What Game Trailers Can Learn From Film History
How can we use the history of documentary filmmaking to make more authentic and influential game trailers? Put on your film student hat, because this post is all about Cinema Verité and Direct Cinema (links go to more in-depth explanation of these terms). These are two styles of documentary filmmaking which emerged around the 1960s because of portable 16mm cameras and sync sound audio recording which allowed filmmakers to get closer and more intimate with their subjects.
They look stylistically very similar and their goals were ostensibly the same: to tell more authentic stories. Direct Cinema is described as a "fly on the wall" style where the camera isn't acknowledged by the subjects and they appear to be going about their lives as "normal." The filmmakers are nowhere to be seen, and there's no voiceover narration in the finished product describing the events. The footage appears to speak for itself.
A good modern example of a direct cinema documentary is The Beginning: Making Star Wars: Episode I The Phantom Menace. In this documentary you won't hear the filmmakers talking, they don't sit down subjects for talking head interviews, and there's only the occasional moment where one of the subjects explains to the camera what is happening. It feels like you're in the moment observing things unfold. This seemingly invisible presence of the filmmaker makes it feel very real and unfiltered.
This scene could’ve easily had narration saying something like: “When George Lucas is given a board of intricately drawn storyboards, he quickly marks up each image with what needs to be made with CGI and what elements will be practical.” But presented as is, we can simply observe and figure it out for ourselves.
Cinema Verité is stylistically very similar with its handheld cameras and "raw" style of camera work, but in Cinema Verité the filmmakers are present, often visibly or audibly interacting with the subjects. The philosophy here is described as "fly-in-the-soup" style because of how the filmmakers insert themselves into the story. The idea here is no matter how unobtrusive, the camera and filmmakers are still a presence in the room, so by the filmmakers and subjects acknowledging the practical reality of being filmed, everything will appear more authentic. The opening scene of Scorsese's documentary Italianamerican is a fun example of this style.
But here's the thing, no matter how uninvolved the camera person or filmmakers appear to be in the footage, the filmmakers influence what shots get shown and how they're edited together. For example, a shot of a cat attacking a child might make you think the cat is vicious, and the child is an innocent victim. But it's a very different story if the WHOLE shot is the child pulling the cat's tail, and petting it in ways it doesn't like and chasing it around. So, as always, the director and editor still ultimately have control over what story is told.
As the saying goes, the films are made/written three times:
First when it's written
Second when it's shot
Third when it's edited.
While Werner Herzog’s documentaries rely heavily on his voiceover, his interviews are very much fly-in-the-soup in style. In this very emotionally raw scene from Grizzly Man, Werner Herzog (the documentary film’s director) is played the audio captured of the main character’s last moments as the camera person zooms in on the widow’s face.
I first started thinking of the connection between these filmmaking movements after releasing my video essay What Makes A24 Movie Trailers So Good? because in that essay I took a close look at the movie trailer for The Florida Project (made by Mark Woollen & Associates) which is notable because of its use of scenes which play out seemingly exactly as they do in the film, rather than being cut together with lots of b-roll or the dialogue used as voiceover to be juxtaposed with other moments from the film. Similar to Direct Cinema, the trailer editor chose some raw scenes and let you come to a conclusion for what it means and what story is emerging (but of course, the editor is invisibly leading you towards an idea by selecting these particular scenes)
The polar opposite to this approach is an in-your-face narrator or series of title cards which cut the footage into tiny bits and tell you exactly what you're seeing and why it's cool. In the Direct Cinema style you are invited to think for yourself based on some seemingly unfiltered footage, and in the other you're being told NOT to think because it's all being explained to you.
So how can you apply these ideas to gameplay trailers?
Well, there's a fine line between well captured and selected "raw" footage, and a bunch of clips that are then cut into a montage that seems to have little to no direction. This recent trailer for UnderDungeon is one which got me to thinking of this style of capture/editing.
In this trailer there seem to be only a small handful of edits. Otherwise, this gameplay feels very in-the-moment. This has the effect of making it look like we're seeing "raw" gameplay. It's fully possible there's some editing trickery here for timing purposes, but to my eyes it feels like it was done in several long takes. And yet, something in this footage makes it feel like it's still a trailer despite having none of the usual frills like added music. The information I need to understand this game is released a nice pace, and I feel like I'm getting a bit of a story.
The gameplay trailers released for Half-Life: Alyx also feel very much in the vein of Direct Cinema style gameplay. But this is still not "raw" gameplay in that I'm sure whomever captured these clips did several takes to pick the best one. I love these clips because they're very well paced. There's suspense, interesting gameplay ideas happening one after another, and lots of peaks and valleys in the action (for a more in-depth analysis you can read my review).
You might be wondering: Aren't you just saying to make our gameplay trailers continuous, unbroken "oner" shots?
Yes and no.
While an unbroken shot like the Half-Life: Alyx Gameplay trailers is one way to do this, I think using longer takes cut together is another way to make a game trailer feel more authentic than a bunch of short clips cut together at a very fast pace. I'm imagining a gameplay version of that trailer for The Florida Project.
For Firewatch's release we experimented with these mini-trailers which each depict one short and complete story moment from parts of the game. I could envision a trailer which stitches together some of these longer moments into one whole trailer.
What does this teach you about the game? It’s a first person narrative game set in the woods where you talk to someone about things you see and what you do in the game dynamically affects the things characters say, and it has funny light-hearted moments.
Because we're dealing with video games, there is another ethically debatable option which is using scripted sequences to create the appearance of "raw" gameplay. A great example is this Horizon Forbidden West Gameplay Reveal trailer. This appears to be "raw" gameplay because it's made with long continuous shots with no cutaways. But there are several times when the camera operator, and the characters controlled by artificial intelligence PRECISELY hit their marks. Everything is always very well composed and timed to perfection. This is a great showcase of gameplay and I think most people won't have the eyes to see any trickery, but the more scripting used, the less "raw" the footage.
If you don't know what scripting is, think of it like digital stagehands or crew mates who pull strings to make things happen at precisely the right time. For example, a game might be working on a system where when you drive a car into a gas station, the station explodes. But if the dynamic system that makes that happen doesn't work yet, you could instead drive the car into the gas station, then have a digital stage hand press a button to make it explode. It's the difference between an automatic door that opens because there's a motion sensor, and two people pulling ropes to open the door when they see a person approaching (or not, lol).
Because of these tools, it is within the power of game devs to create "live" gameplay takes which feel like they were captured during normal play. It is up to each developer to decide how much manipulation is "okay" for the sake of presenting something which feels authentic to the viewer. AAA Game Animation Director Jonathan Cooper has a great writeup for how to detect "Bullshots" in game trailers.
The PRECISION timing of the robot raptor on the far left of this shot is almost certainly the result of some very customized work either because it happens in the game this way or it was done for the trailer. The thought of trying to capture a succession of perfectly timed moments which rely on gameplay I in a completely live take is pretty much unthinkable to me.
This trailer for Bright Memory Infinite is another example which feels very unfiltered and it's plausible this is an unbroken take, but there are also many places within this shot where I can see the possibility for stitching multiple takes together because of things like the screen flashing completely white after certain weapon hits, or glitchy effects on the screen which can mask edits.
This approach might not be the right one for every game, it highly depends on what sort it is. For example, it would be pretty dry to watch long takes of a puzzle game, or realtime strategy game. It also relies on games which allow for a sequence of gameplay ideas to happen in relatively quick succession within one take. For example, if the things you want to interact with for an unbroken shot are very far apart, the amount of time spent walking from one thing to another could drag down the pacing.
Here are some rules for creating a game trailer that feels more like Direct Cinema:
Don't use voiceover or title cards
Use longer takes of gameplay
Tell a short story within each long take
Tell a story by cutting long takes together
I'd put this style of game trailer making in the more Advanced tier, because frankly, as a professional game trailer maker it just sounds like so much more work to both create long takes of gameplay and to find those moments in a game which can be used for good long takes with a beginning, middle, and end. Nevermind making multiple and putting them together to create a bigger narrative.
You can use this sort of idea and apply to to just one section of the trailer. For example, I often use this approach for the cold open like in the gameplay trailer I made for Spelunky 2 or Cybershadow which have unbroken shots of gameplay with several little ideas within the broader story of the entire shot. This also works great for Steam trailers which autoplay with no sound.
The opening shot of this trailer was captured by one of Yacht Club Games’ expert QA people.
Whether the audience is aware of it or not, cuts are the editor manipulating the footage, so the fewer cuts, the more authentic something will appear to be. This isn't an invitation to script the heck out of a gameplay trailer to make it look like "real gameplay"; you do that at your own risk. But it's some food for thought about the best ways to present your game to make a compelling game trailer!
If you want to read more about the subtle differences between Cinema Verité and Direct Cinema, this is a good writeup I found, and I highly recommend watching the documentaries Italianamerican and The Beginning for a deeper understanding of these styles of filmmaking!
