Oricon recently released some charts with their mid-year ranking, and I’m trying to figure them out here. We’ll start with the typical mid-year ranking.

2015 Mid-Year Ranking by Volume
http://katsu.ula.cc/test/read.cgi/piropiro320jp/1432224768/260
This covers the week of 2014/12/08 (12/22付) through the week of 2015/06/01 (06/15付).

The following items represent new data beyond what I have as of right now, and their discs have been updated. Typically these represent very small gains, because the wekly extended ranking adjustments already had them estimated quite closely. Some items were left off because they’ve already ranked

2015 Mid-year DVD Top 100
17,102 Youkai Watch v1 [+38,292 (2014) = 55,394]
10,089 Haikyuu!! v8
*9,705 Haikyuu!! v9
*9,343 Haikyuu!! v6
*8,976 Haikyuu!! v7
*6,677 Free! -Eternal Summer- v6

2015 Mid-year BD Top 50
70,194 Love Live! 2nd Season v7
37,829 Fate/stay night Unlimited Blade Works
24,386 Persona 3 #2 Midsummer Knight’s Dream LE
19,948 Kantai Collection -Kancolle- v1
19,158 Strike Witches Operation Victory Arrow v1
16,649 Sword Art Online II v3
16,172 Sword Art Online II v4
16,149 Sword Art Online II v5
15,630 Shirobako v1
15,573 Sword Art Online II v6
15,291 Sword Art Online II v7
15,203 Kantai Collection -Kancolle- v2
14,945 Shirobako v2
14,601 Shirobako v3
14,391 The IDOLM@STER Cinderella Girls v1 (w/o G4U)
14,309 Shirobako v4

2015 Mid-Year Ranking by Series
http://katsu.ula.cc/test/read.cgi/piropiro320jp/1432224768/261
This does not say what weeks it covers, but I think it’s reasonable to assume it also covers the week of 2014/12/08 (12/22付) through the week of 2015/06/01 (06/15付).

This one is quite the pain to use, because it seems to count data from below the Top 300 cutoff, which would also count any small sales from RE versions. There is simply no way it could be correct otherwise, because the averages it implies directly contradict the extended rankings. For example, Yuyuyu is up to 55,743 total discs (all released in the reporting period) but this list says 58,961, a bump of roughly +536 per disc. Because the extended rankings place the discs within very tight ranges, less than a few dozen units for most, there’s nowhere to gain 3,218. Maybe 200-300 at most! And then you have to consider we don’t know which volumes have how much of that gain, or even which video formats (DVD vs BD) or editions (LE vs RE).

So I’ve settled on calculating a batch adjustment at the series level, and adding that to the total average. I will not try to spread these discs out across individual volumes, as that would end any hope of tracking them reliably via the extended rankings – and would cause total chaos in general. Even as a one-time adjustment, it can be very tricky if not downright impossible to calculate this 100% accurately for titles that span both 2014 and 2015 so I’m not claiming this is on the nose right down to the last disc. But it should be pretty close. Please let me know if you think you see any mistakes though.

Note that these numbers are taken before any additional data from the week of 2015/06/08 and its extended rankings. The before-after averages factor in this week’s top 100 rankings, but I needed to make sure the “previous known” values did not.

120,464 Sword Art Online II
• Previously known sales: 114,370 = 17,179 v8wk1-2 + 95,590 v3-7 + 1,246 v2wk3 on + 355 v1 past 2014 yearly
• Addition: 6,094
• Average: 19,479 → 20,156
• Note: I’m assuming SAO Extra Edition isn’t counted as it explicitly says SAO II, but even if it is, it released late April 2014. Only a negligible number of units could have sold in 2015.

98,151 Love Live! 2nd Season
• Previously known sales: n/a
• Addition: per below, 721 avg gain * 7 volumes = 5,047
• Average: 65,050 before adjustment, 362,248 + 98,151 = 460,399 / 7 = 65,771
• Note: Previous sales not needed as we have full 2014 and 2015 data and it’s a single-format show, which makes things very straightforward since I don’t see any reason to suspect this involves live event discs. It’s almost all going to be from the REs.

92,418 Haikyuu!!
• Previously known sales: 82,915 = 82,471 v6-9 + ~444 v5wk4-5
• Addition: 9,503
• Average: 23,822 → 24,877

92,234 Shirobako
• Previously known sales: 89,110 = 13,986 v6wk1-2 + 75,124 v1-5
• Addition: 3,124
• Average: 14,963 → 15,353

89,383 Free! -Eternal Summer-
• Previously known sales: 84,744 = ~835 v3wk4 on + 83,909 v4-7
• Addition: 4,639
• Average: 21,604 → 22,267

69,223 Tsukimonogatari
• Previously known sales: 67,805 = 67,805 v1-2
• Addition: 1,418
• Average: 33,903 → 34,612

65,060 Kuroko no Basuke
• Previously known sales: n/a
• Addition: n/a
• Average: n/a
• Note: This data includes the Dec 2014 fandisc. That plus 3rd Season v1-2 comes to 60,148, which is a gap of 4,912. It’s impossible to determine exactly how much of that applies to 3rd season, but considering the only additions to v2 could be REs (if there even are any) and v1 could at most be counting REs plus one week of sub-300 ranking in both formats, it stands to reason that virtually all of the gap is for the fandisc (and maybe a handful of season 1 discs? season is not explicitly stated here, unlike with the other series), and thus no adjustments should be made to 3rd Season. I’ve therefore assigned it to the fandisc – which is a rather hefty bump from 18,845 to 23,757.

61,035 Mahouka Koukou no Rettousei
• Previously known sales: 54,510 = 324 v5wk3 on + 54,186 v6-10
• Addition: 6,525
• Average: 11,578 → 12,231

59,922 Kantai Collection -Kancolle-
• Previously known sales: 58,722 = 16,721 v3wk1-2 + 42,001 v1-2
• Addition: 1,200
• Average: 19,660 → 19,860

58,961 Yuuki Yuuna wa Yuusha de Aru
• Previously known sales: 55,743 = 55,743 v1-6
• Addition: 3,218
• Average: 9,291 → 9,827

50,325 Gundam G no Reconquista
• Previously known sales: 47,100 = 47,100 v1-6
• Addition: 3,225
• Average: 8,630 → 8,989
• Note: Previous sales as I have it were actually 51,781 which is due to using Soundscan for v1wk1 due to Oricon’s obvious screw-up, which we believe was leaving out the Amazon-exclusive edition. For the “previously known” number above, I went purely on Oricon in order to calculate the Oricon gap. We can also be sure that the gap isn’t due to Oricon adding back the missing sales, because the amount missing is much greater than the entire gap.

49,185 Yowamushi Pedal Grande Road
• Previously known sales: 47,051 = 47,051 v1-5
• Addition: 2,134
• Average: 9,410 → 9,677

44,730 Shigatsu wa Kimi no Uso
• Previously known sales: 43,454 = 8,056 v4wk1-2 + 35,398 v1-3
• Addition: 1,276
• Average: 10,903 → 11,044

43,331 Aldnoah Zero
• Previously known sales: 36,797 = ~782 s1v3wk3 on + 16,255 s1v4-5 + 19,760 s2v1-3
• Addition: 6,534
• Average: ???
• Note: Similar problem to Kurobas, except this time it’s due to season one discs. s1v3wk3 on + s1v4-5 + s2v1-3 comes to a bit under 37k, leaving quite a large gap. But how to portion out that gap? There’s no good way that isn’t a total guess. More guessing than I’m comfortable doing with the data, honestly. I’m not sure what to do about this one!

43,093 Jojo no Kimyou na Bouken Stardust Crusaders
• Previously known sales: 36,695 = 1,032 s1v5wk3 on + 11,440 s1v6 + 10,884 s2v1 + 9,358 + 3,981 Live Event
• Addition: 6,398
• Average: ???
• Note: The Aldnoah problem again. Oricon isn’t separating the original from Egypt-hen. At first I calculated a 10,379 gap which was just beyond the realm of the possile, but then I remembered the live event disc sold ~4k and probably ranked for well beyond the two weeks we saw, so it probably accounts for a nice chunk of the 6.4k gap. But this one really is impossible to use because it combines both a sequel split *and* a live event! Sorry Jojo, there’s just nothing here I can work with.

40,721 Amagi Brilliant Park
• Previously known sales: 38,391 = 38,391 v1-6
• Addition: 2,330
• Average: 6,399 → 6,731

40,045 Crossange
• Previously known sales: 37,396 = 37,396 v1-6
• Addition: 2,649
• Average: 6,233 → 6,564

39,697 Saenai Heroine no Sodatekata
• Previously known sales: 38,477 = 7,451 v4wk1 + 31,026 v1-3
• Addition: 1,220
• Average: 9,777 → 9,952

37,829 Fate/stay night Unlimited Blade Works
• Previously known sales: 37,440 = 37,440 v1
• Addition: 389
• Average: 37,440 → 37,829

35,334 Youkai Watch
• Previously known sales: 34,811 = 17,709 v2wk1-11 + 17,102 v1 in 2015
• Addition: 523
• Average: 36,662 → 37,194

35,093 Psycho-Pass 2
• Previously known sales: 32,628 = 32,628 v1-5
• Addition: 2,465
• Average: 6,526 → 7,019

32,762 Nanatsu no Taizai
• Previously known sales: 29,060 = 29,060 v1-5
• Addition: 3,702
• Average: 5,812 → 6,223

19 Responses to “2015 Mid-year per-Volume and per-Series Total charts from Oricon”

  1. something something says:

    Took a shot at this. Unfortunately I don’t think there’s much of anything I can do with the Aldnoah or Jojo data.

  2. PP says:

    Yuyuyu so close to 10k ;_;

  3. Ejc says:

    It says “集計期間:2014/12/22付~2015/06/15付” at the top of this anime series list (2015年上半期アニメ作品別TOP30). So it indeed covers from 2014/12/08 to 2015/06/07.

    > The Aldnoah problem again. Oricon isn’t separating the original from Egypt-hen. At first I calculated a 10,379 gap which was just beyond the realm of the possile, but then I remembered the live event disc sold ~4k and probably ranked for well beyond the two weeks we saw, so it probably accounts for a nice chunk of the 6.4k gap. But this one really is impossible to use because it combines both a sequel split *and* a live event! Sorry Jojo, there’s just nothing here I can work with.

    I don’t think the live event is included, since it this is an Anime ranking. So you’d expect only releases that are credited under “アニメーション” would be included. The live event is labelled as “イベント” by Oricon.

    • something something says:

      Regarding Jojo, the idea of the total of added discs being >10k is just… so far beyond what I can imagine. REs and a few extra weeks of sub-300 LE DVD/BDs for just a couple volumes? It seems mathematically impossible. I could *easily* buy shipments being >10k higher than Oricon reports, heck a good deal more, but this is strictly Oricon data so I just can’t picture it.

      Though either way, the split makes the data impossible to use. And I really want to avoid reversing course and counting all of Aldnoah and Jojo SC as single seasons purely for the purposes of this list.

  4. Ejc says:

    >Previous sales not needed as we have full 2014 and 2015 data and it’s a single-format show, which makes things very straightforward since I don’t see any reason to suspect this involves live event discs. It’s almost all going to be from the REs.

    I saw someone ask a question about whether the live event discs are included or not for Love Live 2nd season or not on Twitter, so I will provide the relevant numbers to quell any doubts other readers may have.

    We start off with Vol.2~6. Based on the lower bounds in the extended rankings and what they were at for the 2014 Yearly list…

    Vol.2: 58,944 – 56,286 = 2,658
    Vol.3: 53,264 – 50,131 = 3,133
    Vol.4: 54,689 – 50,286 = 4,403
    Vol.5: 50,493 – 45,778 = 4,715
    Vol.6: 50,867 – 44,150 = 6,717

    Vol.7: 70,194 – 0 = 70,194 (this is obvious)

    Total: 91,820 minimum for 1st half 2015 for Vol.1~7.

    For the live concert discs (I excluded the 1st concert disc since we don’t know what it was at by the 2014 yearly list)…

    3rd concert: 41,623 – 37,009 BD + 4,444 – 4,073 DVD = 4,985 for 1st half 2015.
    4th concert: 49,680 – 42,079 BD + 10,647 – 8,536 DVD = 9,712 for 1st half 2015.

    Total: 14,697 for 1st half 2015.

    Grand total: 106,517 minimum for 1st half 2015, which is > 98,151. Thus there is no doubt that the concert discs are not included in the 98,151 for LL 2nd Season for 1st half of 2015.

  5. moonnya says:

    Where is Nozaki kun? I remember it sells more than 9k each vol

    • Ejc says:

      It could’ve barely missed the cut-off of 32,762.

      Within the 1st half of 2015 (2014/12/08~2015/06/07), Nozaki-kun had 3 new volumes (Vol.4~6) released, with Vol.3 starting its 3rd week on Dec 8th. So we can deduce what its minimum sales was in the 1st half:

      Vol.3: 9,665 – 6,688 BD (2nd week) – 2,813 DVD (extended list 2,813~2,814 DVDs) = 164
      Vol.4: 9,378
      Vol.5: 9,107
      Vol.6: 9,198

      Total: 27,847 minimum.

      It is a tough ask for it to have added an additional 4,915 in residual sales to make that 32,762 cut-off, while it would’ve been ranking in the sub Top 300 BD or Top 300 DVD realm for most of the time.

  6. some1 says:

    Did The ranking by Series include DVD or BD only?

    • Ejc says:

      The list being labelled as “2015年上半期アニメ作品別TOP30” does not indicate a specific format or anything, except that it only includes “Animated” works only. So it is reasonable to assume it includes both BD and DVD.

    • something something says:

      Definitely both, the Haikyuu, Kurobas, and Free! numbers in particular guarantee that. They’re pretty well split DVD and BD that they wouldn’t be so high with just one or the other format.

  7. AholePony says:

    It’s too bad they didn’t serve up the numbers on a platter for you, instead giving mixed data with all kinds of odd contradictions. Thanks for taking one for the team ;-)

    Seeing the per-series results like this really drives home how much money monogatari prints. Tsukimonogatari was only 4 episodes and look at those sales.

  8. Ejc says:

    I just realised what the “アニメ作品別映像ソフト売上” lists that are being posted with the full list data are. Using this example here: http://katsu.ula.cc/test/read.cgi/piropiro320jp/1432224768/254

    This is for the May 25~31 week (it is labelled wrong there). Akatsuki no Yona Vol.1 was released in this week, and sold 1,820 BDs and 1,754 DVDs for a total of 3,574 discs. In that list there, it ranked 19th with 3,574. This means, this list is exactly the same kind of anime series list that was released by Oricon for the half-year, although it is a weekly list!

    19 *3,574 暁のヨナ

    I checked the Oricon Biz sitemap (http://biz.oricon.co.jp/sitemap.asp), and there is indeed such a list being released there, labelled as “アニメ作品別”, under “Blu-ray・DVDランキング” and “基本ランキング”. Too bad we only have 8 weeks of that data, since the usual guy only started posting it recently… but this may be incredibly useful, especially for future Vol.1s.

    An example I found is for Etotama Vol.1. It sold 494 BDs in its release week, and its anime series sales was 628: http://katsu.ula.cc/test/read.cgi/piropiro320jp/1432224768/238

    20 *,628 えとたま

    This means that Etotama Vol.1 DVD must’ve sold 134 in its release week, and thus is a new lower bound in the DVD extended list! We may get lucky one week in the future with a Vol.1 selling high on DVD, but low on BD and being able to figure out its BD sales using this method and thus being able to get a new lower bound in the BD extended list.

    • Ejc says:

      A few other examples too. For the Apr 20~26 week, Idolmaster Cinderella Girls Vol.1 was released. It’s series sales was 13,575:
      http://katsu.ula.cc/test/read.cgi/piropiro320jp/1432224768/241

      *7 *13,575 アイドルマスター シンデレラガールズ

      It sold 11,807 BDs and 1,723 DVDs for a total of 13,530 in the 1st week. However, it also released a RE DVD, so that means it must’ve sold 45 RE DVDs to make the full 13,575. The RE DVD doesn’t appear in the DVD extended list, so without a doubt… these weekly series sales (and the half-yearly series sales) include sales that are in the sub BD300/DVD300 realm.

      For the Apr 27~May 3 week, Sidonia no Kishi 2 Vol.1 was released. It’s series sales was 4,578:
      http://katsu.ula.cc/test/read.cgi/piropiro320jp/1432224768/243

      *3 *4,578 シドニアの騎士 第九惑星戦役

      It sold 4,466 in the 1st week, so that means that the DVD must’ve sold 112… and that one also doesn’t appear in the DVD extended list.

    • something something says:

      Yeah I realized what those were after seeing the mid-year list. I don’t intend to use them, though.

      Going through that amount of nonsense once for a mid-year list containing significant gains was one thing, doing it *every single week* for piddly additions here and there is just too much, since I already spend more time on this than I should. The extended rankings are bad enough about that anyway (ugh I wish my real job had something approaching normal hours, then I’d not care how much time I spent on sales stuff). Figuring out how to incorporate this into my spreadsheet structure (which is full of automated calculations) would be a mess too.

      So they won’t be reflected in my numbers here. Only titles that make a mid-year or yearly by-series list will get anything added. I will use that Etotama DVD lower bound on the extendeds, tho! That adds Etotama v1 and v4 DVDs and Mangaka-san v5 DVDs.

      • something something says:

        Oh fuck me. Even as I was typing the reply about how I’d *totally* not try to mess around with these lists, I was brainstorming ways to use them in my head. -_-

        I’ve added a new sheet and am at least capturing the data points back to when they started but I still have to figure out how exactly to incorporate the data without double-counting or screwing everything up. I’ll definitely have to do it *after* incorporating the extended rankings each week so I don’t double-count, though that can still be tricky. Well whatever, capturing the data at least is quick, I can determine how best to use it later.

        • something something says:

          I devised a way to get these numbers into my calculations, and it works well in normal cases… but as I think about it more, the more I’m concluding that this is actually a very bad idea to pursue.

          See, I can record the number that a series sold in a given week per these charts, and then in another cell I can subtract the “known” sales during that period and get the “addition” adjustment. The fatal problem is that it’s a one-time calculation. Once I determine how many discs are “known”, that’s that. But what if the “known” number changes later? Such as in a mid-year or yearly list. Or when new data in the extended rankings reveals previously unknown sales, or provides a tighter sales range than was known before. I have no remotely realistic way of going back and “adjusting the adjustments” to avoid double-counting these discs because I’d have to only adjust the portion that happened during that week and that level of granularity just does not exist.

          For example if I had been tracking Etotama with these rankings since they first went up, and then suddenly along came the new lower DVD threshold, I’d be double-counting the DVDs on v1 and v4 once I added the known DVD sales, unless I went back and recalculated the per-series adjustment downward to account for the new data I’d added. And this adjustment of the adjustment is just impossible in some cases, particularly for series that rank for many weeks.

          I need to bail on this one – it’s going to do more harm than good in the long run. I’ll do it for the mid-year and yearly lists, because that’s a single adjustment covering a long period of time and thus way less error prone.

          Damn I really miss when this sales stuff was no more complicated than “Record the top 50-100 each week. Add to the previous total. Done.” Monthly lists, yearly lists, extended lists, per-series lists, each makes tracking more complicated. And it’s multiplicative complication too, because they interact with each other in really awkward ways.

          It is a shame. These series lists give us >300 titles, but only for the top 20 series each week and is not volume-specific so it can provide unusable data when it incorporates special releases outside of the TV series volumes. The extended rankings give us data on a wider selection of shows, and at per-volume levels of specificity, but without any of the >300 data. And the two datasets cannot be reconciled because they’re counting different things. So I have to pick one or the other to track on a weekly basis, not both. And while this list would be a lot easier to deal with… well winner’s got to be the extended rankings right?

          At least I figured out the problems now before I got too deep into this I suppose. I will, at least for now, capture the data… I just can’t factor it into the averages.

          • primadog says:

            Maybe just leave this tracking method to the readers and manually adjust if anything big actually pops up.

            • something something says:

              Selectively counting data or not is always kinda iffy, but I think it’ll be good enough if I stick to including 1) data that lowers the extended ranking thresholds, which will be very rare because it can only be accurately calculated for titles with only one disc out and no REs and 2) mid-year/yearly lists which I only have to deal with twice a year for a limited number of titles.

          • something something says:

            Oh man I just realized the other huge problem with this. Calculating an adjustment, dividing it by the number of discs, and adding it to the average totally fails if certain volumes never ranked on any format.

            For example, Etotama has DVDs for v1 and v4 thanks to extended ranking thresholds. Its BD average is 726 (based on 4 of 4 ranking) and its DVD average is 134 (based on 2 of 4 ranking).

            If I add in DVD sales (and some BD but it’s mostly DVD being filled in) from the per-series ranking, I’m double-counting the hell out of the DVDs for v2-3, because a DVD average of 134 *already assumes* 134 discs for v2 and v3 DVDs each, since we treat unranked as not existing, not as zero.

            So while I’ve calculated 562 missing discs for Etotama based on the per-series charts so far, it’s actually only 294/#discs that would get added to the average because an extra 134*2 is assumed from the 134 DVD average.

            Yeah this breaks down totally for series where a given format ranks sometimes but not always. AREN’T SALES FUN?

Leave a Reply