StoryFit at SXSW: How gender and character affects profit & awards

StoryFit SXSW Opening Video from StoryFit on Vimeo.

We’re done with our first SXSW as the company of two presenters, and we’re gobsmacked by the amazing sessions, films, and people we met. On March 11, StoryFit CEO Monica Landers and PhD Data Scientist Grace Lin presented “Sex & Text: Breaking Down Movie Stereotypes w/AI” at SXSW as a part of the film track. 

Their well-made argument is simple: It’s important to invest in women in film…Landers note  ‘Sometimes changing the norm can make more money.’” — Austin Chronicle

StoryFit used AI to analyse the data of over 2,000 films, and over 25,000 roles. If AI can help us identify our subconscious stereotypes, and then prove there’s a business case for smashing them, that can only be a good thing,” — Campaign Live.

“It’s a valuable analysis for SXSW..” – Engadget

With #metoo and #timesup, we are in the middle of powerful changes. It’s sometimes even hard to have these conversations. But this is one of the benefits of data. It’s measured proof that we as writers, producers, directors, and studios are stereotyping people in movies.  If we’re going to improve things, then it’s time to outgrow stereotypes.

The first step in improving things is measuring them. Here are some highlights we measured:

    • Men and women are written in such a predictable way that our algorithms can identify the gender in movie scripts without knowing a character’s name or physical descriptions.
    • Average number of male/female characters per movie: 11/5
    • Roles:  70% were male 30% female; Some of the biggest news is really simple. The 70/30 split is actually bad for business. 
    • Almost all women are stereotyped as highly agreeable, like 85-90% agreeable, whereas men cover the full range.
      • What does agreeable look like?
        • Mrs. Weasley (100%) from Harry Potter and the Chamber of Secrets
        • Mrs. March (100%) from Little Women

Men are strongly linked with aspects of Dynamism /Reactiveness/ Emotional Range, also called Neuroticism. Men get high scores on this. Women get mostly low scores.

    • What does this look like?
      • High RangeDonny (93%) from The Big LebowskiMcClane (80.95%) from Die Hard 
      • Low: Einar  (7.85%) from The Danish GirlBella from Twilight (19.5%)

Didn’t make it to SXSW? We’ll be talking about these results and more at the Atlanta Film Festival on April 16!