The Bluebird's Feathers: Fandom, Ideology, and the Economy of a Seven-Chapter Route

Companion to "The Bluebird Who Outgrew the Throne: On Li Tai in Road to Empress."

The Paradox

Seven chapters against twenty-five. Four hours of content against a full-length route. One kiss scene against at least three. By every quantifiable metric, Li Tai's storyline in Road to Empress: Empress Chapter is a fraction of Li Zhi's. On the game's internal popularity charts, his pairing with Wu Yuanzhao ranked first by a margin so wide fans called it duàncéng dìyī (断层第一) — dominant beyond competition.

The Steam approval rating dropped to 47% on launch day. A Change.org petition asked for the original storyline to be restored as DLC. On Weibo, a dedicated "refund building" (退钱楼) thread materialized in the Yuantai supertopic, tagging the official account until it shut down its own comment section. On Xiaohongshu, the topic accumulated over 2.6 billion views.

Seven chapters, and the internet could not stop talking about them. The question is not why do people like Li Tai — the preceding essay addressed that. The question is what happens when a product quantifies devotion, breaks its promises, and discovers that the character it tried to erase has become more real to its audience than the one it kept. This essay examines the aftermath: how fan rhetoric, platform economics, and developer panic turned a seven-chapter route into the most contested site in Chinese interactive fiction.

The Rhetoric of Necessity

Every fandom war has its slogans. The Road to Empress discourse produced two that deserve to be studied side by side.

A note on terminology: what follows uses labels common in the Chinese fandom — "career-brain" (事业脑), "history fans" (历史粉), "45 CP" and "95 CP" — as shorthands for discursive positions, not organized factions. The same player may occupy several positions at once; the boundaries are porous. What matters is not who holds each view but the rhetorical structure the view produces.

The first slogan circulated among what Chinese fandom calls the shìyè nǎo (事业脑, "career-brain") position — the conviction that Wu Yuanzhao's only valid arc is the one that ends on a throne:

向下的自由不是自由。 Downward freedom is not freedom.

The second is a distillation of the same logic, phrased as destiny rather than argument:

女帝没有不做皇帝的自由。 The empress has no freedom not to become emperor.

Both statements perform the same rhetorical operation: they take a preference — I want this character to claim the throne — and recast it as a necessity. "I choose power" becomes "power chose me." Ambition becomes fate. A character who walks away from the crown is not making a choice; she is committing a category error against her own nature.

One faction argued: "武皇称帝是必然选项，既然利用了历史人物为噱头那就要受得了网友的监督和批评" (The empress becoming emperor is the only valid option. If you exploit a historical figure for marketing, you must accept scrutiny). Another accused the developers of "将权力巅峰的女性爆改为恋爱脑娇妻，消解了女性历史主体性" (transforming a woman at the peak of power into a lovesick "wife," dissolving female historical agency).

The opposing camp offered a simpler thesis: "真情无价，即使是聪慧强大如武则天，也有更爱美人不当皇帝的自由。况且《盛世天下》只是游戏" (True love is priceless; even Wu Zetian has the freedom to choose love over power. Besides, it's a game).

What makes this exchange worth studying is not who is right. It is the asymmetry in what each side treats as self-evident. To be fair: the wariness behind the career-brain position is not baseless. Chinese popular culture has a long history of reducing powerful women to romantic subplots, and the instinct to protect a rare female-emperor narrative from that gravitational pull is legitimate. The problem is not the vigilance itself.

The problem is where it leads when applied selectively: toward the premise that power is inherently superior to love, that public achievement outranks private happiness, that ascending is the natural direction and anything else is descent. At that point, the framework stops being feminism and becomes the value system feminism set out to critique — the equation of worth with rank — dressed in feminist language. Li Tai's existence threatens this framework not because he is a man who distracts a woman, but because he embodies the possibility that someone might voluntarily step off the ladder. If that choice is valid, the entire hierarchy wobbles.

And the hierarchy cannot wobble. So the rhetoric reframes voluntary choice as involuntary destiny. The empress must take the throne, not because she wants it, but because fate demands it.

Remarkably, the game's own text performs exactly this operation. At the end of the Li Tai route — after the prince has been erased and Wu Yuanzhao walks alone — a Taoist priest named Yuan Tianfang (元天方) appears and tells her that he once read her physiognomy as a child: there was "dragon energy" (龙气) above her head; she was born to be emperor. The camera follows her walking ceaselessly along mountain paths and riverbeds, implying she will resume her climb toward the throne. The narrative has just spent seven chapters demonstrating that she chose ambition over love. Then, in its final moments, it retroactively recasts that choice as destiny — she was always going to be emperor; it was written in her bones. The fan slogan "女帝没有不做皇帝的自由" is not an invention. It is a faithful reading of what the game itself wants the audience to believe. The fans did not invent the ideology from nothing. They amplified what the game had already written into place.

The irony, of course, is that Road to Empress is not history. It is an interactive fiction in which Li Tai survives his historical death, Gao Yang outlives her execution, Wu Yuanzhao romances multiple characters across parallel timelines, and the Tang Dynasty has been renamed the Sheng Dynasty to dodge censorship. Every element of the game is a fabrication. But the demand for historical fidelity is invoked exclusively against the Li Tai route — never against the liberties taken everywhere else. "Respect history" does not mean respect history. It means: this character is too popular and I need a principled-sounding reason to object.

The double standard is diagnostic. When a critical framework is applied selectively — when "historical accuracy" matters only where it constrains the character you dislike — the framework is not the real argument. It is the uniform. The real argument is underneath: why is everyone still talking about him?

The Economy of Devotion

Road to Empress monetizes affection. The game features an in-app gifting system where players can send virtual flowers (鲜花) to their preferred character pairings, creating a real-time popularity chart. They can also throw virtual eggs (鸡蛋) at pairings they oppose. Both charts are visible. Both are public. Both cost money.

Li Tai's pairing — the "45 CP" in fandom shorthand — simultaneously topped both. First in flowers. First in eggs. One faction poured resources into proving devotion; the opposing faction poured resources into proving contempt. The result was what one analyst described as "角色与现实八卦交织的特殊景观" (a unique spectacle of character and real-world gossip intertwined).

This is the fandom economy operating at its most transparent. Emotional investment — love, grief, rage — is converted into a visible, quantifiable metric. And once it is quantified, it becomes contestable. The flower chart is not a measure of narrative quality. It is a battleground. The egg chart is not criticism. It is counter-programming. Both sides are engaged in the same activity: spending real money to make a fictional character's significance legible.

Then came the refund campaigns. When the Empress Chapter launched and Li Tai's route turned out to be seven chapters of betrayal and a dream, fans did not simply post angry reviews. They filed consumer complaints on Heilongjiang's Black Cat platform. They demanded Steam refunds. Players who had exceeded the two-hour refund window — because the bait-and-switch only became apparent after 2.5 hours — were denied, and escalated.

A refund is not a review. A review expresses opinion. A refund reverses a transaction. When fans move from the flower chart to the refund form, they are saying something the developers cannot dismiss as "entitled fans": I paid for a product that was advertised and did not receive it. One English-language Steam reviewer called it plainly: "a bait and switch, false advertisement, and fraud." Another: "This was false advertisement for a non-existent storyline which was the main reason they wanted to get the game."

The game's economy of devotion runs on a specific transaction: invest emotionally, receive narrative payoff. The Li Tai route collected the investment and defaulted on the payoff. The flowers were not refundable. The eggs, at least, turned out to be prophetic.

The Amplification Loop

The developers saw a problem. On-screen, Li Tai was consuming more audience attention than the protagonist. Online, the "大泰传" (Legend of Tai) nickname — a sarcastic rebranding implying the game should have been named after him — had its own curated list on Douban. Accusations of favoritism multiplied: the character was an "亲儿子" (developer's pet), the actor had "走后门" (pulled strings). Sina reported complaints that the game "篡改李泰历史" (falsified Li Tai's history) by fabricating imprisonment and suffering that never occurred in the historical record.

The developers' response was to cut. Scenes previewed in Part One's finale were deleted. Thirty-plus planned chapters became seven. The entire romance was reframed as a hallucination. The launch-day statement cited "资源有限" (limited resources) and "坚守创作初心" (staying true to creative vision) — addressing neither the broken promise of equal treatment for both routes nor the specific content that had been removed. When the backlash intensified, the official Weibo account shut down its comment section. The pressure did not dissipate; it redirected. Fans began tagging individual production staff on Xiaohongshu — including the scriptwriter credited with the Empress Chapter — demanding answers the official channels refused to give. Block one outlet and the water finds another.

This was a misdiagnosis. The developers identified the source of Li Tai's popularity as screen time and prescribed the obvious cure: less screen time. But screen time was never the engine. The engine was character writing — the specificity of his desires, the transparency of his emotions, the gap between what he understood and what he chose anyway — amplified by an actor whose performance exceeded the script. You cannot kill that by cutting minutes. You can only make it worse.

And they made it worse. Every cut fed the cycle:

The route is shortened; fans feel the character was treated unjustly; sympathy deepens; the community generates its own content to fill the narrative gaps; the character becomes more psychologically rich in the fandom than in the game; new audiences encounter this enriched version; they play the route; they are outraged by the disparity between what the community built and what the developers delivered; they join the chorus.

Each turn of the cycle amplifies the last.

Bilibili creator 赛万味超人's commentary series on the Empress Chapter — episode titles include "改编可以乱编，戏说可以胡说，请不要按历史来" (Adaptation can be wild, fictionalization can be nonsense, just don't pretend it's history) — accumulated hundreds of thousands of views. The discourse ecosystem around Li Tai now dwarfs the actual content the game provides for him.

A telling detail — presented here as fan narrative, not verified fact. Road to Empress began life as an Orange Light (橙光) visual novel by author "肉肉" (Rourou), where Li Tai was a straightforward antagonist with no romance line — "和女主的感情线基本等于没有，主要就是相互利用" (essentially no romantic arc with the protagonist; purely mutual exploitation). The live-action adaptation reinvented him as a love interest. When the Empress Chapter's writing drew accusations of quality collapse, fans constructed an explanation: the Part Two script had been handed to writers sympathetic to the rival pairing, who deliberately sabotaged the Li Tai route. The original Orange Light author's involvement was recast as the smoking gun. Whether any of this reflects actual production decisions is unverifiable from public sources; that fans needed the theory is the point. A conspiracy of authorial malice is easier to metabolize than the possibility that the production simply ran out of craft. The amplification loop requires a villain, and when the developers' launch-day statement ("limited resources, staying true to creative vision") offered no satisfying antagonist, the fandom invented one.

This is the paradox the preceding essay described from the textual side: "A character the writers kill, unjustly, conspicuously, against the audience's will, becomes something else entirely. He becomes myth." The fandom ecology is the proof. If the developers had delivered the promised thirty-chapter route with a happy ending, Li Tai would have been a popular love interest. Satisfying. Complete. Forgettable within a year. By killing his story, they made him immortal.

The Hollow Product

The gutting of the Li Tai route was not an isolated creative decision. It was one symptom of a broader production crisis that hollowed out the Empress Chapter while leaving its surface intact.

The game marketed itself with "4K实拍、电影级服化道、每套冕服工期45天" — 4K live-action filming, movie-grade costumes and props, forty-five days of labor per imperial robe. One Steam reviewer called it "by far the highest production quality FMV game I have played." Then came the Empress Chapter.

End-credit animations that looked like "劣质AI跑出来的" (inferior AI output). Illustrations described as "AI尸块拼贴" (AI corpse-piece collage). AI-generated soldiers marching into the palace during the main ending. English translations peppered with jarring machine-generated phrases — "Go big or go home," "spill the beans" — in a Tang Dynasty court drama. None of this disclosed on the Steam store page. Another reviewer: "I wouldn't have bought and played this if the genAI use had been properly disclosed." A third: "Your resort to using AI generated imaging and video cheapens and I would go so far to say as ruins the great games and story I grew to love."

The AI controversy overlapped with a separate production crisis. In January 2026, Peng Shiliu (彭十六), the actress playing Shangguan Wan'er — a character threading through roughly 40% of the game — was penalized for tax evasion and widely treated in coverage as a "disgraced artist" (劣迹艺人). The studio faced three options: AI face-replacement at an estimated 500,000 yuan per minute with uncanny-valley risks in close-ups, deleting her scenes, or reshooting with a new actress. The planned Spring Festival 2026 launch was scrapped; the game was delayed to June. Players who noticed the digital substitution were not told it had happened.

Premium costumes, AI-generated endings. Forty-five-day robes, undisclosed face-swaps. A production that sold the handmade and delivered the automated. The Li Tai route was not singled out for hollowing; it was simply the place where the hollowing became impossible to ignore, because the audience that cared most was also the audience paying closest attention.

When a product's substance collapses but its packaging holds, someone has to finish the story. In this case, the audience volunteered. Bilibili creator 某幻君 — whose Part One playthrough drew sixteen million views — sits through the clinic ending, where Wu Yuanzhao makes medicine and saves lives on ground fertilized by the people she sacrificed, and asks the question the script refuses to ask: do you ever think about the Prince of Wei and Gao Yang, who died for this? That question is not in the game. A Douban user frames the core grievance more bluntly: "cpf不允许玩家选择（活着脑补）李泰" — fans being forbidden from even imagining that Li Tai might survive. Taiwanese players on Bahamut ask the question the game's structure forecloses: "如果歷史沒有如果，我們是否仍會選擇成為武則天？" (If history has no what-ifs, would we still choose to become Wu Zetian?)

These are not idle complaints. They are the audience doing the work the product charged money for and did not perform. Somewhere between the forty-five-day robes and the AI-generated palace siege, the developers stopped finishing their own story. The audiences did not.

Coda: The Feathers

A player finishes Li Tai's route. Seven chapters. Four hours. A dream ending that erases everything. They close the game. They open their browser. They begin searching.

They find 2.6 billion topic views. A character who simultaneously topped the love chart and the hate chart. Petitions from Switzerland. A commentary series whose episode titles read like chapter headings from a novel about narrative betrayal. A streamer who paused mid-playthrough to ask the question the writers would not. An entire ecosystem of grief and devotion built around a character who received a fraction of the protagonist's screen time and somehow accumulated more emotional infrastructure than the protagonist herself.

And perhaps the most revealing evidence is the reaction of those who insist none of it should have happened — their fury directed not at a bad character but at a successful one. "向下的自由不是自由" was always a confession disguised as a slogan. It does not describe the empress. It describes the people who cannot tolerate anyone choosing a direction they did not authorize.

The developers can declare the story a dream. The feathers are already everywhere.

The fans already woke up.