Fan Fiction’s Predictive Value for Nielsen Ratings

September 25th, 2009 by Laura Leave a reply »

On January 15, 2009, CSI had one of its highest rated episodes all season.  On that day, people published 26 new pieces of fan fiction, the most stories posted on the same day as an episode had aired. On September 25, 2008, CSI had it third lowest ratings day all season and people posted zero new stories on that date.

Fan fiction is a really popular outlet for fan expression of interest in television shows.  The stories are creative, explore plot lines in the show and, according to many fans, help market a series in a positive way.  Fans often argue that their activities mirror larger interest in a show, and that producers should pay more attention to them and cater to their fannish interests as the example provided seems to demonstrate.  Fan Fiction’s Predictive Value for Nielsen Ratings tests this fan theory and answers the question: Does the volume of fan fiction published in the period around when an episode airs correlate to Nielsen Ratings?

To answer this question, fan fiction daily posting stats were gathered for the one week period around television shows where fan fiction communities existed and Nielsen Ratings were available for that show.  The fan fiction data was compiled from six archives: FanFiction.Net,, FanWorks.Org, FicWad, SkyHawke, and Freedom of Speech Fan Fiction.  The Nielsen Ratings data included over 720 episodes representing thirty-nine shows.  Once this data was compiled, it was analyzed using Pearson’s Correlation and linear regression. 

The results confirmed what many fans already suspected: Levels of fan activity, specifically in terms of the production of fan fiction, mirrors interest specific episodes of television.  Fan fiction can be used to predict Nielsen Ratings.  The predictive value is strengthened in several cases when it is broken down by network, genre or specific television show. The best networks for predicting Nielsen Ratings are CBS, The CW, Disney, Fox and USA. Comedy, crime comedy, crime drama, medical comedy and sports drama are the best genres for predicting Nielsen Ratings.  The strongest correlations for  television shows for predicting Nielsen Ratings are Burn Notice, CSI, Eli Stone, Friday Night Lights, Gossip Girl, Grey’s Anatomy, Hannah Montana, Heroes, iCarly, Law and Order: Criminal Intent, Life, Prison Break, Psych, and Terminator: Sarah Connor Chronicles.

This information is potentially valuable to parties with a vested interest in a television show’s performance.  By analyzing content patterns around periods with high volumes of fan fiction and high Nielsen Ratings, comparing that to periods of low posting volume and lower Nielsen Ratings, producers can make changes to maintain high interest amongst fans.  Non-American television networks and advertisers can better predict how their shows will perform.  This method of analysis can help organizations save money as it is cheaper to monitor and track than other analytic tools.

A copy of Fan Fiction’s Predictive Value for Nielsen Ratings can be found at . The appendix can be found at .

  • I knew that when we talked about this via IM and then read it that I was firm in the belief that there would be a link between some of the shows. I told you I saw in when InuYasha was going and now that InuYasha is around the corner for The Final Act, I am once again seeing an increase in participation for both fanfiction AND fanart amongst the InuYasha fandom. Obviously as you said, it is not all shows that produce the results that others do. This is something big. Great job Laura!

  • I suppose it's only natural to see correlations in this. Art is human expression, and the better the art is, the stronger response you'll generate across the board. Fanfiction is no exception, and you're right: it should be treated as valid feedback.

    If only it were possible to convince the lawsuit-happy types of this :/

    Good analysis, too - as someone who deals with complicated reports on a daily basis, I can only imagine how much fun it was ;)

  • There are some shows that don't have correlation when you get down to that sort of break down. The shows that didn't? They didn't surprise me.

    The lawsuit happy folks haven't really come after fan fiction in many years. They won't do so. Producers are being a bit careful about staying away from fan fiction sites or being only on sites where users, as part of the ToS, give away all rights to their ideas to the creators. The level of engagement feels pretty unprecedented.

    The analysis was pretty damned fun. :) It was the right up process that was the ZOMG! I HAVE NOT DONE THIS IN WIKI FORMAT IN SO LONG! that was scary. :) I love this sort of work and in ideal world, I'd love a job where I can do this sort of analysis for pay in the entertainment industry. Debating putting something together similar for sports and online communities because it just feels like so many people are talking about how you can do this type of engagement and get all these wonderful results... with out, you know, actually showing the ROI or presenting case studies. I wanted to do one of those. :)

  • Jack Newquist

    Excellent correlation!

blog comments powered by Disqus

Canonical URL by SEO No Duplicate WordPress Plugin