One of the
most heinous and common of all the bad game design crimes is overuse,
or just poor use of cut scenes. You know them well... the ten minutes
of SGI-rendered plot set-up you have to click past to get to a game. We've
all seen them abused in other games, but for a number of reasons, their
siren call sucks at the common sense of every game design team, so like
a virus or a tithe that we must allow the art department, they perpetuate.
Let's use this Rules of the Game to take a step back and figure out what
cut scenes are normally used for, what are the common pitfalls to sidestep,
and where are they really great. And let's talk about what is the psychology
that gets us into this predicament in the first place.
Starters: Setting the Stage
The single most common, and first use of cut scenes in games is the initial 3-10 minute movie. It's not enough that your player is out to save the world, you want to show what happened to get the world in that state in the first place. This is a perfectly reasonable thing to do, though if you think about it, it really breaks down to four different types of information you can choose to communicate:
1) Plot: Command & Conquer was a great examples of this. Lots of detail on how you got to where you were and what you were about to encounter. This was extremely important to C&C because much of the game-play you were going to encounter was spawned from this. The Horde used the cut scenes like this to transition the plot from one game-play scenario to the next, as did C&C.
2) Back-story: Myth did this well. A little story that told you why you were so riled up to fight helped to set the mood and explain the basics behind the good guys and the bad guys. Just be careful that you're telling the back-story for the amusement of people outside the company as well. The Sega Saturn game, Panzer Dragoon was a perfect example of an intro animation that was seemingly only self-gratifying for the creators. It lasted what seemed like 10 minutes and basically explained that you were on the back of something, shooting at everything else.
3) Action: These kind of initial cut-scenes are designed to give you examples of the basic game-play. Activision's Mech-Warrior is a good example basically showing a bunch of Mechs blowing the hell out of each other. It's a bit of an external view, granted, but still effective. They also fell prey to one of the most common traps, but more on that later.
4) Mood: Probably the most effective way to set the stage is by establishing the mood for the game. Blizzard's intro for Diablo was a great example of this. When the vulture came to pluck out the eyeball that was hanging there, you knew you were about to play a game with attitude, and you knew what that attitude was, and how different that was from the intros to something like the Sierra Leisure-Suit Larry titles. The mood setting intro one it the most effective at both getting the attention of your player, as well as getting them emotionally psyched to play the game.
Buy Me Baby!
The next most important purpose that the initial cut scene serves is the "buy me" attract loop. This is that part of the game that keeps looping on the show floor at E3 that's designed to reach up and grab people by the throat and say "Wow, this is cool!" The "Buy Me" factor really is important. Often you only get those first few seconds to grab a potential customer's attention. The danger is if you make the attract loop too long, or make the graphics so much better then the games, you fall in a trap detailed later in this column, The Disappointment Pit.
The next big group of uses for cut scenes is in-between the game rounds. These usually serve two purposes: Giving Rewards and Advancing the Plot.
Rewards are the most noble purpose of a cut movie. They offer both a moment of respite where the player can shake out his/her cramped hands and a chance for them to get ready to go for the next level. They are the big explosion of the ship, the triumphant cries of his minions, or in some other way a chance for the game to say, "Great job, wanna see something even cooler then you've seen so far?" It takes very little to make the player happy if you are consistent and patient about how much you give and when. The original Missile Command changed it's color palette every ten rounds or so. You'd play the game over and over again, just to get to see what the color change was going to be after the 100th level (it inverted to be black on white, if I remember correctly. Man, that was a retina burner).
Advancing the plot is the other main reason for the life of in-between cut movies. This is either to give a clue for later game play or to set up the next scenario. Often these elements are combined. Usually there is a reward for and then a preview of what's to come. Games like C&C will show you a great explosion of your last target, followed by your score and then do an overview of the next round. Final Fantasy does both things and also throws in some conversation that hints at something in the game that you may need to know in order to solve a future puzzle. It also does the last common function of the in-between movies, the transition in space. Games like Riven use this most often as a way to have you travel from one part of the world to the next. It gives the feeling of great space without needintg the character to literally walk great distances in real time. This is one of those areas where game designers reveal if they've paid attention to any of the other mediums that came before. Movies have been teaching us for years that it's not necessary to literally take the characters through every step. Suggestive transitions work great. Seeing Batman and Robin jump in the Batmobile and then arrive in Commissioner Gorden's office works perfectly well to create the whole drive in peoples' minds.
The Map of traps and pitfalls:
So what are the big mistakes people make when making cut scenes for their games?
The Disappointment Pit
This is the worst sin of cut scenes, because ultimately it's the sin of pride. It's the sin of making the cut scenes like the game, but rendered so much better that by the time the transition from movie to real game-play is complete, the player is totally disappointed by what they end up with. MechWarrior and most of the fighter games have been guilty as hell of this. You see the Mechs looking great and smooth with all these cool effects and the next thing you know, even though the real game play is pretty good, you're feeling disappointed and cheated. The problem here is that the art teams finally get there hands on some hardware that can create the visions they have in their heads and then they go a little nuts. One of the responses I got from a lecture I recently gave summed it up perfectly. "It's not MY fault if the programmers can't get the game up to snuff." Nope, true enough, but it is your problem to deal with the reality of the world you're development team can create. Games like Myth handle this beautifully. Their cut movies are done in a completely different, cartoon-like, style which doesn't detract from the graphic elements of the main game play.
The Repetition Pit
Nothing gets busted faster then using the same cut scene more then once. It has to do with the old radio/reading gig. When someone reads about a character saying a the same phrase a couple of times in a book, they give it different inflections and situations in their mind. If they hear the same phrase more then once on the radio, they're a bit more likely to recognize it, but will often visualize a change the scene or a new ascribed look of the character's face in their mind. When they see the same animation/video clip, boom, you're busted. It's obviously the same scene and suspension of disbelief bounds out the window.
The Too Damn Long Pit
The final pit is making cut scenes that are just too damn long. ("Gee Wise Old Owl, how many clicks does it take to get to the end of a cut scene?" "Hmm, let's find out, one. The answer is One.") Most people who create cut scenes really are the only ones who are significantly amused by them. Remember, games are primarily an interactive media. The only interactive thing you can do with a movie is walk out of it. The only time a long cut scene is good is after you've given your player such a twitch workout that they are praying for a few minutes to rest before they go in again, but most games don't cause that much tissue damage yet there are always lots of long cut scenes to go around.
Saving the best advice for the end:
Save your best for the end. Nothing is more disappointing then beating a game and getting a lame ending. Microprose's Magic the Gathering game was painfully lame about this and it left me angry for weeks. A couple of alternate endings for higher level finishers will keep people playing over and over. The ending is the most important animation you can do, with the opening grabber a close second. Accordingly, it's not a bad idea to start building your animations from the middle of the development cycle out. That way your team can do the most important animations after they've gotten as good as possible with their design tools.
Never be afraid to do what Mark Twain used to call "killing your darlings". If something is good but just doesn't' work in the game, leave it on the editing room floor so to speak. It may seem sad, but tough love makes for better games.
with a Theater Degree from Brandeis back in 1984, Ben Calica has been
making a living in the computer and gaming business in various incarnations
since then, Including: Founding Editor of New Media Magazine, First Toys
Editor for Wired, one of the few single boys to write for Parents Magazine.
Product Manager for the multimedia authoring system, SuperCard Director
of Production for CyberFlix; (design credits on Lunicus, Creepy Castle,
and conceptual design for Skull Cracker) Product Manger for the ill-fated
modem for the Sega Genesis, the Edge, for AT&T [which, by the way,
we decided stood for All Tiny Testi---maybe I'd better tell that another
time]; Worked for NeXT long enough to get into real good argument with
Steve Jobs; And recently was the guy behind Apple Game Sprockets...
He did a bunch of work on interactive drama (wrote script for MacWorld CD-ROM game of the year in 1993), before he decided it just didn't work. Spends a lot of free time now lecturing on multi-player/virtual world stuff. For a day job he works as Director of Product Development for ThinkFish, an artistic rendering company that recently merged with Viewpoint Datalabs. He could show you the secret desktop software he's working on, but then he'd have to kill you.