In this previous post I discussed the practical differences between making trailers for video games and movies. But what really separates trailers for games from movies is the necessity to show the player verbs.
If you're not familiar with the design lingo, "player verbs" are simply the actions a game character can perform, like the ability to: jump, walk, read, look, push, pull, shoot, move etc. A simpler way to think about it are classic point-and-click adventure games which have a grid of verbs used to perform actions.
Movie trailer shots are there to tell the story and look pretty, but even the most movie-like games eventually have to show what the player is going to DO when they play the game; games cannot exist without the input of the player. As much as I think people should take YouTube comments with a grain of salt, I can't help but notice the frequency with which people express frustration at game trailers that either show no gameplay or take too long to do it, so I worry about this all the time.
How do you show off the player verbs in a comprehensible way, and balance it with making the game look exciting? This varies heavily on the game genre, but the fundamental thing the audience needs to see in a shot is simply:
Change, cause and effect.
For games with a lot of action in them, this is pretty easy. If we see a shot of Nathan Drake shooting a gun at a red barrel which then explodes, we understand this is a game where you shoot things. Mario jumps on a goomba, we understand that we're going to do that. Even in games with very minimalist graphics, seeing something move or change in a way that requires human input will clue the audience into what they're doing in the game. For example: in a puzzle game like "The Room" we understand that a shot of a key turning itself means this is something we're going to do, because keys cannot turn themselves.
But movement alone doesn't show player agency. For example, we know elevator doors open and close without direct human input. Therefore, a shot of elevator doors would not be a good example of showing player input. But if the elevator doors were to stop closing part way through, then re-open, that would be a sign of human input.
Objects changing on screen is just one way to show player verbs, the other way is via shots that indicate there is a player at the controls. What does it take to show this? Surprisingly little, because humans aren't machines that move things at precise intervals. This is why good animation has motion that speeds and slows. As humans, we can't help but do things imprecisely; our eyes are typically very good at distinguishing between "real" and artificial movement. A car moving straight ahead won't look like someone is at the wheel, but it will look human as soon as it varies its speed, pumps the breaks or steers.
Movement is just about the bare minimum you can show as far as player verbs go; without good art, the trailer will quickly get boring. The only exception I can think of is the game QWOP where the simple act of walking is a hilarious and arduous task.
Even just something like the rate a gun fires will clue us into human input. A machine gun shooting in short bursts, or maybe the person pressing the buttons tries to do it in some sort of rhythm or beat; the audience will instantly know there's a human at the controls.
Shots with no verbs can be very pretty and cinematic in games where you can make the camera orbit, track, pan, tilt, fly and zoom all around, but there's only so much the audience can take before they're going to get bored. Think of the difference between admiring the craftsmanship of an intricate 3D rendered yo-yo, versus seeing the yo-yo being used in even the simplest of ways.
In general, game trailer shots with no player verbs have to be:
A: Intercut with story via text or voiceover
B: Used as a short break between clips of gameplay
C: EXCEPTIONALLY PRETTY
I think shots without player verb that don't fulfill the above criteria need to be earned with a healthy amount of gameplay, because giving the audience a bit of what they want establishes trust and expectation. Without that initial promise of showing gameplay, they have no reason to expect any in the rest of the trailer.
So to set this expectation, purely aesthetic shots need to be kept to a minimum, ESPECIALLY at the beginning of a trailer. If the game has astounding art like in the recent reveal trailer for GRIS, audiences are more likely to stick around to find out more. If the art doesn't command the attention of the audience, you're working on borrowed time.
Also, showing player verbs means nothing if the shot isn't CLEAR and READABLE. It's better for the audience to see one thing and remember it, than see two or three and forget them all. So it's better when player verbs happen one at a time. It can be in multiple shots or within the same shot, as long as it's fed in a digestible manner.
You might be thinking: Well can't I just use some title cards or voiceover to describe what you do in the game? The answer is: yes, but I think there's value in letting the audience come to their own conclusions. Another thing to consider is: when you make a trailer you have the power of editing at your fingertips. You might be inclined to get everything in one shot, but through editing you can show the cause in one shot, then emphasize effect in the next with a close up or a different angle; this is one way you can focus the attention on the important information.
To sum up, player verbs in trailers need to show: change, cause and effect. If done well, this will translate into: "If, then, and therefore." The therefore is what will really hook the audience if they make it that far; that indicates they're intrigued by the possibilities, and need to play the game to explore them.