoped for. We will get to the bottom of this sooner or later and see where it leads to. Much work to be done with the collected data while this strange behavior continues to happen. Thanks again for your help in keeping the thread alive at the top of the list. It has been a pleasure interacting with you.
Yeah they wouldn't want to do that, it's bad for the health. I find this quite condescending being one of the other handful of posters in your thread. Good look in your quest to find the Higgs particle hiding with a wallpaper scraper.
Jez, my apologies for what my words "threads like this" may have implied to you and other participants. All of my interactions earlier with you in this thread should show clearly that I was nothing but respectful to your comments as well as others. "Topics like this" may be a better phrasing of what I intended to say. What I meant to say in that poorly worded statement was that trying to come up with speculative theories by discussing them in a public forum is futile because it devolves into such content (which you can see in many of the threads here that attempts to discuss speculative or unsubstantiated theories on the ratings/reviews and manipulations). This is why I tried to avoid speculations on who and why of this pattern or encouraging such, rather just focusing on the objective data analysis process to entice people to participate. It turned out to be more of a private participation for reasons you do understand. Fortunately, it has been somewhat easier than detecting the elusive Higgs particle with a wallpaper scraper (love the metaphor!), more like deciphering intentional graffiti left on a wall. Peace.
OK, well before you close it definitely, Probation period? Who knows. This already occured the 6th of August. Why do they remain private? Why not post publicly and simply retain some sensitive information? How in the hell pointing out that you've suffered -1s in some odd fashion as well be considered an attack against Apple's interests or whatever? Hold on. First of all, that graph shows peaks at ~4.375. But there's not such value in the log at the corresponding date. It also shows four peaks above the 4.6350 mark but there simply are not that many in the log. Is the graph a curve that arcs beyond the points (up or down), as if using beziers or something similar? Secondly, the downtrend was already occuring in the old log, even before the odd -1s started to appear. Your app's average rating starts at 4.660, sometimes a bit more, sometimes less. By the end of it, the average ratings hover at 4.630~4.629. Nothing new here. The lowest average rating found in the freshier log happens to be 4.60591 and is collected on the 5th of October, yet surrounded by many averages in the 4.7xx~4.8xx ballpark. You actually lost more decimal points throughout the entire old log's course than during the new one, and considering the length of the older one, it seems clear that it covers a shorter timeframe. In other words, your app's ranking sank more on a shorter and earlier period of time, before the -1s, than after, during a long period of time while exposed to a phenomenon you consider to be detrimental to your app's ranking. And that's hardly the only point that is completely at odds with your claims and conclusions. I'm not sure how to understand this sentence. I'm afraid there's nothing to see. There's nothing glaring to be seen on this graph. The shaded areas just look as random as they get to me. Why don't they post openly about it? They don't even have to share the name of their apps. Recurring low amounts of ratings are fairly common. Things stagnate at some point, I wouldn't see any unnatural pattern. This is more interesting, especially if those one liners tend to sound rather negative. Examples would be nice though. I don't see such facts from your logs. The greater moments of rating posting are rather random. There were some happening on Wednesday, OK, but, others on Friday, some others right during Sat-Sun, etc. I also checked the months' ends, there are no clear visible peaks in rating numbers.
Ok Walter sounds fair enough with the misunderstanding, I think you swallowed a dictionary though because raising 'tangential' issues and potential thread-contributor-abuse seem to conflict - so it was natural for me to listen which is the case. Indeed you have politely dismissed my theories - which I figured the thread was requesting so as to eliminate any by recreating some localised re-scrape test (for example). I get it all the time too though so I'm happy to see what might have been at best touché was merely a context issue after all. I think people should be free to discuss experiences in public, instead of letting them always go away. If people commented on my old glitch then perhaps a "probably an overly active auto-censor was incorrectly confirmed" would be fine from anyone in the public house. But I think like the 23% thing (for example) it is not auto corrected back...oh tangential, though I think before your reply the thread was feeling a little military, it is yours to set as you wish of course. Ha yes that metaphor would make a great cartoon! Hm about the also mentioned a few times 'fingerprints' (as I described them), the ones I was getting by superimposing the graph over itself at various power-of-two levels, seemed to be repeated too often...BUT then I realise that what else is this kind of data going to look like and I didn't bother posting the layers. I still think this is all too negligible to worry about but you say you're on to something. Good closure. Keep us posted with any result or not. Peace.
