Something I stumbled onto a while a go that made me curious was a seemingly non-trivial connection between myanimelist popularity in the sales boosts of both novels and manga attached to a given anime, which led me to some speculation as to how different Western and Japanese fanbases’ preferences really are.
This time, I’m taking a look at a similar question; how many shows with high levels of Western popularity truly bomb in Japan? To answer this, I took the TV shows in the top 200 most popular on myanimelist, and excluded the ones attached to any series that averaged over 4000 in disk sales, or had a novel or manga chart in its first two release weeks at 20,000 copies or more. What remains is, theoretically, a list of the series which failed to catch on in Japan despite catching on in the West.* Data via myanimelist, someanithing, and the Japanese BD/DVD sales wiki. Note that I count the box releases as part of the disk average for pre-millenial series.
List format: Show Title (Popularity Placing, Disk Sales, Sample Print Source Sales)
Elfen Lied (6, 1794, below thresholds)
Another (34, 2507, irregular)**
Samurai Champloo (39, 2218, original)
Baccano (51, 3389, 12532)
Trigun (65 , 2514, before charts)
Welcome to the NHK (69, 720, before charts)
School Days (75, 3531, game)
Ergo Proxy (86, 1192, original)
Lovely Complex (99, 1038, before charts)
Pokemon (110, N/A, game)
Wolf’s Rain (137, 2406, original)
Chaos Head (139, 1185, game)
Kaze no Stigma (145, 1461, 18729)
Devil May Cry (147, 1723, game)
Monster (150, below thresholds, before charts)
Danganronpa (161, 3944, game)
Hell Girl (163, 2858, below thresholds)
Chrono Crusade (164, 2936, before charts)
Ichiban Ushiro no Daimao (168, 2587, 12086)
Shinsekai Yori (177, 587, below thresholds)
Log Horizon (180, 3922, 16459)
B Gata H Kei (181, 1832, below thresholds)
Shaman King (184, N/A, before charts)
Black Cat (186, 2475, before charts)
Digimon Adventure (190, below thresholds, game)
DN Angel (193, 1751, before charts)
Granted, the above criteria leave on some series tied to games and/or from before I have manga/LN chart data. If we eliminate those from the list (plus Another, for reasons described in the footnote), it gets a bit smaller:
Elfen Lied (6, 1794, below thresholds)
Samurai Champloo (39, 2218, original)
Baccano (51, 3389, 12532)
Ergo Proxy (86, 1192, original)
Wolf’s Rain (137, 2406, original)
Kaze no Stigma (145, 1461, 18729)
Hell Girl (163, 2858, below thresholds)
Ichiban Ushiro no Daimao (168, 2587, 12086)
Shinsekai Yori (177, 587, below thresholds)
B Gata H Kei (181, 1832, below thresholds)
The list itself is fairly diverse, containing fanservice-laden comedies, critical darlings, and Adult Swim headliners. It’s a small sample, but that’s a necessary consequence of looking at the upper tiers of popularity; comparatively few series reach that level to begin with. Depending on how you slice it, somewhere between 10 and 26 of the 178 TV shows in the mal popularity top 200 were arguable examples of shows that got popular outside of Japan without succeeding domestically. That’s between 6% and 15% of all shows in that category. It’s not a negligible number, but it is a fairly small percentage. Even with the smaller sample, that’s a strong enough difference that I feel fairly secure in echoing Hiroyuki Yamaga’s sentiment that the large majority of TV anime that become popular in Western markets tend to be the types that succeed in Japan, through either disk sales or print sales.
*Including the non-US traffic myanimelist does get. One of their moderators cited figures during the draconian romanize-everything name purges to the effect of 82% of that site’s traffic coming from non-US sources (see post #19).
**Another is a weird case. As an actual novel, its reprinting sold the majority of copies after the anime was aired, for a total on the order of about 200,000 copies.
Animetics 
Welcome to the NHK (69, 720, before charts)
I had absolutely no idea that Welcome to the NHK was not so popular in Japan. Why?
Pokemon (110, N/A, game)
It gets pretty TV ratings though, consistently in the Top 10 ratings for anime in Japan. Plus, the games are extremely popular worldwide s
Also, I don’t know where else to put this, so I’ll put it here: Can I request an article? About a correlation of anime sales and the sales and peak Oricon ranking of that show’s OP/ED themes. Think of this way: Madoka is a huge phenomenon, and this extends to the sales of its OP and ED singles, which sold extremely well (Connect and Magia were huge hits). On the other hand, Gen’ei wo Kakeru Taiyou (Day Break Illusion) flopped, and so did its ED single Mirage, which only charted for one week. However, there are some exceptions: the OP of the aforementioned Gen’ei, traumerei by LiSA, sold relatively well, although this may have more to do with the artist’s popularity. Another case is NouCome, which bombed pretty hard, but its OP sold pretty well. Now, my request for an article is: is there a correlation between the sales performance of a late-night anime and the Oricon chart ranking and sales of that anime’s theme songs?
I have no idea why any one thing is popular versus not in Japan or in the US. Any real answer would probably involve a bunch of elements. Likely it had one or more traits that turned a percentage of people buying DVDs in the early 2000s off of it, and the 12-disks for 24 episodes release schedule may have hurt it a bit.
I kicked Pokemon out of the more selective sample for that reason – like Another, it has a solid alternate revenue stream unique enough to miss the criteria.
That article would be doable and nontrivial, and probably wouldn’t take too much time if I limited it to 2012 series only. Do you happen to know where I might find archived versions of the full Oricon music charts? MAL archives the anime CD data in their BD/DVD threads, but I’m going to need the full charts to normalize for artist popularity (as you mentioned, that’s almost certainly a significant factor).
Try Oricon’s full year charts. I think there should be some data hidden there. Preferably though, 2012, 2013 and 2014 series should be included too.
I’m asking if you know of a source that archives those charts. My Japanese is good enough to decipher things, but it’s slow, slow work that I’d rather not do if anyone has already done the job.
I’m very sorry for the late reply, for some reason, WordPress would not post my comment if (as a joke) I use a Wikipedia article as my “website”.
Anyway, I’m pretty sure MAL has the best data that is easily accessible (Oricon’s more detailed data is in a premium section), and even if you do use Oricon, it’s mostly going to be skewed towards mainstream artists. You could base it on MAL data and maybe some Oricon pages as a source, with a disclaimer saying that an artist’s popularity can also affect the chart positions of a single, not just the show’s popularity.
Just a few things I noted about the list: First, Baccano can probably be considered a moderate success, since it did sell more than 3k, and few series that sell more than 3k are considered flops, unless they’re by an extremely popular studio (usually KyoAni) and/or were hyped a lot before release. Also, to my knowledge, Hell Girl is an original anime, although it did have a manga (and I think a novel) adaptation.
Finally, how were the sales of the Sasami-san LN, before and after the anime?
Baccano’s definitely debatable in that it did put up numbers that could have made it a success, but it’s not an unambiguous hit. If I have before sales numbers for the LNs, I’d probably be able to be more definitive about its exact effects – the series continues for years after, and the longer continuation equates to more worth for an anime bump.
Sasami-san never charted in 2013, which probably means under-10k 1 week totals. No telling what the bump was like – could have been zero or 5k.
One thing to note is that some of these shows aired on the pay satellite channels AT-X and WOWOW. These channels are more or less like HBO where you have to pay extra for them. That means that unlike most anime these shows didn’t need to drive disc sales/manga sales/ratings. All they really needed to do was drive new subscriptions or keep current subscribers. It also means that it is hard to tell how successful they actually were in Japan.
Out of the list the ones that aired on AT-X during their initial run were:
B Gata H Kei (Uncut only on AT-X)
Elfen Lied (Only aired on AT-X)
Ichiban Ushiro no Daimao (Uncut only on AT-X)
School Days (Uncut only on AT-X)
And the ones that aired on WOWOW during their initial run were:
Baccano (Only aired on WOWOW)
Devil May Cry (Only aired on WOWOW)
Ergo Proxy (Only aired on WOWOW)
Anime that aired exclusively on AT-X or WOWOW were at least partially funded by them (they are listed on the production committees for those shows). AT-X aren’t listed on the production committee for the other three shows they aired uncut, and those shows were aired (with heavy censoring) on terrestrial TV networks. I’m guessing though that AT-X has to pay some amount to air anime uncut since it has to help drive subscribers.
Oh wait I just checked again and AT-X are listed on the production committees for B Gata H Kei, and Ichiban Ushiro no Daimao. They are also on the production committees for other big fan service shows like Ikki Tousen and Queen’s Blade. So they do seem to lean towards helping fund shows with extra fan service in them that can be actively censored for any terrestrial broadcasts.
Out of curiousity, where did you get that production data, and does it list what order they were on in the production committee? I’m just wondering whether or not the channels got top billing. It’s fairly frequent for a number of companies to be on production committees, but my understanding is that it’s the first 2-3 listed that pay the most money and thus have the revenue streams most relevant to a show’s success/failure demarcation.
In general, a 3-4k average isn’t enough to pay for a typical anime budget after you chop out the retailer’s share 50%-ish of each volume. One reason those numbers are often quoted as a break-even is because we assume that any given show is going to have several alternate means of producing revenue. Even if the shows had alternate revenue streams available, those shows put up poor disk sales, which is a first-order indicator of poor popularity. While it’s true that you can’t necessarily rule out that they were more popular on the satellite channel, it’s more likely they weren’t based on what we can see.
I was just going off ANN. They only seem to list the committees in alphabetical order though.
For Baccano I was actually mistaken and ANN only list WOWOW as a broadcaster. I double checked the dvd ending credits and they don’t list them on the production committee (製作) either. I also messed up Elfen Lied. Same thing for that one, AT-X are only listed as the broadcaster and aren’t on the committee.
For Ichiban Ushiro no Daimao, where AT-X were on the committee, the ordering is:
コンスタン魔術学院 (A holding company or something I guess)
Marvelous Entertainment
Media Factory
Hobby Japan
AT-X
Good Smile Company
Lantis
GENCO
So for this one you have both a dvd publisher and a magazine publisher listed above AT-X on the production committee.
On the Ergo Proxy blu-ray, the production company ordering is:
Geneon Entertainment Inc.
Geneon Entertainment [USA] Inc.
WOWOW
Manglobe
In this case the Geneons, who are both in the business of selling dvds, are listed before WOWOW. Plus, Geneon USA being on the committee would seem to indicate that the show was made for a western audience.
The order for Devil May Cry is:
Showgate
Media Factory
WOWOW
Again the dvd publishers (Showgate/Media Factory) are listed above WOWOW.
But another show WOWOW had a hand in, Crest of the Stars, the committee order in the dvd ending credits is:
Sunrise
WOWOW
Bandai Visual
This committee ordering doesn’t change for the sequels Banner of the Stars and Banner of the Stars II. Then for the 2-part OVA, Banner of the Stars III, WOWOW dropped out of the committee and the ordering in the credits becomes:
Bandai Visual
Sunrise
So for Crest/Banner I’m guessing the satellite revenues did matter more since WOWOW is listed above Bandai Visual, the dvd publisher. Then again Sunrise is listed above WOWOW, and since they are an anime studio I am not sure what their measurement of success would be. I guess maybe it would be dvd sales. Either way Crest had to do decent by at least some measurement since it got three full seasons before going the OVA route.
Overall, I think you are right. The satellite channels never seem to take top billing when they are on the production committees, and the dvd sales probably do still matter quite a bit. I’d actually be interested to know what kind of dvd numbers Crest/Banner did since it is the only series I could think of that got sequels and had a satellite channel listed above the dvd publisher.
dvdbd.wiki.fc2.com/wiki/星界の紋章
“Sekai no _” data is available on the JP BD/DVD sales wiki. Seems like Crest series averaged 3.5k, Banner I averaged 9.5k, Banner II averaged 4.5k, and Banner III averaged 6.5k. That’s high-variance, but it does seem like the profile of an all-around success. Not saying that these ppv channels had no role in making things happen in anime – terresterial channels, at least, partially funded a portion of the shows that eventually started to fill up the nighttime slots back in the late 90s, and that marginal stuff is important over longer timescales.
Side note: does that mean you can look up production committee orders from show credits? That’s useful to know.
Another suggestion I have for an article, although hopefully easier to do: does airing exclusively on a pay channel or airing an uncensored version on a pay channel end up affecting sales later on? I’ve wondered about this ever since the performance of Coppelion. Coppelion was a massive flop in sales (<300 average), but it is interesting to note that it only aired on a pay channel (Bandai Channel IIRC). On the other hand, Initial D 5th Stage also only aired on a pay channel, but it still sold well. So the question: is there a correlation to sales and if an anime only airs on a pay channel?
The pay channel question is one that I could look into (iirc, that data is easily available, but I’d have to check into is some more). As for the general “Do shows requiring more censorship sell better?”, I’ve looked at that before:
https://animetics.net/2014/02/08/fun-with-numbers-do-elements-of-ecchi-contentfanservice-provide-a-boost-for-blockbuster-shows/
In the BD era, shows that unambiguously contain ecchi content have a slightly lower sales average than those that contain relatively little, while shows that contain debatable amounts of it tend to dominate HV sales.
To my knowledge, such a list is pretty easy to do (you just have to check which anime aired only on a pay channel, and seeing what channel(s) an anime airs in is pretty easy).