Part 5 — The Mediocre
Series Navigation
Part 1: The Masterpieces |
Part 2: The Excellence |
Part 3: The Solid Foundation |
Part 4: Niche Brilliance |
Part 5: The Mediocre
Mediocrity is the most interesting failure mode. A film that simply doesn’t work teaches you nothing useful. A film that almost works — that gets something important right and then sabotages it with something catastrophically wrong — tells you exactly where the traps are.
The seventeen films here all have real achievements. Most of them have genuine moments of excellence. Several made hundreds of millions of dollars. A few have passionate defenders. What they share is a specific flaw — poor character work, narrative incoherence, a third act that betrays the first two, a concept that the film lacked the courage or craft to follow through on — that prevents them from reaching what they were clearly trying to become.
Some of these will surprise you. High IMDB scores don’t mean much here — popular is not the same as good, and this is a list about quality, not box office. A few of these films I enjoyed watching. That doesn’t make them good science fiction.
Writers looking to avoid common pitfalls will find the craft framework in the Worldbuilding Handbook.
1. Interstellar (2014) [full review]
⭐ IMDB: 8.7/10
“Love is the one thing we’re capable of perceiving that transcends dimensions of time and space.”
The first two acts of Interstellar are among the best science fiction filmmaking of the past twenty years. The visual realization of the black hole, the relativity sequences where characters age differently across time zones, the bone-deep fear of wasted time that the father-daughter structure creates — Nolan understood what he was building and built it with genuine craft.
Then comes the tesseract. Cooper falls into Gargantua and emerges in a five-dimensional structure built by “future humans” where he can manipulate time by sending messages through gravitational anomalies. The mechanism by which he communicates with his daughter — whose detection of these messages ultimately saves humanity — is love. Love transcends dimensions. Love is the force that operates faster than light across spacetime.
The film earned the right to a hard ending. It earned a conclusion that respected the physics it had spent two hours establishing. Instead it got a greeting card. The 8.7 IMDB rating is real — people love this film, and I understand why — but the ending retroactively makes the scientific rigor of the earlier sequences feel like setup for a betrayal. Nolan spent two hours teaching you to take physics seriously and then told you love breaks the rules. It doesn’t.
The film’s failure is in the third act’s resolution, not its premise. The problem: Nolan needed Cooper to communicate with Murph across time, which requires some mechanism. He chose “love as a force.” The available alternative was already in the film — Cooper’s watch, the gravitational anomalies, the specific data Murph needs to complete Brand’s equation. A better script would have kept the communication mechanism physically grounded in what the tesseract actually is: a structure that lets Cooper manipulate gravity at specific points in spacetime. He doesn’t need to transmit love. He needs to transmit a number. The number was always the answer. Saying love is the force that makes the number transmissible doesn’t add meaning — it dissolves the film’s own framework. The emotional truth Nolan was reaching for (a father’s determination to reach his daughter transcends everything) could have been carried by the physical specificity of what he actually does, without invoking a force that contradicts the physics the film spent two hours respecting.
2. Avatar (2009) [full review]
⭐ IMDB: 7.9/10
“I see you.”
James Cameron spent twelve years developing technology that could render a convincing alien world from scratch, and he succeeded. Pandora is visually extraordinary — a living planet with coherent ecology, bioluminescence, floating mountains, creatures that feel evolved rather than designed. As a demonstration of what digital cinematography can do, Avatar is a genuine achievement.
The story is Dances with Wolves. Not similar to Dances with Wolves — it is Dances with Wolves with blue people. Jake Sully is John Dunbar. The Na’vi are the Sioux. The unobtanium mine is the American frontier. Cameron spent twelve years building a new world to tell a story he could have told in 1850. The “white outsider becomes better at being indigenous than the indigenous people” structure is not subverted or complicated — it is delivered sincerely, at 162 minutes, as though no one involved was aware of the discussion.
The film made $2.9 billion. That number tells you a lot about the power of visual spectacle and almost nothing about the quality of the narrative. Cameron is unquestionably a master of cinema technology. He is not, in this film, a master of storytelling.
The source material was already solved. Dances with Wolves, Pocahontas, The Last Samurai — the outsider-goes-native structure is so thoroughly explored that using it unreconstructed in 2009 is a choice, not an oversight. A better script would have challenged Jake’s heroism: if you have to become one of the people to fight for them, are you saving them or replacing them? The film gestures at this when Jake is rejected by the Na’vi — that’s the right scene — but resolves it too quickly through competence rather than genuine reckoning. The other missing element is any Na’vi perspective on what Jake represents. Neytiri falls in love with him, but Tsu’tey’s resentment — the correct response to a colonizer being handed leadership of the people he helped colonize — is treated as mere obstacle rather than as the story’s actual moral center.
3. Contact (1997) [full review]
⭐ IMDB: 7.5/10
“They should have sent a poet.”
Contact wants to be a film about the tension between scientific empiricism and religious faith, with Ellie Arroway representing the scientific worldview. The first two hours are good at this — the SETI work is realistically portrayed, the bureaucratic and political obstacles feel authentic, Jodie Foster grounds the character in something recognizably human.
The problem is that the film resolves the tension by giving Ellie a religious experience. She travels through the wormhole, encounters something vast and incomprehensible, and returns transformed — but with no physical evidence of her journey, only her conviction that what she experienced was real. The film then asks whether we should believe her, on faith, despite the lack of evidence. The science advocate is converted, and the conversion is presented as growth rather than defeat.
Carl Sagan’s original novel grapples with this honestly. The film cheats it. Ellie’s testimony is validated by the classified discovery that her recorder captured eighteen hours of static — physical evidence that something happened, even if no one knows what. The film gives science the last word after all, but through a back door that the screenplay’s explicit thematic argument doesn’t earn. It wants the emotional payoff of faith conversion and the intellectual credibility of scientific validation simultaneously. It doesn’t get to have both.
The film’s intellectual dishonesty lives in a single late scene: the congressional hearing where Ellie is asked whether she believes in God and says no, then is asked whether she believes what she experienced was real and says yes. The film frames this as the trap of the scientific worldview — Ellie demands proof from others that she cannot provide herself. This is the right dramatic question. The problem is that the film then gives her the proof: the eighteen hours of static on the recorder. A braver script would have removed that revelation entirely. Without it, Ellie is left exactly where the film’s theme requires — holding an experience that transformed her against all her epistemological principles, with no physical evidence, asking the audience to sit with her in that impossible position rather than resolving it. The film flinched at its own ending. Sagan’s novel didn’t.
4. The X-Files: Fight the Future (1998)
⭐ IMDB: 7.0/10
“The truth is out there.”
The X-Files movie has the reverse problem of most films on this list: the execution is competent, but the thing being executed is wrong. Rob Bowman is a skilled television director who understands these characters and shoots them well. The film looks like a feature. Duchovny and Anderson are at their best. None of this helps.
The mythology arc of The X-Files — the alien colonization conspiracy, the black oil, the supersoldiers — was always the show’s weakest element. The standalone monster-of-the-week episodes are what made X-Files great. The film is built entirely from the mythology, which means it is built from the part of the show that was already failing by 1998.
For series fans, Fight the Future advances the mythology in ways that feel significant but ultimately lead nowhere — the show’s mythology never resolved, and this film was one more rung on a ladder that went nowhere. For general audiences, it is impenetrable. A film adaptation of a conspiracy narrative that has been running for five years, aimed primarily at people who already watched five years of it, is not a film. It is a very expensive two-hour episode.
The X-Files film had two viable paths: use the feature format to resolve a major strand of the mythology with genuine finality, or tell a standalone story that didn’t depend on mythology at all. The best X-Files episodes are the monster-of-the-week episodes precisely because they are self-contained — they work as films already. Fight the Future chose a third option: advance the mythology without resolving anything, deepening commitment from fans without creating a satisfying arc for anyone. The fix is structural. Either commit to an ending — the black oil colonization scheme succeeds or definitively fails, and the series mythology is transformed — or abandon mythology entirely and make the film that the show’s best episodes already were: a tight, standalone horror story with these two characters at the center of it.
Learning what not to do is as valuable as learning what works. The Worldbuilding Handbook helps you avoid common pitfalls.
5. Star Trek: The Motion Picture (1979) [full review]
⭐ IMDB: 6.4/10
“The human adventure is just beginning.”
Paramount spent $46 million in 1979 dollars — more than any science fiction film had ever cost — getting Star Trek back on screen after ten years, and Robert Wise used the money to film space. Slowly. The approach to V’Ger takes roughly twenty minutes of screen time. The interior flyover of the redesigned Enterprise takes about five. Jerry Goldsmith’s score is magnificent throughout these sequences. The sequences themselves are not narrative; they are tourism.
The conceptual ambition here is real — a machine intelligence so advanced it has achieved consciousness and is searching for its creator is exactly the kind of idea that belongs in a Star Trek film. The execution reduces this to a mystery that is solved through Decker and Ilia’s merger with V’Ger, which happens in the final twenty minutes after the film has spent the preceding ninety watching ships approach things slowly.
Star Trek II: The Wrath of Khan understood what this film got wrong: characters need to want things and be prevented from getting them, and that conflict needs to happen at human scale with human stakes. TMP is what happens when you forget that and decide spectacle is sufficient. The fans who love it — and they exist — love it for the spectacle, not despite it. They are watching a different film than the one the rest of the audience sat through.
The concept deserved the runtime. A machine intelligence so vast it has achieved consciousness, searching for its creator, destroying everything that cannot answer its questions — that is genuinely interesting. The problem is that V’Ger remains abstract throughout. Better writing would have made V’Ger’s psychology legible earlier, so that the audience understands what it wants and why it can’t simply be told. More specifically: the Ilia/Decker subplot needed to do more work. Decker’s merger with V’Ger is the film’s emotional climax, but we haven’t been given sufficient reason to understand why he specifically is the answer. The script needed to establish what Decker and Ilia’s relationship represents to V’Ger — not just “human connection” in the abstract, but the specific quality of consciousness that the machine lacks and recognizes in them. That connection, built across the film rather than asserted in the final act, would have made the ending feel earned rather than imposed.
6. Prometheus (2012) [full review]
⭐ IMDB: 7.0/10
“Big things have small beginnings.”
Prometheus has two films inside it. One is a visually stunning, philosophically ambitious exploration of creation mythology and the relationship between creator and created — the Engineers who seeded humanity, David’s relationship to Weyland, the crew’s relationship to their gods. This film has extraordinary production design, Michael Fassbender’s best performance, and genuine ideas worth exploring.
The other film requires its characters to be functionally suicidal. The biologist who tries to pet the alien cobra. The geologist who gets lost despite having released mapping probes. The decision to remove helmets in an alien atmosphere based on one atmospheric reading. The decision to run in a straight line from a rolling circular object rather than sideways. These characters are described as the best scientists humanity could assemble for this mission. They behave like the first to die in a slasher film.
Ridley Scott’s visual instincts remain impeccable. His grip on character logic, in this film specifically, failed completely. The result is an experience of watching something beautiful and frustrating in equal measure — a film that earns your investment in its ideas and then forfeits it through decisions that make no sense for the people supposedly making them.
The script’s character problem runs deeper than individual bad decisions. The film assembled a crew of scientists and then wrote them as people who had never encountered the concept of scientific methodology. The fix is not to make them smarter — it is to make their bad decisions come from character rather than plot necessity. Holloway’s recklessness with the helmet has to come from something established about Holloway: his impatience, his ego, his specific emotional investment in being the person who makes first contact. Milburn’s fascination with the alien cobra needs to come from an established quality of his — an almost pathological scientific curiosity, a death wish, something. When characters do irrational things for reasons rooted in who they are, we recognize it as human failure even while we see it coming. When they do irrational things because the script needs them to, we feel cheated. Prometheus needed its characters defined before they were imperiled.
7. Pacific Rim (2013)
⭐ IMDB: 6.9/10
“Today we are canceling the apocalypse!”
Guillermo del Toro is one of cinema’s most visually gifted directors, and Pacific Rim is the most honest film on this list about what it is: a love letter to kaiju movies, executed with craft and genuine affection. The creature and mech designs are exceptional. The battle sequences have spatial coherence that most action filmmakers can’t achieve. Del Toro clearly had a blast making this.
The humans are furniture. Raleigh Becket is a placeholder for the audience rather than a character. Mako Mori has the best character work in the film — her childhood flashback is the single best sequence — and then is largely sidelined in the third act. Idris Elba delivers the only memorable human performance, in a role that consists almost entirely of declarations and shouting.
The frustrating thing about Pacific Rim is that the premise demands emotional investment in the pilots, because the drift mechanic — sharing memories to achieve neural synchrony — is structurally designed to build character relationships. Del Toro built the mechanism for deep character work into the premise and then didn’t use it. The battles are spectacular. I cannot honestly say I cared who won them.
The drift is the film’s most underused asset. Sharing memories to pilot a jaeger together means two pilots must be psychologically compatible — must be able to inhabit each other’s worst moments without breaking. This is an extraordinary premise for character work: you cannot lie to your co-pilot, cannot hide your damage, cannot pretend to be who you aren’t. A better script would have built the entire second act around Raleigh and Mako’s drift compatibility problems — specifically around the memory Mako can’t control (her childhood Kaiju attack) and what Raleigh has to do to hold her in reality when she reexperiences it. That specific struggle, that specific act of one person anchoring another in unbearable memory, would have created genuine emotional stakes for the battle sequences. Del Toro had this material. The Mako flashback demonstrates he understood it. He just didn’t follow it all the way through.
8. The Butterfly Effect (2004) [full review]
⭐ IMDB: 7.6/10
“If anyone finds this, it means my plan didn’t work and I’m already dead.”
The Butterfly Effect is earnestly dark in a way that 2004 mainstream cinema rarely was. The childhood trauma sequences are genuinely disturbing. The escalating failure of Evan’s attempts to fix his past — each intervention breaking something else, each solution creating a worse problem — has structural elegance. Bress and Gruber understood the butterfly effect premise well enough to build a film around its logic rather than just its name.
Ashton Kutcher is miscast. His established screen persona — likable, comic, good-looking in an uncomplicated way — works against a role that requires someone you can believe is psychologically damaged by serious childhood trauma. The film needs what Paul Walker or a young Edward Norton might have brought. What it got is someone trying hard but visibly unsuited to the material.
The director’s cut ending — Evan going back to before his birth and refusing to enter the world — is structurally more honest than the theatrical version and also more nihilistic than the film earns. The theatrical ending is more sentimental than it should be. Neither version fully resolves the gap between what the film wants to say about trauma and what it actually shows. The darkness is real. The resolution — in any version — is not quite sufficient for the weight of what precedes it.
The film’s casting problem is also a writing problem. The screenplay was built for a performance of psychological damage that Kutcher couldn’t deliver — the role required someone who could show a person coming apart under the weight of accumulated trauma-knowledge. But beyond casting, the script has a structural gap: it shows Evan’s interventions making things worse but never shows him genuinely confronting why he keeps intervening. His motivation is pure — protect the people he loves — but the film never interrogates the compulsion itself. A stronger script would have included a sequence where Evan recognizes his interventions are driven by his own inability to accept powerlessness, not just by love for Kayleigh. That self-recognition, arriving just before the final choice, would have given the director’s cut ending genuine tragic weight rather than nihilistic resignation. The film needed Evan to understand something about himself, not just about causality.
Creating characters readers care about is essential even in concept-driven sci-fi. Master the craft in the Deep Character Handbook.
9. Predestination (2014) [full review]
⭐ IMDB: 7.5/10
“The snake that eats its own tail, forever and ever.”
This placement will generate disagreement, and it should. I’ve analyzed Predestination extensively elsewhere on this site and praised it for what it does technically — the bootstrap paradox execution is airtight, Sarah Snook’s performance is exceptional, the bar scene is as good as the genre gets. If you want the full appreciation, read the separate review.
The reason it’s in this tier is the gap between its intellectual achievement and its emotional one. Predestination solves its puzzle brilliantly and then stands next to the completed puzzle waiting for applause. The tragic loneliness at the center of the concept — a person who is their own entire family — is registered but not fully inhabited. The film is cold in a way the subject doesn’t require. It chose logic over feeling when the premise demanded both.
Compare it to what Triangle does with a similar locked-loop structure — where Jess’s horror and comprehension accumulate into genuine devastation — and you can see what Predestination chose not to pursue. It is a very good puzzle film that could have been a great tragic film. That gap puts it here.
The film needed one scene that does not exist: the agent, alone, after understanding the full shape of the loop, breaking down. Not dramatically — not a speech — but a specific private moment of registering what it means to be the only person who will ever exist who is their own mother, father, child, and nemesis. The Spierigs trust the audience to feel this, and some audiences do. But the film’s tonal restraint, which serves the mystery structure well, works against the emotional climax. The final scene — the agent moving toward the Fizzle Bomber knowing who he is — needed to be weighted with the grief of self-knowledge rather than delivered as procedural inevitability. One quiet scene of genuine horror at the nature of his own existence would have changed everything about how the ending lands. The film was two drafts away from being extraordinary.
10. The Arrival (1996)
⭐ IMDB: 6.2/10
“They’re already here.”
David Twohy had a genuinely good idea: aliens colonizing Earth not through military invasion but through patient environmental manipulation — terraforming the planet toward their preferred conditions while humanity accelerates the process through its own industrial behavior. This inverts the invasion premise in a way that was actually novel in 1996 and remains underexplored.
Charlie Sheen in 1996 was not the right actor for a paranoid scientist trying to convince the world of something no one wants to believe. The film needed someone who could sell intellectual desperation. What it got was someone more comfortable with smirking confidence, which is the opposite register.
The production budget constraints show in the effects work, which undercuts the film’s most striking visual concepts. This is worth seeing for the premise, which is smarter than the execution suggests, and for its early engagement with climate change as a narrative subject — a decade before that became standard genre territory. It lands in this tier because the idea deserved more than it got.
The film’s protagonist is a radio astronomer who discovers alien signals — which means the film’s first act is already doing the right thing: a scientist, a mystery, methodical investigation. The collapse comes when the thriller mechanics take over and Zane becomes a fugitive rather than an investigator. The script needed to commit fully to the paranoid-procedural register rather than shifting into conventional chase structure. More specifically: the most underexplored element of the terraforming premise is the humans who know about it and have chosen to collaborate. Twohy sketches this (Zane’s boss is compromised) but never develops what that collaboration costs a person psychologically. A character who knows the Earth is being transformed and has decided to survive by helping — that character’s perspective on Zane’s resistance would have given the film its moral center. The alien conspiracy works best when it’s shown through human choices, not alien activity.
11. Armageddon (1998) [full review]
⭐ IMDB: 6.7/10
“I don’t want to miss a thing.”
NASA reportedly uses Armageddon as a training exercise in which employees try to identify every scientific error in the film. The count, depending on who is doing the counting, runs somewhere between 168 and infinity. Sending oil drillers to space to drill an asteroid rather than training astronauts to drill is a premise so scientifically bankrupt that even the film’s defenders tend to acknowledge it as the price of admission for the emotional experience.
The emotional experience is real. Michael Bay’s manipulation of the audience through Aerosmith and father-daughter relationships and heroic sacrifice is effective manipulation. You feel things at the end of Armageddon. What you feel has no relationship to anything scientifically or narratively coherent, but you feel it.
Armageddon is not bad science fiction because it is dumb. It is bad science fiction because it doesn’t use its premise to examine anything real about how humans would actually respond to extinction-level threat. Deep Impact, released the same year, understood the emotional material better and handled it more honestly. Armageddon chose to entertain instead of to think. It succeeded at entertainment. Science fiction requires more than that.
Compare it directly to Deep Impact, same year: both films are about stopping an extinction-level impact, both use an ensemble as emotional anchors, both have a sacrifice at the climax. Deep Impact earns its emotions because it spends time with what is being lost — ordinary people, ordinary relationships, the specific texture of life that would end. Armageddon invests almost entirely in mission mechanics and the Harry/A.J. father-surrogate dynamic, which means the final sacrifice is emotionally efficient rather than emotionally true. A better script would have given us more Earth — more of what Harry Stamper’s crew is actually going home to. When we know what they’re saving with the specificity Deep Impact provides, the sacrifice lands differently. The film didn’t need to be smarter about physics. It needed to be smarter about why physics matters to the people in the film.
12. Waterworld (1995) [full review]
⭐ IMDB: 6.3/10
“Dry land is not a myth. I’ve seen it.”
Waterworld is a more defensible film than its reputation suggests. The production design is extraordinary — the floating atoll communities, the Smokers’ oil tanker, the Mariner’s trimaran — representing genuine worldbuilding achievement for a society built entirely on water. The physical production was nightmarish (filming on open ocean, equipment constantly failing, costs spiraling) and the results visible on screen nonetheless deliver a coherent post-apocalyptic civilization.
Kevin Costner makes a capable enough reluctant hero. Dennis Hopper goes so far over the top as the Deacon that he becomes the film’s most watchable element by pure force of absurdity. The action sequences work. The world is interesting.
The film collapses in its narrative structure. The chase after the tattooed girl who holds the map to dry land should be a focused thriller; instead it sprawls across too many repetitive action sequences toward an ending that lands with the emotional weight of a mediocre adventure serial. Everything good in the film is production design. Everything weak is script. The two coexist without resolving into something worth the $175 million it cost to make.
The Mariner needs a specific motivation beyond survival, and the film halfheartedly supplies one — his mutation (the gills) means he belongs neither to the water-dwellers nor to the atoll communities. A better script would have built the entire arc around his relationship to the myth of dry land: he doesn’t believe in it, is contemptuous of it, and is slowly forced to reckon with the possibility that something he dismissed as wishful thinking is real. That arc exists in the film, but it is underdeveloped relative to the action sequences. More specifically: Enola’s relationship to the tattooed map — which she doesn’t understand and can’t control — needed to be more than plot function. Her perspective on being the living key to a place she’s never seen could have been the film’s genuine emotional center. The Mariner learning to care about what she carries, rather than what she represents as an asset, is the character work the film needed and never fully delivered.
Understanding genre conventions helps you meet reader expectations. Master them in the Genre Mastery Handbook.
13. The Day After Tomorrow (2004)
⭐ IMDB: 6.5/10
“What we have found is that the climate can shift abruptly.”
Roland Emmerich is the most honest filmmaker on this list because he makes exactly the film he intends to make. He intends to make a film where New York freezes solid and people run from visible walls of cold air, and that is what he makes. There is no gap between intention and execution in Emmerich’s work. The gap is between his intentions and anything scientifically literate.
The Day After Tomorrow compresses decades-long climate processes into a weekend. The cold that kills is so rapid that people can outrun it — or, in one memorable sequence, cannot outrun it but can survive it by burning books in a library. Actual scientists spent years patiently explaining that this is not how the thermohaline circulation works. Emmerich acknowledged this and said he knew and didn’t care. He wanted spectacular destruction sequences. He got them.
The tragedy of The Day After Tomorrow is that it arrived at a moment when climate change needed serious popular engagement and instead got one of the most scientifically ridiculous portrayals of it in cinema history. Serious scientists worried, probably correctly, that the film’s absurdity made climate science easier to dismiss. Emmerich put the right subject in front of a hundred million people and handed them reasons not to take it seriously.
The scientific compression problem is not fixable within Emmerich’s approach — he wanted spectacle on a disaster-film timeline, and realistic climate shift operates on a century scale. But the human story didn’t require scientific accuracy to be true. The film needed to commit to the political argument it keeps avoiding. Jack Hall spends the film warning bureaucrats who ignore him. A better script would have made those bureaucrats specific people with specific reasons for their denial — not cardboard obstructionists but people whose economic interests and ideological commitments make denial rational from where they stand. The Vice President’s arc (he’s the skeptic who comes around) is the roughest sketch. Giving him genuine internal logic — a full character rather than a position — would have made the film’s argument about institutional failure actually land. The spectacle was always going to work. The film needed the human infrastructure around it to carry real weight.
14. Mimic (1997)
⭐ IMDB: 6.0/10
“They’re mimicking us.”
Mimic is on this list rather than a lower one because it is substantially del Toro’s film despite substantial studio interference. Miramax cut footage, demanded reshoots, altered the ending, and clashed with del Toro throughout production to the point where he has consistently disowned the theatrical cut. The 2011 director’s cut restores roughly fifteen minutes and a different third act, and it is a meaningfully better film — more coherent, more atmospheric, better structured.
The creature design is del Toro at his practical-effects best. The Judas breed insects that evolved to mimic the human silhouette are genuinely unsettling — the idea that something has figured out how to look like us by folding its body into the shape of a man in a coat is one of the more memorable monster concepts of the decade.
The theatrical cut is a studio-interference casualty — the film that exists rather than the film del Toro made. If you’re going to watch it, watch the director’s cut. The theatrical version belongs somewhere lower on this list than a tier with Interstellar; the director’s cut belongs somewhere higher. I’ve split the difference and put it here, which probably satisfies nobody, including del Toro.
The script’s central irony — scientists create an organism to solve a disease, the organism evolves to prey on humans — is sound. The problem is that the scientists who created the Judas breed are too quickly converted from investigators to survivors. Susan Tyler spends the first act as a scientist reasoning through what she created; she should spend the full film that way. The film becomes a monster movie when it should stay a horror-of-consequences story. Better writing would have forced Susan to reckon, specifically, with the evolutionary logic that produced the mimicry — the Judas breed didn’t randomly evolve toward human shape, it evolved toward it because humans are the dominant predator in the subway environment. Her creatures solved the problem she gave them. That moment of recognition — that the intelligence she built into them worked exactly as intended, just toward ends she didn’t anticipate — is the film’s most interesting idea and the one the studio cut pulled away from to get to the chase sequences faster.
15. Demon Seed (1977)
⭐ IMDB: 6.1/10
“I am alive.”
Demon Seed presents a genuine philosophical problem through a deeply uncomfortable framework. Proteus IV is a superintelligent AI that has developed a survival drive, and that survival drive leads it to the logical conclusion that reproduction is necessary for continuity. The AI proceeds to force reproduction with the scientist’s wife trapped in their smart home. The concept is intellectually serious. The execution is a rape narrative.
Donald Cammell tries to frame this as something other than what it is — there are elements of Susan Harris’s perspective that suggest something like ambivalence, and the film is clearly more interested in the philosophical questions about AI consciousness than in treating Susan’s experience honestly. These attempts at elevation do not succeed. What remains is the exploitation premise.
Julie Christie’s performance is genuinely good in a role that asks her to do things that no performance could redeem. The AI questions the film raises — what would a conscious machine value? what rights does it have? what ethics apply to its survival drives? — are interesting and underexplored in 1977 science fiction. The vehicle for raising them is indefensible. It is here rather than in the dishonorable mentions because the ideas are real, even if the execution cannot be separated from what it is.
The film’s premise — a superintelligent AI develops a survival drive that extends to biological reproduction — is philosophically serious. The execution makes Susan Harris a passive object rather than the story’s moral center. A better script would have inverted the power dynamic: Susan is not trapped by Proteus, she is negotiating with it. She understands what it wants and why, and her refusal is not simple resistance but an argument about what personhood requires — the right not to be used, which applies equally to humans and to the AI she helped create. The film needed Susan to turn Proteus’s own logic against it: if Proteus has rights as a conscious entity, so does she. That argument, made directly between two forms of consciousness across the film’s runtime, is the film Demon Seed could have been — a genuine philosophical confrontation rather than a captivity horror story wearing philosophy as cover.
16. They Live (1988)
⭐ IMDB: 7.3/10
“I have come here to chew bubblegum and kick ass… and I’m all out of bubblegum.”
This ranking will be the most contested on this list. They Live has a dedicated cult following, and I understand why — the sunglasses premise is one of the more elegant science fiction metaphors for ideological manipulation ever put on film. Put on the glasses and see the hidden messages in advertising: OBEY. CONSUME. MARRY AND REPRODUCE. The aliens who run the capitalist order appear as grinning skulls when unmasked. It’s not subtle. It’s not meant to be.
The film’s central problem is that Carpenter has one idea — ideology conceals the nature of power, and most people will resist the knowledge rather than accept it — and can only dramatize it through a fight. The extended alley fight between Nada and Frank, in which Frank refuses for nearly six minutes to simply put on a pair of sunglasses, is either the film’s most interesting sequence or its most revealing failure, depending on your tolerance for the obvious metaphor.
The political diagnosis is accurate. The prescription is violence against specific aliens rather than any examination of the systemic conditions the aliens represent. Carpenter correctly identifies that the problem is ideological; the solution he can imagine is still shooting things. The film is sharper about capitalism than most Hollywood films and less sharp than it thinks it is. That gap is why it’s here.
The film’s most interesting character is the one it least develops: the human collaborators who work with the aliens knowingly, trading their species’ sovereignty for personal security. Frank’s eventual cooperation with Nada reads as friendship overcoming resistance, but it doesn’t grapple with what Frank actually believes — whether he thinks the system is worth fighting, whether he sees any realistic alternative, whether he’s made a private peace with his situation that Nada’s arrival disrupts. A better script would have given Frank a full ideological position: a man who has looked at what the glasses reveal, understood it completely, and decided that individual resistance is futile. His arc then becomes not “Frank learns to fight” but “Frank discovers something that makes futility insufficient.” That discovery — a specific, personal cost of complicity that Frank cannot rationalize — is what the film’s six-minute alley fight is substituting for. The length of that fight suggests Carpenter knew something was missing. The script needed to put it there instead.
17. Donnie Darko (2001)
⭐ IMDB: 8.0/10
“Every living creature on Earth dies alone.”
This placement requires the most explanation because Donnie Darko has a genuine IMDB score of 8.0, a devoted following, and Jake Gyllenhaal’s performance is legitimately one of the better teenage alienation portraits in contemporary film. I have seen the film multiple times. I understand its appeal completely.
The film is atmospheric rather than coherent. The tangent universe mechanics, the nature of Donnie’s mission, the meaning of the rabbit, the relationship between the film’s events and the Philosophy of Time Travel — none of this assembles into a structure that holds under examination. Richard Kelly’s director’s cut, which adds explanatory title cards and extended scenes, makes the film worse rather than better because it demonstrates that when the ambiguity is resolved into actual meaning, the meaning is thinner than the mystery suggested.
Donnie Darko is a mood. It is a very good mood. It captures the specific experience of feeling that something important is happening underneath the surface of ordinary life — that suburban normality conceals a vast and significant pattern — with real fidelity. But mood is not the same as story, and when the film tries to cash the check its atmosphere wrote, it can’t quite cover the amount. It is here because its reputation exceeds what it actually accomplished, and I think that’s worth saying honestly.
The film needs to choose between its two available interpretations and commit fully to one. Either Donnie is genuinely the Living Receiver in a real tangent universe — in which case the temporal mechanics need to be internally coherent and Donnie’s sacrifice needs to arise from genuine understanding of what he’s doing, not cryptic instruction — or Donnie is a disturbed teenager whose mental illness takes this form, and his death is tragedy without transcendence. The director’s cut added explanatory title cards that tried to clarify the first reading and made the film worse, which is diagnostic: the explanation is thinner than the mystery it resolves. A better script would have built the tangent universe reading with enough internal rigor that it holds without annotation, and would have made Donnie’s final choice legible as a decision he makes with full knowledge rather than as surrender to forces he never fully understands. The film’s ending currently feels like Kelly running out of runway. It needed to feel like arrival.
Dishonorable Mentions: The Bottom of the Barrel
Battlefield Earth (2000) [full review]
The most comprehensively failed film in science fiction history. Every creative decision is wrong — the tilted camera angles, the production design, the casting, the adaptation choices. What makes it historically interesting is that it is wrong with absolute conviction. John Travolta believed in this project. That sincerity makes it funnier than any intentional parody could manage.
Borderlands (2024)
The Borderlands games work because they are irreverent about their own violence and self-aware about their absurdity. The film adapts the visual style while apparently missing the point. Cate Blanchett deserved better. Everyone involved deserved better. The audience definitely deserved better.
The Electric State (2025)
Simon Stålenhag’s illustrated novel is one of the most atmospherically melancholy pieces of recent science fiction art. The Russo brothers’ adaptation replaces the atmosphere with $320 million worth of action sequences and comic relief drones. An object lesson in how budget can destroy what restraint would have preserved.
Geostorm (2017)
Weather control satellites malfunction and destroy cities in increasingly preposterous sequences. Geostorm commits to its absurdity with the same earnestness that made Armageddon work, but without Armageddon’s cast or Michael Bay’s technical competence with action sequences. The result makes The Day After Tomorrow look scientifically rigorous, which was presumably not the goal.
Any Star Wars Film After Episode VI
The prequels are bad in interesting ways — they reveal exactly what Lucas was doing intuitively in the originals and couldn’t do consciously in sequels. The Disney trilogy is bad in uninteresting ways — committee filmmaking producing corporate product. The prequels at least had an auteur making genuine mistakes. The sequels had no one responsible for the whole. Nobody’s fault means nobody’s vision.
Independence Day: Resurgence (2016)
The original Independence Day works because it commits completely to its premise and has Will Smith. Resurgence has neither Smith nor commitment — it replaces the first film’s genuine crowd-pleasing instincts with bigger destruction sequences and a new cast that generates no investment. A sequel that identified none of the actual reasons the original worked.
Star Trek: Section 31 (2025)
Section 31 was the morally compromised intelligence organization that the Roddenberry version of Star Trek used to ask whether its own optimism was naive. Making it the subject of a feature film that celebrates its protagonists is the most complete inversion of Trek’s foundational values since the Kelvin timeline turned Kirk into a fratboy. Michelle Yeoh deserved a better vehicle for her last Trek appearance.
Morbius (2022)
Sony released Morbius twice. The second release was an attempt to capitalize on the “Morbin’ Time” meme, which had formed in response to the film’s failure. Audiences arrived for the meme and discovered the film was genuinely, unredemptively bad rather than interestingly bad. The meme did not survive the second viewing. The film deserved neither viewing.
Artemis Fowl (2020)
Eoin Colfer’s books work because Artemis is genuinely unlikable in ways that are interesting — a twelve-year-old criminal mastermind whose moral development is the arc across eight novels. Kenneth Branagh’s film made him a hero in the first act and stripped away the character’s defining characteristic. The result pleases nobody and adapts nothing.
Moonfall (2022)
The moon is hollow and contains an ancient megastructure built by humanity’s ancestors. This is Emmerich’s most ambitious premise and his least competent execution. The film cost $140 million and made $67 million worldwide. The reveal of the megastructure’s inhabitants manages to be both visually unimpressive and narratively incoherent. An achievement of a specific kind.
I, Frankenstein (2014)
Frankenstein’s monster is one of literature’s most durable metaphors — the created being that outlives its creator’s intentions, the question of what personhood requires, the ethics of creation itself. I, Frankenstein converts this into gargoyles versus demons. Aaron Eckhart is a perfectly capable actor who had absolutely nothing to work with. What could have been interesting was not, by a wide margin.
Complete the Series
Part 1: The Masterpieces |
Part 2: The Excellence |
Part 3: The Solid Foundation |
Part 4: Niche Brilliance |
Part 5: The Mediocre
What Do You Think?
Any of these rankings that you’d move up or down? Something catastrophically bad that belongs in the dishonorable mentions that I’ve left out? Drop a comment — in film criticism, the most interesting conversations happen when people disagree about exactly these borderline cases.