This project started when I learned academia was applying serious scholarly rigor to sleazy 70s slashers. I’ve been fascinated by that collision ever since.

Academics argue horror gives audiences a place to process collective fears. When a fear runs deep enough, similar films cluster around the same moment. They call these horror cycles.

This project fed 70+ years of horror films and related data into a bunch of machine learning models to try to find these horror cycles.

Here are a few notable ones.

The 1940s gave us the atomic bomb. Hollywood responded with giant, irradiated creatures.

Nuclear anxiety made manifest in giant ants and city-stomping lizards.

The 1970s brought Vietnam into America’s living rooms and serial killers onto its front pages. Hollywood responded with a man in a mask and a kitchen knife.

They’re still arguing about why nobody ever just turns on the lights.

By the mid-1990s, horror had seen enough of itself to start laughing. Scream didn’t just revive the slasher — it autopsied it.

The postmodern cycle that followed was the genre interrogating its own history.

That’s the pitch. Seventy years and 56 inferred cycles.