Fun With Numbers: Very Screwy Vampire Hunter D Numbers for the Week of 9/21-9/27

So this week’s Vampire Hunter D data is… different. DVD data didn’t update when the rest of the DVD data did. BD figures *did* update, but dropped off about 60 freaking percent. Last week saw the BD lifetime total at 4404 copies. This week the Nash DB has the same total down all the way to 1768! That’s not some small-time mistake, yo. I um, I saw this a few days ago and I don’t know enough four letter words to describe how this makes me feel. Look at this biz:


It’s a nightmare, really. I’m still flabbergasted. But after seeing these figures a few days ago, I think I have a crack theory about what might be going on.

This is kind of an outrageous happening – sixty-plus percent of sales don’t actually just disappear overnight without a newsworthy story. It’s certainly not a rounding error. There’s a couple of different ways this could have happened, many of them accidents which would depend on how exactly the data is entered into the Nash DB.

I have a particular theory as to why this might have happened exactly the way it did. It’s a bit of a reach, but it does kinda fit.

We start with an odd point reaching back to 2 weeks ago. While the original Vampire Hunter D film was rereleased on BD on August 25th, the sequel film, (Vampire Hunter D Bloodlust), was released on BD on September 8th. When the data for the release week of Bloodlust was added, it was not included in the database (I checked for it under several possible titles). It’s worth noting two things here;1) that the original film was released by Section23, while Bloodlust was rereleased by Eastern Star, and 2) that the Bloodlust DVD was released 2 weeks after the BD (in the week currently being examined).

One thing I’ve noticed is that Nash tends to add releases as a package deal when they do add them – for example, when BD data for the August VHD release was added, the DVD data was also added even though DVD data for that week had yet to come out. It’s possible that someone (or some formula) at Nash saw data for two separate releases with similar names coming in, took a look back, and plugged the back BD data for Bloodlust releases into the DB, replacing the old numbers for the original VHD release. DVD data not updating for the week at all could be that same person leaving the data pat while they sort things out further.

Our basic understanding of amazon ranks plus the observed rank data for Bloodlust on BD collaborates the possibility of this new figure being the 3-week total for the Bloodlust BD release. In those 3 weeks, the release averaged ranks of #490, #1370, and #1934. These would correspond to weekly BD totals of approximately 1600, 329, and 191 copies. That’s an estimate of 2120 copies in total, pretty close to the 1768 figure we see janking its way into the column this week. It’s not *that* far-fetched.

We’ll have more information when next week’s data comes out, but we’ll never have confirmation as to exactly what happened. All we know is that, at some point somewhere along the line, someone got lazy or effed up in a pretty visible way. This will definitely joins the Neko no Ongaeshi hiccup in showing some basic fallibility in Nash’s system. It’s always important to keep the obvious failures in mind while using their data to build a model.

4 thoughts on “Fun With Numbers: Very Screwy Vampire Hunter D Numbers for the Week of 9/21-9/27

  1. Either that or someone made a very drunken order for 2700 Vampire Hunter D copies. If your theory is true (which is likely imo), you just accidentally found some validation to your model. Like the late Bob Ross liked to say, no mistakes, only happy accidents.

    • Yeah, this whole project has been full of that sort of productively frustrating turn. Right now, it seems like I’ve basically got it down as much as it can be pinned down, and most of the remaining scatter I see is a function of the significant week-to-week flux of US home video market as a whole.

        • I don’t think they’re totally irrelevant, but they do seem to be a smaller factor than general flux of the system. What I think happens, based on what we’ve seen from the ranks of these releases where preorders can’t be applying, is that they get folded into the day 1/2 ranking of the title as the first copies ship out.

          Otherwise it’d be odd – the curve is shallow enough that 3-4 weeks of the preorder ranks new releases are getting should show up in sales ranks double what the curve would predict, and we haven’t seen much of that with either VHD or Bloodlust.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s