[OC] Top-Talker – Fernseh-/Filmcharaktere, sortiert nach ihrem Dialoganteil

I was curious as to how TV/movie character speaking-time was distributed, so I built a tool to ingest datasets of subtitles and calculate the split. I also added „cumulative speaker-share plots“ across time, which help visualize how an episode/film’s speakers change throughout the content. For some titles, speaker data wasn’t available, so LLMs were used to infer the speaker with ~reasonable success (please feel free to flag any mistakes).

Some interesting findings:

* Michael, Dwight + Jim make up ~half of [The Office](https://top-talkers.jamiepinheiro.com/title/the-office?shared=true)
* [Friends](https://top-talkers.jamiepinheiro.com/title/friends?shared=true) has a pretty even split between the leads

Check it out here! [top-talkers.jamiepinheiro.com](http://top-talkers.jamiepinheiro.com)

(Here’s the [raw dataset](https://top-talkers.jamiepinheiro.com/?raw=1). Each title’s underlying data-source is linked on [it’s page](https://top-talkers.jamiepinheiro.com/title/the-office) in the footer)

https://preview.redd.it/bq2239l4nhdh1.jpeg?width=1206&format=pjpg&auto=webp&s=b93a3fac8a30d0bc3c3d13bf93995d833c33e591

lmao Wolf of wall street is basically all Jordan

Disappointed you did Friends and the office and himym, but not seinfeld.