The Lowest-Scoring Films of 2026

A bad film is almost never bad everywhere. That is what makes writing about them worthwhile and what makes a snark list useless. Of the eight below, six have a genuinely accomplished element (a performance, a score, a first act) and every one failed for a reason you can name and point at. None failed because the people making them were lazy. They failed because one decision went wrong early and nobody caught it.

Below 5.0 on our scale, a film has usually stopped functioning rather than merely disappointed. Here are the eight that went furthest under this year, and what went wrong each time.

How a Film Ends Up Below 5.0

Our scale runs to one decimal place and the bottom band is narrow by design. A 5.5 is a boring film: competently made, no reason to exist. A 4.5 is a broken one: the seams show, the decisions contradict each other, and the thing stops holding together in the room. Below 4.0 you are watching a production actively work against itself.

What pushes a film across that line is rarely the acting, and rarely the photography, which is why so many of these look beautiful in isolation. It is structure. A film is a machine for spending attention, and every one of the eight below makes the same error in a different costume: it spends the audience’s attention on something it never pays back.

1. The Glass Harbour: 3.1 / 10

Noor Haddad’s port-town drama opens with real authority: a shipping firm collapses overnight, four hundred jobs go with it, and the camera stays on the people who were told at 6am. For forty minutes it looks like the best film of the year.

Then it stops deciding. The fraud thread, set up with three scenes of legwork, is dropped about seventy minutes in and never mentioned again. The film slides sideways into a family story it has not built, and ends on a confrontation in which nobody says the thing everyone has been circling. The diagnosis is a script that never chose: there are three films here (a procedural, a family drama, a town portrait) competing for the same two hours. Remove any one and the whole thing tightens by a full point.

2. The Ninety-Ninth Floor: 3.4 / 10

A premise so clean it hurts: two strangers trapped in a stopped lift on the ninety-ninth floor of a half-finished tower, with a bad radio and a storm coming. Tobias Rennick shoots it beautifully, and Priya Menon and Rowan Pike commit entirely to the confinement.

The problem is arithmetic. That premise contains one act. Ninety minutes in a box needs a steady drip of new information or a steady escalation of pressure, and Rennick has neither. At minute fifty-five the characters begin repeating earlier arguments in different words; at minute eighty-five somebody opens the door from outside and the film ends on a shrug. Every ingredient is right and the container is the wrong size: a twenty-two-minute short inflated because shorts cannot be sold.

3. Midsummer Static: 3.6 / 10

Camille Ferrand’s film announces itself as slow cinema and then mistakes that for a genre rather than a method. There are seven minutes, early on, of a woman making coffee. Nothing happens in them, nothing is revealed, and the camera sits at a distance that makes the frame feel like an accident.

The tell is what the film does with silence. In a genuinely patient film, a long silence is where the scene does its thinking. The audience works, filling in what the characters will not say. Here the silence is where the film does nothing at all. Subtext requires something underneath the text, and Midsummer Static has quiet on top and more quiet below. Slowness is the price you pay for depth; a film that charges the price and delivers no depth is not slow cinema, just long.

4. Copperhead Run: 3.8 / 10

Dmitri Vale’s desert thriller is a well-made film accidentally assembled around nothing. A courier, a stolen case, two hundred miles of asphalt, three parties who want the case. Marianne Trott is extraordinary as the courier, and Nikolai Petrov does career-best work largely through a windscreen.

The film never establishes what is at stake, for anyone. Vale withholds so thoroughly that he withholds the reason to care, and by the time the pursuit begins the audience is watching strangers drive with no preference about who wins. This is the cruellest failure mode here, because the craft is real: practical stunt work, an angry sound design, all of it in service of a film with nothing driving it.

5. All the Wrong Weather: 4.0 / 10

Hannah Okpara spends eighty minutes building something genuinely tense: a wrongful dismissal claim turning on one document, with Tamara Besson as the lawyer who knows it is real and cannot prove where it came from. Quiet, procedural and gripping.

Then, at minute eighty-two, the film walks away from its own case. The hearing is abandoned off-screen. Besson and Felix Aurand drive north for twenty minutes while the film attempts a reconciliation neither has earned, in a register it has not used at any point before. The document is never mentioned again. The road trip is not a bad idea for a film; it is a bad idea for this one.

6. The Vanishing of Ella Fray: 4.1 / 10

The original, a Portuguese film from 1974, withholds the fate of a missing girl for a reason: the village’s refusal to answer is the film’s argument about what a community will admit. The mystery is not a puzzle. It is a moral position.

Rafael Sousa’s remake keeps the withholding and discards the argument, denying the audience information purely for atmosphere, because fog reads as serious and an unanswered question feels more sophisticated than an answered one. Juno Kestrel is badly miscast in the central role, playing grief at a pitch the film cannot hold. The crime of the misread source is not changing the material; it is reproducing its shape with the meaning taken out.

7. Tidewater: 4.3 / 10

A marshland horror film that has decided fog is a story. Ingrid Martensson can compose a wide shot (the drowned fields and grey water at dusk are the best argument for the film) and there is one night crossing of a flooded causeway that belongs in a much better one.

Everything around it is atmosphere with nothing inside it: ninety-four minutes, roughly fifty of them shots of water, mist and empty rooms, used to fill time between the three moments that matter. Horror needs a mechanism: a rule, a cost, a thing that will happen if the characters do the wrong thing. Tidewater has a monster you barely see and no rules at all, so nothing anyone does can be wrong and no scene can be tense. Mood is a multiplier, not a subject.

8. Last Train to Cortina: 4.4 / 10

Julius Kane’s snowbound whodunnit is the highest-scoring film here and the most frustrating, because it is close to working. The carriage is a wonderful set, the first hour is genuinely gripping, and Margarethe Lund gives the best performance of her career as the conductor.

The failure is that the solution does not matter. The reveal turns on a relationship between two characters who have had ninety seconds of screen time between them, and the film has spent no runtime making it feel consequential. Every performance in the final twenty minutes is doing exceptional work on material with no weight. Nine good performances do not add up to a story, and a whodunnit is a promise that the answer will matter.

The Failure Modes, Named

Laid out together, the eight stop looking like eight separate accidents. There are six recurring faults, and between them they account for the entire bottom of our 2026 list.

The Unmade Decision kills outright: three films in one script, and no edit can save it. The Stretched Short is the most common mid-budget failure, because a premise with one act of fuel is the easiest thing to get financed. The Atmosphere Substitution arrives attached to real talent, which is why it produces the most argument. The Well-Acted Void is the most likely to be defended, since the performances are visible and genuinely good. The Third-Act Defection and the Misread Source are failures of nerve rather than skill. None of the six is really about money. Two of the eight cost under ten million and look like they cost five times that.

What a Bad Film Can Teach You, and When It Is Just a Waste

The honest position sits between the two. A broken film is a fine object of study and a poor use of a Saturday. Watch the first act of Copperhead Run, because the craft is real and instructive. Watch all of Last Train to Cortina only if you want to see how a promising film dies at the end, in front of you, politely.

The Verdict

Reader FAQ