Most recommendation systems optimize for one thing: what you already clicked. They watch your last ten choices and serve you the eleventh, slightly louder. The result is a feed that feels like a hall of mirrors — familiar, frictionless, and oddly exhausting.
CurationFig starts from a different premise. You rarely choose a film because it has a 94% score. You choose it because you want to feel a certain way tonight — calmer, sharper, less alone. Mood is the search term you can't type into a genre box.
Feeling is the real filter
When we say "mood-first," we mean the emotional temperature of a work is the primary key, not its release year, its studio, or its aggregate rating. A 1953 film like Tokyo Story and a 2023 game can sit in the same emotional drawer if they ask for the same kind of attention. That is why our lists are organized by feeling — patience, melancholy, cozy focus — rather than by franchise or format.
This is not relativism. A slow-burn crime series and a 90-minute comedy make different promises to your evening. Naming the promise is the curatorial act.
Why algorithms can't do this
An algorithm can learn that you watched a lot of sci-fi. It cannot tell whether you wanted wonder or dread, and it certainly cannot tell that last Tuesday you wanted wonder but this Tuesday you want comfort. Those are context signals no clickstream captures cleanly.
Our editorial layer is the part no API generates: the reason a specific work earns its place in a feeling, stated in plain language tied to that list's thesis. That reasoning is the product. When you read "Stalker is the list's most generous on-ramp: the long take feels like breathing, not endurance," you are reading a human decision anchored to the mood — not a sort order.
What to do with this
Start from the night you actually have, not the night a homepage assumes. If you want a quiet watch, our slow cinema list is built for patience over plot. If you want focus, the late-night lofi list is a backdrop, not a distraction.
The point is not to abandon discovery. It is to make discovery legible — to say out loud why a thing belongs, so you can disagree intelligently and build your own version.