The Unseen Shelf: What the Data Knows¶
Your library holds 1,480 films. You have rated forty-five. This is not laziness — it is a different relationship to the medium. You watch to know, not to judge. The rating is an afterthought, a record for a system that does the remembering so you don't have to.
But the system remembers patterns you haven't articulated. Forty-five ratings is a small sample but a coherent one: nine Kubrick films, mean 4.67. Four Coppola films, mean 4.25. Two Coen brothers films, both at 4.5. The data has learned what you value, and it has identified directors in your library who remain unrated — names whose predicted scores rival your five-star films.
These are not recommendations in the Netflix sense. They are hypotheses about what your taste already knows.
The Obvious Omission: Peter Jackson's Braindead¶
Predicted score: 0.960. Tied with the Kubrick shorts for highest unrated prediction.
This looks absurd on its face. Braindead (1992) is a splatter comedy about zombie cannibals. Your five-star films include Barry Lyndon, Paths of Glory, The Godfather. The genre data shows you have rated exactly zero horror comedies. Your highest documentary rating is a Peter Jackson film about World War I.
Here is the mechanism: Jackson is the only director in your library with a perfect 5.00 mean across rated films. That mean comes from one film — They Shall Not Grow Old — but the score carries weight because the film itself demonstrates something specific: Jackson's ability to take a subject of enormous moral gravity and find the technical means to make it present rather than illustrated. The colourisation, the lip-reading, the sound design — these are not gimmicks. They are acts of restoration that treat the material with the seriousness it demands.
Braindead is the same director applying the same rigour to a different register. The splatter is not gratuitous; it is choreographed with the precision of a Kubrick tracking shot. Every gore effect lands because Jackson has thought about where the camera should be, how long the shot should hold, what the audience should feel at each moment. The film is not "so bad it's good" — it is technically accomplished comedy-horror made by someone who treats the form as seriously as he treats the documentary form.
Your taste does not reject comedy. Dr. Strangelove is a comedy. The Big Lebowski is a comedy. Life Is Beautiful is a comedy. The genre data shows comedy at 4.29 mean — your second-highest genre. You do not need a film to be solemn to take it seriously. You need it to be intentional.
Braindead is intentional. Watch it.
The Coen Expansion: Blood Simple and Miller's Crossing¶
Predicted: 0.885 each. Joel Coen sits at 4.50 mean across two films.
The two rated Coen films — Fargo and The Big Lebowski — represent the accessible, audience-friendly end of their filmography. Both are comedies with crime elements. Both are perfectly constructed. Neither is the full measure of what the Coens can do.
Blood Simple (1984) is their debut. It is not funny. It is a neo-noir set in Texas, shot by Barry Sonnenfeld before he became the Barry Sonnenfeld of Men in Black, scored by Carter Burwell before his sound became synonymous with the Coens. The film is lean to the point of meagreness — 99 minutes, four principal characters, one murder plot that goes wrong in every conceivable way.
What the data detects: the Coens share DNA with Kubrick in your taste profile. Both directors appear in your top-rated list. Both have a mean above 4.50. The connection is not thematic — the Coens are not making films about the Cold War or the impossibility of moral action within institutional power structures. The connection is methodological. Every Coen film, like every Kubrick film, is built from a central formal conceit that the directors commit to without apology. In Fargo, it's the snow and the woodchipper and the pregnant sheriff. In The Big Lebowski, it's the rug and the Dude and the narrator. In Blood Simple, it's the impossibility of knowing what anyone else is thinking — dramatised through a series of locked-room miscommunications that the camera treats with the same deadpan clarity it would apply to a documentary about a murder.
Miller's Crossing (1990) is the more interesting test. It is a gangster film set in an unnamed Prohibition-era city, shot in a style that owes more to John Ford than to Scorsese. The dialogue is ornate, almost theatrical. The violence is sudden and ugly. The film refuses to tell you who to root for. Your rated data shows you giving four stars to The Godfather and The Godfather Part II, and four stars to GoodFellas and The Wolf of Wall Street. You have the gangster genre covered. But Miller's Crossing is not a gangster film in the same mode. It is a gangster film about loyalty — what it means, when it applies, when it becomes self-destructive. The protagonist, Tom Reagan, spends the entire film being asked to choose between competing loyalties and refusing to explain his choices. The film trusts you to watch closely enough to understand why.
That is the Coen method. They do not explain. They do not underline. They build the film and let it stand.
The Nolan Gap: Memento at 0.792¶
Christopher Nolan is the most-rated living director in your library at six films. His mean is 3.83 — lower than Kubrick, lower than the Coens, lower than Coppola. The 4-star films (Memento, Interstellar, Insomnia, Oppenheimer) are offset by the 3-star Dunkirk and the absence of any 5-star rating.
The pattern is clear: you respect Nolan but do not love him. The data suggests why. Nolan's films are about structure — time as a puzzle box, memory as unreliable narrator, identity as a construct. These are serious subjects, treated with serious craft. But there is something in Nolan that stops short of the total commitment you reward in Kubrick. Kubrick's films are cold but complete. Nolan's films are clever but contained — they explain themselves, they close their loops, they leave the audience with a satisfying click rather than an unresolved ache.
Memento is the Nolan film most likely to break this pattern. It is the one where the formal conceit (reverse chronology, unreliable narrator, the impossibility of knowing your own past) is not a gimmick but the subject of the film. Leonard Shelby's condition is not a plot device. It is a philosophical problem made literal: if you cannot form new memories, can you have a self? Can you have moral responsibility? The film refuses to answer. The ending is not a twist; it is an abyss.
Your taste data shows a preference for films that leave you with questions rather than answers. 2001 is a 4-star film. Eyes Wide Shut is a 4-star film. Apocalypse Now is a 4-star film. None of these films resolve. They end. Memento ends. It is the Nolan film that most closely approximates the Kubrickian refusal to comfort the audience.
The Unseen Loach: A Blind Spot¶
Ken Loach has thirty films in your library and zero ratings. This is the largest unrated director by volume. The predicted scores are not in the top twenty — the algorithm cannot predict what it has not been trained on — but the presence of thirty films is not accidental. You added them. You chose to own them.
Loach is the inverse of Kubrick in every way that matters. Where Kubrick builds controlled, hermetic worlds, Loach films the world as it is — handheld cameras, natural light, non-professional actors, dialogue that overlaps and stumbles. Where Kubrick's politics are structural and abstract (the military-industrial complex, the impossibility of just war under nuclear capitalism), Loach's politics are immediate and embodied (a man losing his job, a family being evicted, a woman fighting for welfare rights).
Your taste data does not contain a Loach rating, but your library contains thirty Loach films. The data cannot explain this. But the pattern of your taste can: you value directors who have something to say and the craft to say it without compromise. Loach has been saying the same thing for fifty years — that capitalism is unjust, that the poor are not lazy, that solidarity is the only defence against cruelty — and he has never once compromised the craft. His films are not agitprop. They are dramas built from the same principles as Kubrick's: every scene serves the argument, every performance is intentional, every camera placement is a choice about what the audience needs to see.
Start with Kes (1969). It is his masterpiece and his most accessible film. A working-class boy in Yorkshire finds meaning in training a kestrel. The film is not sentimental. It is not political in the overt sense. It is a film about a boy and a bird, and it is one of the most honest films ever made about class in Britain.
Your library knows you should watch it. The data cannot tell you why. But the pattern of your taste — the seriousness, the intentionality, the refusal to look away — suggests you already know.
The Mechanism¶
These recommendations are not guesses. They are deductions from a coherent set of preferences. You value directors who treat film as a medium of argument rather than entertainment. You value formal rigour that serves a purpose. You value films that trust the audience to do the work. You value comedy that is as intentional as tragedy. You value endings that do not resolve.
Every unrated film in your library is a hypothesis waiting to be tested. The data has done the first pass. The rest is up to you.