0.1 Percent: The Arithmetic of a Film That Refused to Disappear

By early October 2026, Niu Lai had become a fractions problem. The film, a hand-made animated feature that opened on August 5, was still playing under a second key extension that carried its screenings to November 4. Daily takings had fallen to about 35,000 yuan, and both the film's share of national screenings and its share of the national box office had dropped below 0.1 percent. The cumulative total stood at roughly 64.18 million yuan as of October 7.

Those numbers are easy to nod at and hard to feel. Feeling them requires a denominator, and the denominator is public. On January 1, 2026, China's Film Bureau released its annual data: the country's urban cinema chains had ended 2025 with 93,187 screens, and audiences had bought 1.238 billion tickets over the year. Divide the year out and the market runs at roughly 3.4 million admissions a day across those 93,187 screens.

Now place the October numbers against that denominator. The film's distribution notice sets a settlement standard of 20 yuan per admission for digital 2D. At 35,000 yuan a day, Niu Lai was selling on the order of 1,750 tickets daily. In a market absorbing 3.4 million admissions a day, that is about five hundredths of one percent of the country's moviegoing: one admission for every two thousand the market as a whole was selling. A film whose profile had been large enough to trend repeatedly in August was, by October, statistically invisible.

The curve, in three points

The distance traveled is clearest in three numbers. On August 15, after a hashtag about the film's collapsing box office drew national attention, a single day took 717,000 yuan across 2,322 screenings. On August 16, the peak, takings reached 6.087 million yuan across roughly 14,000 screenings. By early October the same film was taking 35,000 yuan a day, with a screening share below one in a thousand. That is a fall of about 95 percent from the August 15 day in daily revenue, and more than 99 percent from the August 16 peak, with the screening count down by a comparable margin.

A run that begins with 236 admissions in nine days and peaks above six million yuan in a single afternoon has already won something no business plan predicted. The last month of such a run is where the interesting mechanics live, because nothing about the decision to keep screening was irrational.

Why a 0.1 percent film keeps its slot

The first question a reader asks is why anyone bothers. The answer is that every party to the decision is spending close to nothing.

Key production services are charged as a percentage of box office revenue rather than as a flat fee, so the cost of keeping a film authorized scales down with its earnings. The distribution side has already manufactured the digital cinema package, and delivering a key to a theater that asks for one is a marginal operation rather than a project. Cinemas carry their own logic. A screening in an off-peak slot costs the same whether the auditorium holds thirty people or three, and a title with a small but proven following fills seats that would otherwise sit empty. The managers who rebuilt their August schedules around audience requests had little reason to evict those audiences in October.

The mechanics underneath are more precise than the word extension suggests. A digital feature travels as an encrypted package, and each theater's playback server holds a certificate that unlocks it for a fixed date range. The key that authorizes a screening is matched to a specific server and a specific window, so extending a run means issuing fresh keys to every server still carrying the film. The service is priced as a percentage of box office revenue, which is why the cost of keeping a film authorized shrinks along with its earnings. For a title taking 35,000 yuan a day, the authorization fee is small against the revenue it protects.

There is a contractual dimension as well. A cinema run is exclusive by agreement, and the production side may not license a platform or release a physical edition while the run continues. A theater still selling tickets is, from the distributor's side, both revenue and evidence that the exclusivity is being honored. The film was under no obligation to leave a screen while anyone was still buying a ticket to see it.

What the fraction costs is measured in a different unit. Every day held in theaters is a day the post-cinema life has not started, and the audience that made the run possible has already bought its tickets. Several commentators argued that an earlier online release might have earned more than another month of near-empty screenings. The claim cannot be proved, since the alternative was never run, but the logic behind it is sound and the reporting around it has been consistent.

A market of extremes

The denominator does something else to the story: it places the film inside a market that is itself extreme. In 2025, according to the Film Bureau figures reported by People's Daily, China's box office totaled 51.832 billion yuan, eight films passed one billion, and four of the year's top ten titles were animated, led by a Ne Zha sequel that earned more than 15.4 billion. Against that backdrop, a hand-inked feature earning 35,000 yuan a day in the same market is an outlier at the edge of the system. It is also evidence of how wide the system has become. The same 93,187 screens that carried the year's biggest animated hit past 15 billion were, in the following year, carrying a two-person production at five hundredths of a percent of daily admissions, and both facts describe the same industry.

The comparison cuts both ways, which is the point. A market wide enough to hold a 15 billion animated phenomenon and a 1,750-ticket day is a market with room for experiments that no studio committee would approve. The infrastructure that makes the blockbuster possible is the same infrastructure that keeps the experiment visible, at 0.1 percent, for as long as anyone asks for it.

The same annual data points to a second distribution layer that rarely enters these calculations. Beyond the urban chains, the Film Bureau counted 264 rural digital cinema lines operating 44,000 projection teams in 2025, staging eight million public screenings for an audience of about 394 million. A two-person hand-inked feature aimed at families and children has an obvious home in that layer, and the layer runs on its own economics, built around public service rather than ticket revenue. Whether a film of this shape ends up there after its cinema run is one of the quieter questions its extension leaves open.

What the extension was, and what it was not

Stating the record plainly: extending a key is a permitted and routine step that several films take every year. Cinemas choose their own schedules. Audiences who wanted to see the film could still buy tickets through November 4. The criticism that the production side was overstaying its welcome is a judgment about taste rather than a finding about conduct, and reasonable people land on both sides of it.

The structural finding is different from the moral one. A cinema ledger counts yuan per day, and on that ledger a month of small screenings is a small positive. A release strategy counts momentum, and on that ledger the same month is a subtraction. The mistake available to any observer is to read only the ledger that posts a number.

The honest summary

0.1 percent is not a failure metric. It is what a film looks like when the machinery of release is working exactly as designed, after the audience has gone home. The interesting shortfall belongs to attention rather than commerce. A film that 93,187 screens can still show and almost nobody will still see is no longer a release in any meaningful sense. It is an archive entry with a key attached, and the second extension moved that date by a month without changing what it was.

---

Figures, names, and characterizations above are drawn from overseas and Chinese media coverage compiled as of October 2026, including China Film Bureau data (People's Daily), Tencent News, Sohu Entertainment, and distribution notices. They are presented as online reports and viewpoints and have not been independently verified.