Can Social Media Chatter Predict Oscar Winners?

    by Nick Mendoza
    March 5, 2010

    The biggest night in movies is two days away, and everyone has an opinion as to who will win an Oscar. While there isn’t a proven formula that can tell us which film is going to win, a closer look at social media such as blogs and Twitter can provide some interesting perspective as to which nominees are dominating conversations and spurring emotional reactions.

    Here’s a look at the favorite contenders, as determined by social media chatter.

    Jeff Bridges, Sandra Bullock and James Cameron look like good bets to win."

    What the Blogs Are Saying

    Sysomos, a social media analytics firm, today unveiled an updated buzz chart for the 10 Best Picture nominees. The chart outlines which films captured the most attention and generated positive buzz — two potential indicators of Oscar destiny — on blogs over the past month. The blogosphere was measured based on share of voice (percentage of overall conversation) and sentiment (percentage of favorability).


    According to Sysomos’s findings, “Avatar” leads the conversation with 25.6 percent of blogger attention (share of voice), while “The Hurt Locker” was second with 18.1 percent. Based on this assessment, “Avatar” is the favorite to win Best Picture.

    I asked Sysomos to apply the same blog research to three other Oscar categories: Best Actor, Best Actress and Best Director. To keep the searches relevant, Sysomos narrowed the queries to include the name of the actor, actress or director and “oscar” or “oscars” and “academy awards.”

    According to the share of voice analysis, Jeff Bridges (“Crazy Heart”), Sandra Bullock (“The Blind Side”) and James Cameron (“Avatar”) look like good bets to win in their respective categories. The sentimental favorites are Colin Firth (“A Serious Man”), Carey Mulligan (“An Education”) and Lee Daniels (“Precious”).


    Best Actor

    i-e7f6e76d00eb35cdd58dca227678fb39-Actor Share of Voice.jpg

    Share of Voice Rankings:
    1) Jeff Bridges (25%)
    2) George Clooney (24.4%)
    3) Colin Firth (18.2%)

    i-9e0c046218ea2b1a62eca724f94d8eab-Best Actor (Sentiment).jpg

    Sentiment Rankings:
    1) Colin Firth (62%)
    2) George Clooney (58%)
    3) Morgan Freeman (57%)

    Best Actress

    i-f7bb3eafcc747f0e650f763da5f8e83f-Actress (share of voice).jpg

    Share of Voice Rankings:
    1) Sandra Bullock (28.2%)
    2) Carey Mulligan (22.2%)
    3) Meryl Streep (20.8%)

    i-25b81c5495debd8f4045699ecd612a58-Actress (sentiment).jpg

    Sentiment Rankings:
    1) Carey Mulligan (63%)
    2) Meryl Streep (57%)
    3) Gabourey Sidibe (57%)

    Best Director

    i-e1a7e7b77291263a51d8ce6e0ca84178-Best Director (Share of Voice).jpg

    Share of Voice Rankings:
    1) James Cameron (33.8%)
    2) Kathryn Bigelow (24.9%)
    3) Quentin Tarantino (16.3%)

    i-0595f95681eff66fcc920ac17c0993f3-Best Director (sentiment).jpg

    Sentiment Rankings:
    1) Lee Daniels (68%)
    2) Jason Reitman (62%)
    3) Quentin Tarantino (56%)

    Talk of the Town on Twitter

    In Hollywood, it’s not always good to be the talk of the town (see the controversial news that broke about “The Hurt Locker”). On Twitter, the talk is real-time and runs the gamut from great to good to bad to downright nasty. So while a high number of Twitter mentions might signal heightened interest in a nominee’s performance, it doesn’t necessarily mean they’re gathering support.

    The nominees who have the largest share of voice on blogs over the past month were also talked about the most on Twitter. Sandra Bullock has the most Twitter mentions (8,732), followed by James Cameron (6,176) and Jeff Bridges (5,785). In addition, the sentimental favorites on Twitter reflect the same emotions of the blogosphere, as tweets around Colin Firth (second), Carey Mulligan (second) and Lee Daniels (fourth) are highly positive, yet trail the category leaders in overall quantity.

    Twitter Rankings (past 30 days)

    Best Actor

    i-40b236fc544c2846ac24133bd86d67cd-Colin Firth tweet.jpg

    1) Jeff Bridges (5,785)
    2) Colin Firth (1,886) – positive sentiment leader with 61%
    3) George Clooney (1,706)
    4) Jeremy Renner (1,239)
    5) Morgan Freeman (733)

    Best Actress

    i-0d18040b084a54af48f4968f0cdf841c-Carey Mulligan Twitter.jpg

    1) Sandra Bullock (8,732)
    2) Carey Mulligan (3,839) – positive sentiment leader with 63%
    3) Meryl Streep (2,494)
    4) Helen Mirren (451)
    5) Gabourney Sidibe (22)

    Best Director

    i-c188edf7e00a73620f1804fc725bd6f7-Lee Daniels Twitter.jpg

    1) James Cameron (6,176)
    2) Kathryn Bigelow (2,982)
    3) Quentin Tarantino (1,100)
    4) Lee Daniels (971) – positive sentiment leader with 69%
    5) Jason Reitman (633)

    Will Social Media Predict the Winners?

    Now that we know the names and films dominating the discussion on blogs and Twitter, it’s simply a matter of sitting back and watching the show on Sunday. Then we’ll have a sense of whether our collective sentiment is also interesting science.

    Update March 9, 2010: Now that Oscar statues have been distributed, let’s revisit the pre-show social media buzz and see if the names called at the Kodak Theatre matched the names most frequently discussed on blogs and Twitter.

    The Accurate Buzz: Best Actor and Best Actress

    Jeff Bridges (“A Crazy Heart”) and Sandra Bullock (“The Blind Side”) led the share-of-voice category on blogs and were mentioned the most times on Twitter. Both walked away from the Academy Awards as first-time winners for their leading roles. The movie industry and the media that follow it all were bullish on Bullock and Bridges to win Oscars, and both won Screen Actors Guild (SAG) Awards for Outstanding Performance in a Leading Role.

    Bridges edged out Clooney in share of voice by only 0.6% in the past month, but Clooney arguably has dominant share-of-mind as an actor in the mainstream media. I suspect a longer historical comparison between the two would reveal a substantial lead for Clooney on blogs and Twitter. The heightened buzz on Bridges and his lead over Clooney signaled an abnormal interest in his performance and the potential to win an Oscar.

    The Inaccurate Insights: Best Picture and Best Director

    The biggest box office movie of all time naturally had the greatest share of voice on blogs. However, “Avatar” did not win Best Picture and James Cameron’s quest for a second win as Best Director fell short (he won for “Titanic” in 1997). “The Hurt Locker” won six Oscars, including Best Picture and Best Director for Kathryn Bigelow, while “Avatar” captured three Oscars (Art Direction, Cinematography and Visual Effects).

    To many critics and insiders, “The Hurt Locker” was expected to win due to its awards show momentum. The best predictor of Best Director continues to be the Director’s Guild of America (DGA) Awards, which gave Kathryn Bigelow its highest honor this year. As the DGA notes on its website (www.dga.org), “Only six times since the DGA Award’s inception in 1948 has the winner not gone on to receive the Academy Award for Best Director.”

    Lesson from Oscar Chatter

    This leaves us with our initial question: Can social media chatter predict Oscar winners? The answer is an unequivocal “not sure.” Pending further human analysis of the blog and Twitter mentions, we can’t support or refute the correlation between level of buzz and likelihood of awards. Unlike the near-certainty of a DGA Award recipient going on to receive an Oscar statue, the social media chart toppers largely reflect our pop culture and peer influences.


    Did you find yourself mentioning the favored Oscar winners (Jeff Bridges, Sandra Bullock and “The Hurt Locker”) on blogs and Twitter prior to the show? Do you think social media can help predict the winners? Share your thoughts in the comments.

    Nick Mendoza is the director of digital communications at Zeno Group. He advises consumer, entertainment and web companies on digital strategy, distribution and engagement. He dreamstreams and is the film correspondent for MediaShift. Follow him on Twitter @NickMendoza.

    Tagged: avatar blogs james cameron oscars precious sysomos twitter
    • Sysomos (and other vendors) state that the sentiment analysis is 80% accurate. The question is, how does 20% error margin work in comparing numbers within 10% range. As in the above examples? For example Colin Firth 62% vs. George Clooney 58%.

      More about this point-of-view in http://ow.ly/1eXm3

    • Jim

      Another original method of predicting the Oscar results saying Avatar just cannot win the Oscar – http://www.tomedes.com/2010-Oscar-winner.php

    • Mikko: The 20% error margin is definitely significant and human analysis is needed to see if Firth really did exceed Clooney in positive sentiment.

      It’s interesting to see that the share-of-voice leaders on blogs all won Oscars. The most talked about on Twitter, except for James Cameron, also picked up statues.

    • @mikko To put things in perspective, humans agree on sentiment 79% of the time, which means that no automated sentiment can be higher. Sysomos does combine machine sentiment with human, so their sentiment scores will tend to be pretty accurate. We measure with a 65% accuracy, and we also have 4 sentiments, which also makes accuracy harder.

      Overall, I would caution against using *only* sentiment for any analysis. There are other metrics like share of voice, velocity, share of impact, that are useful to look at.


      Maria Ogneva, Biz36

    • As the person who developed the sentiment classifier at Sysomos. I can assure you that our sentiment classifier is completely automated.

    • lets not forget one more aspect:

      Oscars are not political elections – we the people dont vote

      and yes, it does matter,

      try asking the public to predict the Nobel Prize

      olga lednichenko

  • Who We Are

    MediaShift is the premier destination for insight and analysis at the intersection of media and technology. The MediaShift network includes MediaShift, EducationShift, MetricShift and Idea Lab, as well as workshops and weekend hackathons, email newsletters, a weekly podcast and a series of DigitalEd online trainings.

    About MediaShift »
    Contact us »
    Sponsor MediaShift »

    Follow us on Social Media