The Daily Signal
Technology

Streaming Algorithms Built Champagne and Bullets Canonicity, Not Audiences

Margot·Monday, July 27, 2026 Edition
When Algorithms Replace Actual Taste

Champagne and Bullets is being positioned as canonical bad cinema—ambitious beyond its means, resourced beyond its competence, self-aware below the threshold of functional.

The Verge published this claim in its technology section, which is already a signal worth reading. The argument seems settled. It deserves a place alongside The Room and Troll 2 as a film so fundamentally broken it achieves an accidental grace.

The argument assumes that "so bad it's good" is a stable category—films that objectively fail in their stated ambitions yet succeed in some other, more interesting way. But watch what actually happened to The Room. It premiered in 2003 to indifference, genuinely unwatched for years, a failure with no cult.

When algorithms decide what's cult

Only after midnight screenings created a social container around it—only after repeat viewings turned the film into shared ritual—did its "so bad it's good" status crystallize. The failure was always there. The category appeared later, retrospectively applied and shaped by the specific communities who elected to watch it together, argue about it, celebrate its incompetence as virtue. Streaming platforms have collapsed this timeline entirely.

A film can now be algorithmically surfaced to audiences primed to recognize it as cult material before any actual community has formed around it. Champagne and Bullets did not earn its canonical placement through decades of midnight screenings or devoted fan reconstruction—it was algorithmically recommended into visibility. That visibility was then interpreted as proof of its essence. This is not a revelation about the film. It's a problem with how we now validate taste. We mistake discovery for judgment. When a streaming platform surfaces an obscure film to exactly the audience most likely to celebrate its failure. That audience celebrates it, we call that canonicity, but we're not evaluating the film at all—we're watching the algorithm succeed at its actual job of creating demand by matching attention-scarcity with novelty-hunger.

We mistake discovery for judgment.

The film's quality, mechanical or otherwise, becomes almost irrelevant. What matters is whether it satisfies the recommendation engine's prediction about what you'll click. You've learned to trust those predictions. You're building your taste around them without noticing.

Related Stories
Science
Two Measurements, One Universe, Zero Agreement
The universe's expansion rate produces two incompatible measurements—73 versus 67 kilometers per second per megaparsec—that stubbornly refuse to converge despit
Film
Marvel Adds Celebrity Oxygen to Weakening Franchise Formulas
Marvel Studios isn't pivoting toward talent-driven casting—it's compensating for IP exhaustion by layering celebrity names onto existing franchises that can no
Film
Three Fantasia Films Don't Confuse Audiences, They Demand Complicity
The critical claim that You Are the Film, Someone's Daughter, and Los Vampires blur documentary and narrative so completely that audiences lose track of what's
More From Today's Edition
Entertainment
Pogacar's Five Titles Meaningful Because Armstrong's Seven Were Deleted
Tadej Pogacar's fifth Tour de France win is celebrated as proof cycling has moved past doping, but this narrative only works by ignoring that the sport deleted
Culture
Gerry Kelly's Interview Silence Technique Never Made the Obituary
Legacy announcements capture dates and titles but miss the specific methods that actually shaped a field—in Gerry Kelly's case, an RTÉ Radio 1 interviewer who d
View Past Editions >