The Office: Analysis
By Tim Krock

When talking about NBC's the Office, one of the first things said is usually that "Steve Carell's performance made the show." or "Oh, I didn't watch it after Michael left." Using the transcripts and IMDB ratings of each episode, we can decide whether that's true.

This graph shows the expected impact of each character's appearance on the IMDB ratings of an episode of The Office.

Hover for Info

Notes:

In the regressions behind both visualizations, Michael had a positive impact on the ratings to a 99.999% confidence level, so we can conclude that the show did indeed get worse when Michael left.
Jim's effect in the first visualization is so stark because he's only absent in arount three episodes, and those episodes were poorly recieved.
Dwight speaks in every episode, so he was ommited from the regression.

This graph shows what we could expect to happen to the IMDB rating of the Office we gave another line to each character.

Hover Over/Tap heads for Info

Bonus: Shared Scenes Between Characters

This graph explores the interactions between the main characters of the show. Each chord represents the number of times each of the main characters spoke in scenes with each other character. Wider chords mean that the two connected characters shared more scenes.

Hover over the Chord Diagram

About the Data

The Office: AnalysisBy Tim Krock

Notes:

Bonus: Shared Scenes Between Characters

This graph explores the interactions between the main characters of the show. Each chord represents the number of times each of the main characters spoke in scenes with each other character. Wider chords mean that the two connected characters shared more scenes.

The Office: Analysis
By Tim Krock