Back to IndieWire

Can Statistics Be Used to Compute a Hit Film Script?

Can Statistics Be Used to Compute a Hit Film Script?

Online media streaming services like Netflix and Pandora have long been using algorithms to determine what customers are watching and listening to, and to calculate what content will succeed on their platforms. Netflix used this formula to bank $60 million on their original series “House of Cards.” (Read our TOH! story about how Netflix used this process to pick up exclusive streaming rights to “The Killing” here, and a article about Netflix’s data analysis here.)

Now, according to an article just published in The New York Times, Hollywood may start to employ such data analysis in as early stages as script development.

A chain-smoking former statistics professor named Vinny Bruzzese — “the reigning mad scientist of Hollywood,” in the words of one studio customer — has started to aggressively pitch a service he calls script evaluation. For as much as $20,000 per script, Mr. Bruzzese and a team of analysts compare the story structure and genre of a draft script with those of released movies, looking for clues to box-office success.

Screenplay writing has previously been regarded as an area in which creativity reigns supreme. But what will the moviemaking landscape look like if screenplays can be generated, right down to the microlevel of a scene or a plot device, using a calculator?

“Demons in horror movies can target people or be summoned,” Mr. Bruzzese said in a gravelly voice, by way of example. “If it’s a targeting demon, you are likely to have much higher opening-weekend sales than if it’s summoned. So get rid of that Ouija Board scene.”

Though Bruzzese insists that this process will be done by people, not machines, this could be a sorry day for the human screenwriter, “The Best Exotic Marigold Hotel” writer Ol Parker points out:

“This is my worst nightmare” said Ol Parker, a writer whose film credits include “The Best Exotic Marigold Hotel.” “It’s the enemy of creativity, nothing more than an attempt to mimic that which has worked before. It can only result in an increasingly bland homogenization, a pell-mell rush for the middle of the road.”

Read the full article over at NYT here.

Meanwhile, as we watch Hollywood crash and burn (as it collects oodles of cash), Danny Boyle frets that movies are succumbing to the influence of spinning toy movies like the “Star Wars” saga and animated family features from Pixar. And there’s always Steven Soderbergh, who articulates his own set of issues with the studio system. Our favorite Spielberg observation, from an old Premiere interview: Hollywood movies look like they’re directed from the window of a limousine.

This Article is related to: News and tagged ,


Vincent Bruzzese

To clear up any misconceptions about the New York Times article may have left you with;

The script assessment process is not an algorithm, nor is it a statistical analysis of the script. It is an identification of the genre based elements within the script as they relate to playability and marketability, using historical information collected through moviegoer responses over the past decade. The analysis can be granular (down to the scene level – although I am not sure where the article got the comment on "bowling scenes", that just isn't true) or more macro.

Unlike the current process which leaves the writer (and usually the producer) out of the loop, we meet directly with those who had the original vision for the film and stay within that context. In many ways, this process empowers the artist.

As it stands now, a script goes into coverage where it is judged without any standard for criteria or expertise. If it survives that filter, the script is then subject to the opinions of everyone who touches it (most of whom never speak to the writer), and then if it is lucky it is…..rewritten! From there, the few scripts that make it through to production are then changed on set, eventually screened for an audience (and changed again) and then marketed. I ask you, where is the art in that? Where is the writer?

This process is the only one that brings the writer in and simply advises what will help the script sell, or potentially improve audience reaction. The analysis brings up issues that will be noted later on and changed; thus, allowing the writer to change the film in their vision, rather than letting someone else, who never spoke to them, change them.

It is not cookie-cutter and it is not used to create bland homogeneous product. Yes, there is a fine line between art and commerce, but one can mitigate risk and maintain art.


This is an absolutely TERRIBLE idea. For all the reasons outlined above and in the article.


Not opposed to the idea of linguistic analysis and statistical profiling on film scripts. I feel it can add value to the process. The problem here is that these guys are ripping the studios off. An individual with intermediate level knowledge of "R", Mathematica or MatLab could run this analysis with 10 lines of code. Is the value added to justify the $20K fee the database of scripts they have built? No. Most of that information is freely available and can be pulled together easily.

If there really is a market for this service, Bruzzese will be undercut on price quickly by other vendors with more established relationships with the studios such as MarketCast.

Your email address will not be published. Required fields are marked *