The Office US (2005–2013)

The office was one of the best sitcoms of the 20th century — certainly on my infinite repeat list. The mundane routine of a mid sized paper company documentary truly makes us empathize about our work lives. Here’s my analysis on the public data available on the Office!!!

Word Cloud of all User review titles (Number of reviews 1,214 Reviews)

The Office began in 2005 and shot its last season in 2013 — Every time I recommend the office to someone I know; there were one of two outcomes-either they sustained through first few episodes and got hooked on the sitcom or they assumed that the Office is one more superficial sitcom to exist.

Metrics by Season:

From the above charts its evident that the Viewership by Season went on a downward trend after Season 5; To investigate the reason for the downward trend I created metrics by Episodes.

Metrics by Episodes:

The overall best episode by Rating, Viewership and Votes is “Stress Relief” from Season 5 — yes the episode where Dwight creates a Fire drill by smoking a cigarette in the office; “Today smoking is going to save lives”

User Reviews of the Office by year has surely seen exponential growth — from 5.62 Average user review in 2005 to 8.72 Average user review in 2021; the Office saw a jaw dropping 55.16% increase.

Until recently I had no idea that “Michael Slur” who plays Mose in the series is writer on “The Office” and also went on to co-create “Parks and Recreation” — From the Analysis Mindy Kaling(“Kelly”), Ryan Howard(“B.J Novak”) and Paul Lieberstein(“Toby”) are the top 3 writers by number of episodes written.

Popularity Contests are always Fun!!! To find the most popular Character in the Office, I had to web scrape all user reviews from IMBD — Formulate the count of character mention by using str_extract function. Here’s how’s the result looks like - Not much to surprise Michael, Jim, Dwight and Pam are the top 15% of the list.

Analysis Tool used: R Programming.

Visualization Library: ggplot2, wordcloud2.

User Reviews: Scrapped by using Selenium and docker.

Dataset: Office_data

Data Scientist | Master of Engineering Management at Purdue Graduate School