I found this paper titled When To Sell Your Idea - Theory And Evidence From The Movie Industry, and I simply had to read it. You can too [PDF link].
I’m always interested in learning more about the ideas economy and pathways to success in open marketplaces, and I’m curious about how research is actually done on topics like this, so it was a pretty interesting read. It was an analysis of spec script and pitch sales in Hollywood, so who isn’t interested in learning how it works :)
TL;DR
Most specs get written either by rookie writers with no writing credits, or by extremely experienced writers with lots of credits.
Studios (and other buyers) are more likely to listen to an idea from a rookie if the idea is more developed (e.g. spec script over pitch), so there’s an incentive for rookie writers to try selling specs instead of pitches.
Experienced writers have studios interested in them for both specs and pitches. The ideas they are more confident in, they develop into spec scripts before selling, whereas the ones they aren’t, they pitch.
Agents matter a lot for rookie writers.
The measure of a successful writer is how many writing credits they have. Other factors don’t matter that much.
Strong IP and contract laws might actually hurt rookie writers than help them.
The Problem
We all know there’s a ‘paradox of disclosure’ when it comes to sharing ideas - it’s hard to know how much an idea is worth before you hear it, but once you have, there’s no incentive for you to pay for it. At one extreme, you have open source software, and at the other, you have the people who make you sign NDAs before every conversation.
Every idea goes through several stages, and at each stage, the person with the idea needs to decide whether to develop it further, or try to sell it. If you sell an idea early, you have zero sunk costs, but if you wait, you could get more value for it. And more developed work has better IP protections. Like, a completed manuscript is better protected legally than a book proposal, and a spec script has better protections than a pitch.
The people who buy ideas, like studios or publishers, don’t give every idea a hearing, because it’s expensive for them. It takes a lot of time and money for them to give everyone a fair hearing, and to evaluate each and every idea. And they worry about lawsuits that might come from accidentally or otherwise using those ideas in unauthorized ways.
How do these forces interact with each other? What effects do these forces have on the ecosystem and the world at large? When’s the right time to sell your idea? When’s the right time to buy an idea? Who are the right people to buy from? This paper tries to questions in that vein.
The Model
This was my first time really reading about Game Theoretic Models. It wasn’t as intimidating as all the physics nerds in college had me believe. As an inorganic chemistry nerd, I wonder now why the physics nerds were into game theory, which is a math concept, and I also wonder why these concepts didn’t really sink in for me in spite of multiple watchings of A Beautiful Mind.
The model is essentially a set of assumptions, which get represented as equations. Then you vary the values of different variables and see how things change, and try mapping that to a real-life scenario and wonder about how likely or unlikely that set of variables is.
Over here, the assumptions are:
A writer has an idea, which they can 1) sell now 2) develop further or 3) drop.
A studio observes certain features of the writer and estimates how good their idea might be, before they share it.
A writer has an idea or a property, which the studio can’t obtain without the writer selling it.
The idea costs the studio more the more the writer develops it.
The studio can decide whether or not to listen to the writer’s idea.
It costs the studio something to listen to the idea.
The studio has information about market demand for the idea, which the writer wants to know about as well.
I won’t go into the mathematical depictions of these assumptions, but as I understand it, each writer has a perceived value to the studios (based on experience, track record, etc), each idea has a value (e.g. box office performance), and each pitch session costs the studio something. So the equation they are modeling is something like this: Given a writer, an idea (which is either a pitch or a spec), and a studio, what’s the probability of the studio buying the idea? They vary the value of the writer’s renown, the greatness of the idea, and the expense of listening to pitches/reading specs for the studio, and see how this probability of buying ideas changes.
The Model’s Predictions
This model predicts that while not every writer gets to share their idea with a studio, the most and least experienced writers are the ones who will sell a lot of spec scripts. Once an inexperienced writer starts gaining experience, they sell less spec scripts, because the studios take them seriously enough to listen to their pitches.
But as experience level improves for a writer, they realize they can get more money from selling a spec rather than a pitch, so the ideas they feel have the highest potential for success, they write spec scripts for. The ideas that they don’t value as highly, they try to pitch before they develop.
Does The Data Match These Predictions?
So usually, you’d have to use anecdotal evidence to see if these predictions make intuitive sense. But in this case, they found some data they could actually use to evaluate this hypothesis.
There’s this website called Done Deal Pro, which lists details of spec and pitch sales, and they pulled up a couple of thousand sales of movie pitches and spec sales from 1995 to 2005. They then used IMDb to get information about the writers involved in these sales. Further, they used this site called The Numbers to get details of the box office performance of the movies that actually got released.
Some of the features they collected about each sale:
Writer data: How many writing credits does this writer have? How many years have they been writing professionally? Do they have credits on TV/directing/producing/acting?
Sale data: Is it a spec or a pitch? How much did they pay for it?
Movie data: Did the movie get released? Did it have big stars? How many? What was the budget? How much did it earn in the US? Internationally? How experienced was the director?
I suppose they fed this data into a moderately sized Excel sheet, and made some pivot tables and all that jazz.
Results From The Data:
Holy crap, just as they predicted, writers with 0 credits and writers with >4 writing credits were the ones with the most spec script sales. Though TBH I find this 4 credits business a bit arbitrary. But their graph basically starts off at a high number of spec script sales for those with 0 credits, goes down for 1, even lower for 2, and then starts picking back up.
Of the ones that actually became released movies, they found an interesting trend. If it was sold as a spec originally, the box office success of the movie went up sharply as the writer had more writing credits. With pitches though, the box office performance didn’t relate as strongly to the experience levels of the writer. Could this actually mean that experienced writers go on to spec better ideas, and pitch the ones whose quality they aren’t sure of? I think it could.
The likelihood that the movie actually gets released goes up as the writer has more writing credits. And the spec sells for more money as the writer is more experienced.
When an inexperienced writer with 0 movie writing credits was represented by a top talent agency, they were more likely to make a pitch sale. Other factors included if they were in a team with other writers, or if they had an acting credit, or had worked at a major TV network.
What if a writer writes a lot and is really good, but has very few writing credits, or if a writer isn’t great but has quite a few writing credits? The author of the paper thought about this. If that was the case, there probably shouldn’t be as strong a correlation between the success of the movie the writer wrote, and the number of writing credits they have, because the box office is an unforgiving mistress. BUT THERE IS! I guess that could mean we need to take the number and quality of writing credits a screenwriter has very seriously, and if we are screenwriters, focus on notching that number up, because that is directly related to success.
What does this mean for Intellectual Property and Contract laws?
So we usually think having stronger contract and IP laws would mean more innovation, and more free-flowing ideas, because writers and other creators can share their ideas without worrying about being ripped off, but if this model is to be believed, the opposite is true.
And I guess this model is believable, because they simulated some results based on some assumptions, and the real-world data seems to show that these assumptions are solid, and this model is robust.
Stronger IP and Contract protections means a writer is able to get more money for their ideas. But it also means that studios are more worried about facing lawsuits for IP infringement that might affect their reputation as well as the success of their other projects, which makes them more reluctant to listen to pitches from unproven writers. In an industry where it is hard to distinguish which ideas were ripped off from a rejected pitch and which ideas came from the writers working for the studio, listening to more pitches is just an invitation for more lawsuits. Further, strong contract law means a writer would win higher damages in a lawsuit, and this is a stronger incentive to sue.
The net effect is that the writers whose pitches were rejected feel encouraged to develop them into spec scripts, as better contract laws mean they get better money for a spec, but for the population that has more success with selling pitches, it’s bad news as the studios listen to less pitches overall.
The Role Of Agents
In an industry plagued with information asymmetry, the agents are who oil the wheels to keep it running smoothly.
Studios feel unsure of unproven writers, but when they see that a rookie writer is signed with a big agent, they find it easier to assume she is a good writer.
They have connections with the studios, which gives them an idea of what value an idea might hold, and are able to communicate this to the writer, and are also able to use this information to negotiate better deals.
Agents discourage their writers from filing frivolous lawsuits, because it reflects badly on them.
Since they have an idea of what studios like, they give a writer the feedback they would usually have to pitch to a studio to obtain.
So we can see that rookie writers are the ones who benefit the most from agents and agencies. Experienced writers need agents only to be able to negotiate better deals, and if anything, they may be a thorn in their side, because they discourage pitches they think are unsellable and make it harder for them to make a profit off of pitching.
This lines up with the real world data - during the fight between the WGA and Talent Agents, the experienced writers with jobs managed to find work okay even after they fired their agents. The writers with no credits, however, struggled to break in to the industry. If anything, this whole lawsuit was based on ‘packaging’, where the big talent agencies would bundle experienced talent with inexperienced talent, and sell them as a package deal to the studios. Which obviously is bad for the studios and experienced talent, but is great for inexperienced talent.
The one thing to watch out for is a spec from a big writer who works with a big agency, because the agency presses the writer to develop spec scripts for even their bad ideas, just to get a bigger cut of the sale. The moderation of quality that naturally occurs by writers writing spec scripts only for their best ideas is disrupted by agents.
Applying These Findings IRL
Recipe For Success
If you’re an unproven, rookie writer with no credits, you need to signal your quality with some other signs, like
Teaming up with more established writers.
Working with top agents.
Developing your ideas more before selling them.
It ties in to this Twitter thread by Chris McQuarrie, where he says you shouldn’t think about selling your script, but about actually making them into movies. I find most writers and screenwriters agreeing with him on this, but all I see is the competition is so high now that you need to take your ideas to the next level of development just to have a chance.
In machine learning, I see the standards going much the same way. Where a decade or so ago, all you needed to be admitted into a machine learning grad program was some enthusiasm and decent grades, now you need a publication record, a high enough Kaggle score, a portfolio of independent projects, and a host of MOOC certificates in machine learning and data science. Granted, the ML research world is much less zero-sum and much more egalitarian than Hollywood, but it proves to me that while barriers may shift, the amount of effort and planning required to break into an industry only goes up with time.
It also seems very important to just get more and more credits writing in the industry, because that seems to be the only way to get better at it. I can see how writing scripts by yourself is nothing like the real thing, because the trial by fire of the brutal feedback process and focus on sellability makes you learn a whole bunch of other skills that are hard to come by when you focus on the craft of writing alone. As a programmer, I equate that to the steep learning curve I had going from grad school to writing code that couldn’t afford to fail in production.
With regard to signaling your quality, I see this play out big time in the startup scene. Lots of kids in colleges all around the US create innovative businesses, with varying levels of success. But the ones at Stanford receive a disproportionate amount of press coverage, mentorship, investor funding, and other resources to ensure their success. Their ideas are not inherently better than those of others their age; this disparity is heavily due to the brandname they are associated with. I see how this compounds throughout a career, and it’s nothing if not extremely depressing to watch.
What This Means For Diversity
When I read the Chris McQuarrie Twitter thread, the first thing that occurred to me was that the movies that would catch the eye of the big Hollywood producers wouldn’t be some dinky film made by two teenagers from Saskatoon, but probably more like Tiny Furniture by Lena Dunham.
While this model only looks at the survivors and has its target audience be those that didn’t choose to drop their ideas, developing ideas is not free of cost, and people drop ideas that seem unviable all the time.
My original idea was to moonlight as a screenwriter. But as an immigrant on a work visa, I had to prioritize my programming career. When that took me to a city without a vibrant writing/filmmaking scene, I didn’t anymore have a network I could tap into to collaborate with on screenplays, or make short films with. I switched to writing fiction as a result, as it was more conducive to solo work, where I could develop my ideas to completion on my own schedule. I imagine mine is not a unique story.
That said, given the wide proliferation of YouTube and TikTok, I wonder if making movies is second nature to children growing up now, and this is a much smaller barrier than I assume it is. Heck, someone was telling me their toddler says “Like, Share and Subscribe to my channel” on random home videos they make that aren’t even going to go on Youtube.
While that’s a hopeful thought, it strikes me it probably takes more and more resources to be able to develop an idea that the industry notices you by, and that kind of time, expertise, finesse, exposure, connections, and training is probably going to be more easily available to those whose family is already in or adjacent to show business.
Oh well, don’t listen to me, go ready your pitches, write your spec scripts and shoot your movies. And while you’re at it, team up with an experienced writer, and get an agent. Go rack up those writing credits and rake in the greenbacks.