Rocket Bomber - article - manga - rankings analysis - One more year - one more gnarly mess of online sales data and another brave attempt to de-tangle it.

So, having finally caught up [in part] to the massive data backlog that is my daily blogging life, I hit a milestone:

Week ending 26 December, 2010.

I was six (almost seven) weeks late, but being late didn’t really matter. Given the tyranny of calendars and the inevitable retrospectives they engender (related to our collective love as a species for big round numbers) my reaching the end of a calendar year meant, of course, that I should dump everything in a bucket (statistically speaking) and immediately dish up a Top 10. We love numbered lists, especially Top 10 lists (even more than we love Big Round Numbers). A list is order imposed on the entropic universe, and the absolute certainty provided by a “Number One” makes buying decisions clear, and supports the illusion that our critical discernment (or personal taste) is, in fact, correct.

(Unless the Number One is obviously ‘wrong’, in which case we immediately tear down the authority or methodology that picked ‘wrong’.)

Anyway, I took all the data from my 2010 manga sales rankings, compiled it, and then looked at the top 10. And top 50. And top 3000 — and the numbers without context were basically meaningless to me.

At the end of the year, it didn’t matter so much that Ecchi Omnimanga Robot vol 8 outsold Mystical Shoujo Romance Kingdom vol 3 by 30% or 40% or whatever — both individual volumes sold well, and more important to me was how each series performed (all volumes inclusive) or whether the books were selling enough — enough to support the production of the series, enough to keep the imprint alive, enough to encourage editors and publishers to license more series, and to re-license titles so they remain in print. Also, a year-end tally says nothing about when a title sold 500 units, just that it did. What was the market like when the books sold? Was there an anime out in Japan, or in the US? What other books sold that week, that month?

This is why I like the weekly grind [yes, I like it — when I’m posting on time] because it’s a snapshot of a week, a much easier time period to wrap one’s brain around. And at least for my charts (I won’t mention again how useless the NYT manga bestsellers are) I also do series rankings, I stretch into a top 100 and 500 to show how other individual volumes in the same series are performing, and I attempt to provide depth of analysis — well, I haven’t had time but the data is there and I post *long* lists so anyone with the ability to save a webpage or copy-paste the text could do the same.

A single 2010 list, no matter how long, flattens all of that data out just to pick a Number One. But what do we know after such a list posts, that we didn’t already know or guess before? Does knowing the Top 10 of 2010 change the sales numbers? Popular books become more popular? Imprints that were closed or companies that went out of business suddenly re-open because, hey, our top book of last year was #7 on some sales ranking list? Hey, I didn’t know that; why the hell did we close if we had a Top 10 Title?!

A top 10 means nothing; “Number One” is nice, but does not reveal anything about the industry, or about the books.

It was Naruto vol 48 if you care that much.

This isn’t why I invest dozens of hours a month.

##

More important than Number One is how the book got there; what is the sales history? Even better, is the surprise at #17, and how a little known niche book manages to grow over a couple of releases into a solid midlist title — all the volumes in the top 25 even if (especially if) no single book ever does crack the top 10.

As was pointed out to me, though, the weekly charts — especially the top 10 — barely change from week to week.

Think of it as a satellite map: zoom in too close, and you can see your neighbor’s pool [“hey, I didn’t know they had a pool!”] even though you can’t see the nearest major intersection — the corner just outside your neighborhood — but zoom out too far and all you see is a splotch. Maybe it’s a splotch you recognise, because it’s a familiar shape seen on other older maps, but it doesn’t tell you how to get to the new burger joint just reviewed in the local alt-weekly.

Nothing will replace the weekly manga bestseller charts, and I hope to catch up soon and resume posting those in a timely manner. But a lump-it-all year-end chart is the same as a city-satellite-splotch. There is a role, and a need, for something in between.

After spending a couple of late nights with this problem, and with all that 2010 data, I came up with what I think is a solution. The new chart is a significant chunk, a satisfying chunk, and averages out some of the spikes and anomolies, but is not so big that all detail and grain in the rankings is lost. My new chart will slot into the calendar with which we’re all familiar, but isn’t a year-end [dismissive] summary just for the sake of posting something at year end.

It’s a new cycle, and a new commitment. I posted the first two of these charts yesterday.

2010 Q3 Summer Cumulative Manga Bestseller Chart
2010 Q4 Autumn Cumulative Manga Bestseller Chart

Each quarterly chart covers 15 weeks, and corresponds to the season it’s named after. So, for example, the Autumn chart will be the 15 weeks leading up to (roughly) December 22. You’ll note that 15 weeks is just a bit too long for a quarter though (it’s almost four months) so there is some overlap – which has the handy side effect that the Autumn chart will always include one extra week in December – taking us up to and through the week of Christmas.

Yes, I put some thought into this.

So, every 13 weeks I’ll post a quarterly chart as an extension of the work I’m already doing on the weekly charts. I apologize in advance that some of these bestseller charts will post late, but that’s because I have to work a 40-hr-a-week job to pay the bills and no one is paying me [*HINT, HINT*] to collect the data, compile the rankings, and post the reports. So you have to take what you get.

Looking at the two quarterly reports generated so far, I think I’m onto something with this. Among other things, consider the two imprint reports:

[from the 2010 Q4 Autumn report]

Top Imprints
Number of titles ranking in the Top 500:

Viz Shonen Jump 101
Tokyopop 56
Viz Shojo Beat 53
Yen Press 47
Viz Shonen Jump Advanced 33
Vizkids 31
Del Rey 30
HC/Tokyopop 18
Viz 15
Dark Horse 14

[more]

The Backlist
Number of titles ranking in the full 2500:

Tokyopop 372
Viz Shonen Jump 296
Viz Shojo Beat 222
Del Rey 202
Viz 188
Yen Press 185
Dark Horse 126
Viz Shonen Jump Advanced 124
Viz Shonen Sunday 99
Viz Signature 99

[more]

— I post a ‘publisher’s scorecard’ as part of each weekly chart: number of books from each imprint that place in the top 500; but look at the same sort of list pulled from the top 2500 titles from the quarterly chart: Suddenly, Tokyopop is number one. Viz Signature & Shonen Sunday jump into the top 10, while Vizkids & the HarperCollins/Tokyopop OEL manga slip away. Del Rey, on the strength of its backlist, jumps ahead of relative newcomer Yen Press — even in a quarter where Del Rey (recently hobbled by the introduction of Kodansha Comics) had virtually no new releases compared to the dozens of new titles, monthly, from Yen.

Let’s look at the series ranking:

Top 100 Series:

1. ↔0 (1) : Naruto – Viz Shonen Jump [13,184.0] ::
2. ↑1 (3) : Maximum Ride – Yen Press [11,388.7] ::
3. ↑1 (4) : Bleach – Viz Shonen Jump [9,250.4] ::
4. ↑3 (7) : Alice in the Country of Hearts – Tokyopop [9,066.0] ::
5. ↓-3 (2) : Vampire Knight – Viz Shojo Beat [8,943.7] ::
6. ↑6 (12) : Black Butler – Yen Press [8,069.0] ::
7. ↓-1 (6) : Black Bird – Viz Shojo Beat [7,653.8] ::
8. ↑28 (36) : Hetalia Axis Powers – Tokyopop [7,098.5] ::
9. ↓-1 (8) : Warriors – HC/Tokyopop [6,783.4] ::
10. ↓-5 (5) : Negima! – Del Rey [6,568.8] ::

[more]

Here we can see not only that Black Butler & Hetalia became Big Deals during the Autumn, but also the gains made since Summer: from #12 to #6 for Black Butler, from out of nowhere (gaining on 28 other series) for the new Hetalia Axis Powers.

##

For me, it’s not about the top 10: it’s about the midlist, and the backlist, and the context. This is why I spend so much time acquiring the data and manually entering it into my spreadsheets, and why I have always posted not just more than a top 10, but the sources and my methods. (not that I’m going to bag on the NYT for how useless their top 10 manga charts are. again. I also won’t mention that the New York Times bestsellers are a fairly amateur setup that can be — and has been — gamed.)

At some point, maybe 6 months from now or so, I will in fact post the 2010 chart (and I’ve also realized I’ve data from 2009 that was never compiled into a year-end chart) much as I’ve done for 2007 and 2008 — but that’s for the sake of the historical record, and not so much for daily blog posts or book retailers who’d like know what to order next month.

I like the new chart, and feel it’ll be a nice addition to the weekly charts. I also think I might post some interstitial charts (6.5 or 7 weeks into the quarter) just so we can identify the trends without having to wait a full three months. These quarterly rankings make more sense to me than a large globulous full-year chart.

Your comments are welcome; if you’ve looked at the two quarterly charts previously posted and had questions or wanted clarifications, please ask.