Under the heading of Prediction Talk, I’ll occasionally post thoughts on various phenomena we see in the DVD/BD rankings on Amazon Ranking Stalker and elsewhere.

Today what caught my eye, not for the first time, is the proliferation of Amazon-exclusive editions of releases. Amazon has done these before, but until recent seasons, nobody really bought them and they played very little role in sales predictions. One of the exceptions in 2013 was Date A Live, which accumulated 28.7% of its BD prediction on the Amazon-exclusive version. For the purposes of Oricon reporting, the split doesn’t matter. The Amazon edition LE and normal edition LEs are both reported together as the LE BD. But for the purposes of tracking sales performance on Amazon Ranking Stalker, the split can have a deceptively large impact.

While we don’t have an exact mapping between ranking place and Stalker points awarded, I have a set of 8,621 data points (I hope to get this around 20,000 eventually as more data points means better estimates) from a variety of releases mapping daily rankings to the number of points the release got that day. That gives a close estimate for each ranking of how many points are awarded.

Using that, I took some releases that currently have Amazon editions and normal editions and looked at where it would rank if you combined the two versions. It’s sometimes surprising how much higher the ranking would be. All rankings below are the current rankings as of Jan 21, 2014 at 10-12 AM JST.

Chuunibyou Demo Koi ga Shitai! Ren Vol. 1 BD
Amazon LE: 508 (22 pts)
Normal LE: 222 (42 pts)
Rank if combined: approx 125 (64 pts)

Sekai Seifuku Vol. 1 BD
Amazon LE: 245 (38 pts)
Normal LE: 418 (26 pts)
Rank if combined: approx 125 (64 pts)

Nobunaga the Fool Vol. 1 BD
Amazon LE: 911 (15 pts)
Normal LE: 1834 (9 pts)
Rank if combined: approx 480 (24 pts)

As you go up in the rankings, diminishing returns kick in, because the point values from rank to rank grow so much faster. Even so, there’s some boost:

K-ON! BD Box
Amazon LE: 12 (294 pts)
Normal LE: 56 (104 pts)
Rank if combined: approx 8 (398 pts)

Then there’s Hoozuki no Reitetsu which not only has Amazon vs non-Amazon splits, but also two distinct editions of the LE outside of store exclusives. Note that Oricon will not report A ver. and B ver. together; they will be handled as two separate products, like an LE and RE would. Now in this case, the B ver. is so overwhelmingly less popular than the A ver. that it’s not much different from the normal LE/RE situation where nobody even bothers to track the REs because they never sell or rank. Still, if this were to happen to a release with two versions of close popularity, we’d have to take four releases into account.

There are of course other cases where there’s no Amazon edition, but you still want to check all versions. This is more common with box sets and movies. One good example is:

Gundam Build Fighters BD Box Vol. 1
“Standard” Edition: 60 (95 pts)
“Master” Edition: 46 (110 pts)
Rank if combined: approx 20 (205 pts)

To sum this post up, you should always check if there’s another version ranking decently besides the highest ranked edition. It’s the same theory behind making sure to check DVDs. If they sell around 600-800 that’s generally well in excess of 10% of your typical late night anime’s sales. Most of the time it’s reasonable to use LE BDs as a quick ballpark, as in the majority of cases that’s 90% of what matters – but not always!

Leave a Reply