Inspired by MangAniMY, I was playing around with charts for awhile now. They didn't turn out like exactly i wanted, but whatever. Below are sales figures for some current magazine series.
Before starting i shoud note:
SE = Special Edition, Limited Edition.. etc.
Red colored numbers from oricon, rest are estimates from shoseki lists.
So, considering the sizable increase in sales of the manga following the anime's airing this past winter, as well as the fact that the BDs and DVDs for the anime itself sold extremely well, this shouldn't come as a surprise:
anyway, since you have been making all the Shoseki estimates, I actually have to ask... (not that I'll be doing shoseki chart since they really can't do LTD)
is there a universally accepted way to these shoseki estimates?
I just eyeball then most of the times, so I never bothered to actually calculate them. I figured it would best be done by fitting to a "log" scale, but I won't get into why that just a few problems to it automatically,
I know for a fact that the way bookrank does estimates is different . I didn't really look at how they're doing it, but i vaguely remember that they're using gaussian distribution which technically should be more accurate. I am not using it because it's hard to do that by hand. I am sure there are programs out there that i could use to calculate. I just never bothered with it because difference isn't big.
Aside from bookrank, all the other people that have been posting estimates in here are using same crude way. I know because i was comparing my numbers with theirs all the time. As for what that way is i have trouble explaining it with my english. I realized that i don't know the english names for even basic math terms. A basic search in internet didn't help me at all, so i am gonna explain it with a example.
If 100th place is 10,000 and 200th place is 5,000; estimates for 150th place is 7,500, 175th place is 6250,.. etc. How should i express this method in english? Using a linear scale to make estimation, is the best guess i have for now.
I know for a fact that the way bookrank does estimates is different . I didn't really look at how they're doing it, but i vaguely remember that they're using gaussian distribution which technically should be more accurate. I am not using it because it's hard to do that by hand. I am sure there are programs out there that i could use to calculate. I just never bothered with it because difference isn't big.
Aside from bookrank, all the other people that have been posting estimates in here are using same crude way. I know because i was comparing my numbers with theirs all the time. As for what that way is i have trouble explaining it with my english. I realized that i don't know the english names for even basic math terms. A basic search in internet didn't help me at all, so i am gonna explain it with a example.
If 100th place is 5,000 and 200th place is 10,000; estimates for 150th place is 7,500, 125h place is 6250,.. etc. How should i express this method in english? Using a linear scale to make estimation, is the best guess i have for now.
the only reason why I never bothered doing this is because I'm the kind of person who can't make up their mind on what method is the best, because since it is mostly mathematical, I don't think I could handle staying to a same method when I eventually found a better way
To be fair, the most important thing in charting on this scale is consistency, so as long the method is kept the same throughout, it won't be a problem and even then, the discrepancy won't be in range that will actually matter since the upper and lower boundary are already set.
basically you are using (from the few lines you wrote) interpolation based on a straight line between 2 points.(tldr: linear interpolation) which is understandable since it is the simplest way to do it.
even when I am eyeballing, I use linear for everything under 100 (since 1-50 are on oricon & for 51-100, I could use the extra jump stuff for corrections to better guesstimate)
I'm just wondering if there is a consensus on how the estimation are made.
even if I use them myself, I hated using linear interpolation for this, since we know the bigger picture is at least log.
Honestly i have not bothered with it because difference isn't big. At most it's a few hundreds. Since the point of shoseki estimates is to give people a vague idea about how series doing,I think it's negligible. For example a few comparison of my numbers with bookrank's:
Biggest problem with calculating shoseki estimates is actually something different. Regardless of the method you're using, estimates for higher ranks are extremely inaccurate. For below 100th rank, error in estimates at most a few hundreds, while in top 50 it's thousands.
Normally i use oricon numbers and only if oricon isn't available i use shoseki estimates. But it leads to a lot of problems, for example sometimes a series that didn't make the oricon list, looks higher than a series does make it. It's really a frustrating situation. The solution to this problem is actually simple, you just have to use shoseki all the time, ignoring oricon numbers. Bookrank does that. But like i said estimates for top 50 is extremelly inaccurate, so i don't want to do that. There are also times a series does top 50 in shoseki, while missing oricon, that's the most frustrating thing. For example:
Orient v4 week 2 in shoseki: 21,125
cut-off number for Oricon Top 50: 19,352
Normally i would have used shoseki estimates in this situation, but lately i started to use cut-off numbers. I think it's slighlty less annoying.
So, considering the sizable increase in sales of the manga following the anime's airing this past winter, as well as the fact that the BDs and DVDs for the anime itself sold extremely well, this shouldn't come as a surprise:
My body is so ready for this. I wonder if the manga backlog will see another boost after the second season airs. I swear all of the volumes were in the top 100 for Shoseki for at least the first month after the first episode aired, or something crazily amazing like that
My body is so ready for this. I wonder if the manga backlog will see another boost after the second season airs. I swear all of the volumes were in the top 100 for Shoseki for at least the first month after the first episode aired, or something crazily amazing like that
Honestly i have not bothered with it because difference isn't big. At most it's a few hundreds. Since the point of shoseki estimates is to give people a vague idea about how series doing,I think it's negligible. For example a few comparison of my numbers with bookrank's:
Biggest problem with calculating shoseki estimates is actually something different. Regardless of the method you're using, estimates for higher ranks are extremely inaccurate. For below 100th rank, error in estimates at most a few hundreds, while in top 50 it's thousands.
Normally i use oricon numbers and only if oricon isn't available i use shoseki estimates. But it leads to a lot of problems, for example sometimes a series that didn't make the oricon list, looks higher than a series does make it. It's really a frustrating situation. The solution to this problem is actually simple, you just have to use shoseki all the time, ignoring oricon numbers. Bookrank does that. But like i said estimates for top 50 is extremelly inaccurate, so i don't want to do that. There are also times a series does top 50 in shoseki, while missing oricon, that's the most frustrating thing. For example:
Orient v4 week 2 in shoseki: 21,125
cut-off number for Oricon Top 50: 19,352
Normally i would have used shoseki estimates in this situation, but lately i started to use cut-off numbers. I think it's slighlty less annoying.
eh, the discrepancy between oricon and shoseki is expected, since they track different amount of stores(and I'm not sure if both, either, or none of them uses a multiple for the final numbers/estimates)
Yeah, I'm still in awe about that. I'm glad that Kishuku had some form of a backlog boost as well during its run, but I wish it could've made some bigger waves than what it did. I thought its adaptation was really nicely done for the most part.
Difference between shoseki and oricon obviously expected. Problem mostly come from, going back and forth between shoseki and oricon. It creates discrepancy between sales data. Sticking with one obviously would solve the problem. But ultimately just ignoring it is the best option i guess, Estimates are called estimates for a reason, they're not very accurate in the first place.
Tokyo Tarareba Girls will get 3 more chapters as a side story, starting in the October 25 Issue of Kiss. https://www.animenewsnetwork.com/news/2018-09-24/akiko-higashimura-tokyo-tarareba-girls-manga-gets-3-new-chapters/.137217
mangahelpers.com
Go-Toubun no Hanayome and To Your Eternity won in the Shonen Category and Hajime no Ippo got a special mention award for the 110th anniversary of the creation of Kodansha.
Kanojo, Okarishimasu (Lead Color)
Toukyou Revengers
Days
Mako-san wa Shindemo Jiritsu Shinai
Enen no Shouboutai
Go-Toubun no Hanayome
Orient
Daiya no A - Act II
Bakemonogatari
Nanatsu no Taizai
Runway de Waratte
Eden's Zero
Kishuku Gakkou no Juliet
Hitman
Shinanaide Azukawa-san
Gamblers Parade
Senryuu Shoujo
Shichi ha Gojuroku
Seitokai Yakuindomo
Blue Lock
Danshi Koukousei wo Yashinaitai Onee-san no Hanashi.
Domestic na Kanojo
Soredemo Ayumu wa Yosetekuru
Fumetsu no Anata e Kindaichi Shounen no Jikenbo Gaiden: Hannin-tachi no Jikenbo (Special from Magazine Pocket)
Preview 24
Lead Color: Ahiru no Sora
--- Double Post Merged, , Original Post Date: ---
Oricon Top 50, April 29 - May 5
#9 - Nanatsu no Taizai v36 - 51,880 / 349,629
#21 - Go-Toubun no Hanayome v9 -31,015 / 351,606
#22 - Daiya no A - Act II v16 -30,972 / 209,697
#44 - Days v32 -15,920 / 108,798
There's been an update to our forum rules to broaden crediting/sourcing to all series news/announcement, etc. Please read HERE
To manage thread loading speeds and general user experience, we'd apppreciate if volume covers could be wrapped in spoiler threads, going forward. If you have any questions, reach out to a staff member. Thanks for your cooperation.
We use cookies
Essential cookies keep the site working. Analytics cookies help us improve.