Summary: A joke about Lumiore’s extremely readable “dangerous big-sister dragon” design turned into a production-workflow investigation. The public evidence points away from a lone illustrator acting freely and toward a collaborative pipeline: planners, scenario staff, art directors, illustrators, production staff, and animation designers all contribute. Blue Archive goes further, openly describing a creative culture that combines what the staff themselves genuinely like. Personal taste can become an organized production process.
0. Starting point: “Why does Lumiore read like a succubus archetype?”
Lumiore’s visual language is immediate: gold, wing- and horn-like shapes, a confident older-sister personality, and a costume silhouette that strongly emphasizes body lines. Those signals can easily recall the familiar fantasy succubus vocabulary.
Official material says otherwise. In Shadowverse: Worlds Beyond, Lumiore & Argente is Dragon class, and its flavor text describes a nation of dragonfolk and a one-winged family born from a four-winged lineage. The accurate description is therefore not “she is a succubus,” but a dragonfolk character whose visual design overlaps with signals audiences may associate with the succubus archetype.
The official card page credits Koyafu as illustrator. That answers who drew the WB card, not necessarily who invented every part of the original character concept.
1. “Who drew it?” and “Who designed it?” are different questions
Commercial character production can separate the need for a character, role/personality/age/gender, visual direction, concrete illustration, art supervision, and animation/implementation.
A public Cygames workflow interview about Shadowverse leader skins shows a kickoff involving planners, scenario planners, production management, the art director, and animation designers. The planner’s brief can include personality, atmosphere, age, gender, illustration concept, pose references, and other reference material.
The normal model is therefore concept → brief → illustration → supervision → approval, not “give the artist total freedom and publish whatever comes back.”
For Lumiore, public evidence firmly identifies Koyafu as the card illustrator. It does not identify a single public author of every original design decision or reveal exactly which details came from the internal brief versus the illustrator’s proposals.
2. Does a meeting actually decide “make this character a woman”?
Published workflow makes it clear that choices such as gender exist upstream: gender can be written into the brief alongside age impression, mood, role, pose, and visual concept.
What is not public is the meeting transcript everyone would love to read:
- “Strengthen the leg silhouette.”
- “This slot is the powerful big-sister type.”
- “Make the one-winged dragon readable in 0.3 seconds.”
- “Approved.”
There is no evidence of a literal “fetish confession meeting.” The funnier documented fact is that the finished design passes a multi-person commercial review process. Taste can end up inside quality control.
3. Shadowverse characters are built to be readable in seconds
The current WB character page presents eight major figures: Lovesign, Marie Malisse, Kaori Yuihara, Galan, Diawl, Esperanza, Dreizehn, and Eudie. The official descriptions make five of them women/girls, two men, and one a monster.
Their archetypes are highly legible: elegant noble executioner; futuristic flame powerhouse; mysterious mature woman; artificial girl with twin guns; strange senior guide; decadent prince; elderly herbalist; fog-born monster.
That is not necessarily shallow design. It is information compression. In a mobile card game, hair, palette, clothes, weapon, expression, pose, and motif all function as UI so that “this one is for me” can happen before a player reads pages of lore.
4. Blue Archive practically explains the “democracy of taste” out loud
Blue Archive executive producer Kim Yong-ha has described two major character-creation routes: a designer proposes a visual/concept first, or scenario staff request a character with a needed role/personality. The design and story sides then develop the character together.
He also recalled an early concept discussion in which a school setting was attractive partly because it allowed many uniforms, with a scenario writer even teasing him about liking uniforms. Kim said the current Blue Archive emerged from combining what all the staff, including himself, genuinely like.
The important principle is not “show more skin.” Glasses, uniforms, hairstyle, mannerisms, voice, relationships, personality, and silhouette create a huge design space beyond exposure alone.
This section is about production philosophy. It is not an invitation to sexualize Blue Archive’s student characters.
5. A committee of nerds can outperform one person’s obsession
Kim described an earlier project where his own preference dominated too much and four of five main characters ended up wearing glasses. He later became more conscious of drawing a line between personal taste and final product design.
One person can overconcentrate a portfolio. Several people can distribute it across cool, gothic, glasses, uniforms, mecha, animal traits, older-sister energy, princes, elders, monsters, and more.
So “democracy of taste” is more than a meme. It can function as a character-portfolio strategy: the product does not need everyone to love the same character. It needs enough different people to find their character.
6. “Why so many girls?” is an industry question, not an ethnicity question
Saying “Asians are perverts” explains nothing and turns a production/business pattern into an ethnic stereotype.
A better explanation is that game and animation industries in Japan, Korea, China, and elsewhere in East Asia have spent decades refining ways to convert attachment to characters into retention, merchandise, fan art, events, and long-term IP value.
The remarkable thing is the industry’s ability to turn “I like this exact kind of character” into a requirement detailed enough to survive briefs, review, marketing, and monetization.
7. Is the word hentai itself Japanese?
The written term 変態 is older than its modern sexual slang sense. Historically it means a change of form or state; Japanese dictionaries cite usage around 900 and also older Chinese classical sources. It remains a technical word in biology and materials science.
In modern Japan, a sexual sense developed through expressions such as 変態性欲. Krafft-Ebing’s Psychopathia Sexualis was published in Japanese in 1913 under the title 『變態性慾心理』. A Japanese dictionary cites a 1931 literary example for 変態 used by itself in the sexual sense.
So the old Sino-Japanese word was not invented in Japan, while the modern sexual meaning is strongly tied to modern Japanese usage. Historical pathologizing sexology should not be used to declare contemporary orientations or identities “abnormal.”
8. The scariest character ranking was the one AI invented
During this discussion, an AI-generated poster confidently presented Dreizehn #1, Kaori #2, Eudie #3. It looked polished enough to be official, but the ranking had no data behind it. That is a classic generative-AI failure: visual completeness creates an illusion of evidentiary completeness, so the ranking was withdrawn.
The closest current official popularity signal is the 2026 Shadowverse Awards category for favorite leader skin. Voting ran June 1–7, 2026, and the result graphic shows:
- Orchis
- Adahime
- Filene
This is not an all-character ranking. Lumiore’s WB exchange ticket was added on June 29, after voting closed, so the result cannot be used to say Lumiore “lost” that poll.
9. Lumiore cannot be explained as one illustrator going rogue
What can be verified: Koyafu is credited for the WB card; Dragon class and dragonfolk lore are official; Lumiore received an exchange ticket on June 29 with leader-skin, card-style/home-background, and avatar-costume options; and Shadowverse production publicly uses a multi-role planning and supervision pipeline.
What cannot be verified is the identity of a single person who first said “Lumiore should look exactly like this,” or exactly which costume and silhouette details originated in the brief versus the artist’s own design work.
The most accurate joke is therefore:
This was not proven to be one artist’s crime. It passed organizational review.
10. Conclusion: “democracy of taste” is rational character business
Strong character products communicate their archetype immediately, reinforce the first impression through lore, allow multiple creators to add different tastes without breaking the world, and expand into cards, skins, backgrounds, goods, and stories.
Shadowverse demonstrates the collaborative structure. Blue Archive is unusually explicit that staff members’ own “likes” are part of the creative fuel.
Did the meeting literally become a fetish-confession tournament? We do not have the minutes. But do teams sometimes turn creators’ specific tastes into organized, reviewed, marketable design? Absolutely.
Which leaves one final thought whenever a particularly committed design reaches the screen:
“Multiple people looked at this and said yes. Incredible.”
FAQ
Q1. Is Lumiore a succubus?
No. Official WB material places her in Dragon class and describes dragonfolk lore. “Succubus-like” is an observation about visual language, not canon.
Q2. Who illustrated Lumiore & Argente in WB?
The official card page credits Koyafu.
Q3. Did Koyafu create the original character design too?
That cannot be established from public credits alone. Card illustration, concept, internal brief, setting, and art direction can be separate contributions.
Q4. Does one person design every Shadowverse character?
No public evidence supports that model. Published workflow material shows multiple specialties collaborating.
Q5. Does Blue Archive really use staff preferences as creative input?
Kim Yong-ha explicitly described combining the staff’s own “likes,” while also discussing the need to keep any one person’s preferences from dominating the whole product.
Q6. Is Orchis the official #1 Shadowverse WB character?
Not in a universal ranking. Orchis placed first in the 2026 favorite leader skin category; Adahime was second and Filene third.
Q7. What was the earlier AI ranking?
An unsupported generated ranking. It looked convincing but was not based on polling data, so it was corrected and withdrawn.
References (12)
- https://shadowverse-wb.com/ja/deck/cardslist/card/?card_id=10844120
- https://shadowverse-wb.com/ja/news/detail/?id=01KVFAGHSRGF2XWHVGMVG174JN
- https://shadowverse-wb.com/ja/news/detail/?id=01KVCDEM37YGTQ70Z1C6N68Q09
- https://shadowverse-wb.com/ja/character/
- https://gamemakers.jp/article/2024_02_15_57227/
- https://recruit.cygames.co.jp/seminar/2758/
- https://news.denfaminicogamer.jp/interview/250428b
- https://news.denfaminicogamer.jp/interview/250428b/3
- https://kotobank.jp/word/%E5%A4%89%E6%85%8B-131270
- https://ci.nii.ac.jp/ncid/BN02900766
- https://shadowverse-wb.com/ja/news/detail/?id=01KSRY5ZVY5XV8N4JG2XC4E1FC
- https://anniv.shadowverse-wb.com/ja/survey-results/
