Those Too Darn Hot stats - the last one I hope

I have just FINALLY figured out some of what's happening with the "What's Hot" list and why it sometimes shows me a list that seems radically different from time to time. One thing is very simple, and comparatively innocent, and the other ... possibly somewhat less so.

The different lists I've occasionally seen are a function of whether or not you happen to be logged in at the time. (And/or whether the browser cookies and cache have caught up to the switch from one state to the other.) If you're logged in, the store's "Hide what I own" function activates for EVERYTHING but a direct search for a specific word. If you're not logged in, it can't tell who you are, so you get a complete picture of the store.

So, for my logged in browser, the top five items today appear to be:
1. Karen 7 Pro bundle
2. Hongyu's V7 bikini
3. Shade 3 Pro
4. i13 New Visions Bedroom
5. Carrara 8.5 Pro

Karen continues to be able to fend off all comers, for the moment. The Bethany pro bundle, on this list, has, alas, tumbled all the way down to number 8.

However, on the browser that is NOT logged in, the top five is just a bit different.
1. Karen 7 Pro Bundle (you'd expect that to remain the same, since number 1 is number 1 and I haven't bought the bundle)
2. Hongyu's V7 bikini (see above comment)
3. Scott 6 Pro Bundle (... now that IS a surprise, on two counts; see below)
4. i13 New Visions Bedroom
5. Victoria 7 Pro Bundle (residual from the last few sales, with people deciding to pick her up since they could get her for less than her original intro price)

The difference between logged in/not-logged-in is huge, really. Slots 5-9 on the not-logged-in list are all filled by things that I've bought (...Young Teens 5 at number 9. Original flavor Genesis. Really. OK, then.) There is one anomaly in that list that's is puzzling. I don't own any version of Shade 3, and yet on the Not Logged In (NLI) list, it falls from number 3 to number 9. Statistically, it shouldn't have fallen out of the top five at all -- it shouldn't have fallen, period, especially given that the bedroom directly behind it didn't fall. 

I do think, to the extent this list can be relied on at all, it shows tiers and clumping. At the moment, it appears that the Karen Pro Bundle is in a class by itself -- possibly to be joined by the new Hongyu bikini. (As I commented elsewhere, there's always a new bikini.) Scott is also showing surprising endurance, for an end-of-generation figure that is also male, and with the imminent release of his successor generation at hand. (Yes, I know it's only been a week, but, again, end-of-generation, and male, and ... confused presentation, let's say. Endurance of any sort was not to be expected.) Below Karen, I think there's a very long tail of stuff that can switch positions with very small shifts in sales numbers.

That said ... the answer for the people who assert that DAZ manipulates the list seems to be both "Yes, but somewhat incidentally" and also a qualified, "Well ... possibly on purpose, yes." Incidental, because most of it is an artifact of the "Hide what I own" check, and that affects every part of the store. "Possibly on purpose, yes" because what happened to Shade 3 on my not-logged-in list doesn't seem to make sense, at first glance. It shouldn't have dropped behind the bedroom, which I also do not own. (I should note that it held its position relative to Carrara, which I also don't own, in the NLI list. That said, Carrara dropped behind iClone Pro, which I also do not own, in the NLI list, when it was ahead of iClone in the Logged-in list.)

The argument for "on purpose, yes": the What's Hot list could be a combination of true stats and weighting based on browsing and purchase history. True stats, in that most, although not all, of the stuff in the list that I don't own holds their position relative to each other. Weighting because ... well, put it this way. I own Hexagon, because I got it during the period when Daz was giving away all of their programs except Carrara for free. Never used it, never installed it, never registered it, never bought any addons for it -- if it even has addons -- but I own it. Based on that, it would be logical for the store to skim my "purchase" history in an automated way and say to itself, "Store, this guy has Hexagon. Perhaps he would be interested in other modeling software as well. Let us push Shade 3D at him in a relatively gentle way, and perhaps he will take a look and purchase this wildly expensive software package." Similarly, I have items that have separate Carrara components, as well as Studio/Poser components, as well as one item that I purchased long ago before I realized that Carrara even existed, and never got around to returning it. So it also makes sense for the store software to say to itself, "Store, he has Carrara stuff. And yet he has not the Carrara itself. Perhaps we should push that at him in a gentle way and see if he can be persuaded to buy this less wildly expensive software package." (Not working on either count, by the by.) This is, to be sure, a MUCH lower pressure sales tactic than shoving ads for stuff you don't have at you the way many stores do. It would make a type of sales sense to work this way, but it would not be good for a store's reputation for truthfulness if people were to discover this manipulation, mild though it may be.

The argument for "possibly not": How the list generation part of Magento may work, as configured in this store. As I say, I think after Karen and maybe the bikini, the differences in numbers of sales between each rank is probably very small indeed. I also think that, in order to reduce load on the store somewhat, the list is only partially dynamically generated. That is, the list is generated the first time you get to the store, a cookie is set telling the store when you got there, and that list is the one you see for some number of hours, until it gets generated again at whatever interval the store has set. It's entirely possible that, in the few minutes between me going to the logged in version, and then thinking to check what the list would look like if I weren't logged in, the relative ranks of the items could have changed, with just a small number of sales. So I could be seeing a true list both times, despite the changes that seem to make no sense.

Oh, and as a side note: on the NLI list, Bethany 7 Pro falls all the way to number 20.