Predicting Nielsen Ratings from Pilot Episodes Scripts: A Content Analytical Approach

Abstract

Textual and content data were extracted from the pilot episode scripts of 183 new, dramatic television series and used to predict the 18-49 demo ratings for the first five episodes of each series’ first season. As expected, the originality of a series’ premise, the track record of success of the its creator(s), and the cognitive complexity of its pilot episode script each explain a statistically significant proportion of the variance in the Nielsen ratings over the first five episodes.

Keywords

• pilot episode
• script
• screenplay
• television
• television series
• television ratings
• nielsen ratings
• statistical analysis
• panel data
• regression analysis
• content analysis
• network analysis
• semantic network analysis