What if animation in games was not just eye-candy, but an integral part of gameplay?
Games are looking better and better. Along with the game worlds, the characters that populate them are becoming ever more detailed in their movement and appearance.
Yet despite this growing sophistication there is little change in the way animation is used in games. A walk cycle with a million polygons is still a walk cycle...
It seems more often than not animation is treated simply as content to put in the game, with little or no bearing on the interactive experience. It shouldn’t have to be like this.
What if animation in games was not just eye-candy, but an integral part of gameplay?
Far too many games rely on cutscenes to tell stories and icons and text to tell the player what's happening around him…
Why do we need menus to tell us what the characters should be telling us instead?
INTERACTIVE ANIMATION
Imagine a strategy game… Units start out as feeble new recruits and gradually turn into battle-hardened veterans as they gain experience. The usual stuff.
But what if you could actually see that experience in the movement of the characters?
New recruits run around nervously, trying desperately not to get killed, while more experienced warriors jog calmly through the carnage. In a hectic battle where split-second decisions need to be made at every turn, visual feedback like this is invaluable.
Ed Hooks explains in Acting for Animators (2003) that “scenes begin in the middle”. Whatever the scene is, something happened before it, and something is going to happen after it. Taking this into account makes the ‘middle’ that much richer.
You see a character running in a game… Where did he come from? Where is he going? What’s his goal? How the person runs will differ dramatically depending on the answers.
Take the following animations for example... The first suggests that the character is running to something, while the other seems to be running from something. A player might see this and conclude there must be something interesting in the direction the first is headed, whereas it’s probably wise to not go in the direction the other came from…
In RPGs the objective is usually to develop skills in combat so that you can take on increasingly dangerous foes. Early in the game, it often happens that you run into enemies that are too tough for you... and, later in the game, what was previously a worthy opponent now becomes a pushover. Unfortunately, as a player, it's very difficult to gauge whether an enemy is too powerful before that power has been experienced first hand.
Now... what if, depending on the relative power of an enemy, (on your approach) that enemy will either cower and look around frantically for a means of escape or it will take on a threatening stance and growl. I think it won't be difficult for a player to guess which is the more dangerous.
Another thing you often see is popularity stats. Your actions in the game affect the attitude of other characters towards you. Help them and they’ll like you, harm them and they won’t. You could show this in the way the characters greet you. One might smile broadly and hold out his arms as if to hug you, another might frown and cross his arms…
Of course, you don’t just want to show how NPCs feel about the player… you want them to show how they feel about each other. Most character-centric games involve at least some degree of social interaction. The player has to understand these relationships if he is to manipulate them…
Let’s say you walk in on a man and a woman parting ways. The woman is stuck in an idle cycle and the man is walking casually away. You’re not really getting any information out of this... But what if they hold hands, gazing intently into each other’s eyes… a final farewell before battle, like Hector and Andromachè… or what if she stands facing away from him, her arms crossed and an angry look of determination on her face, the man holding out his hand begging her to come with him.
Either way, you’re telling the player about these characters. You’re delivering the narrative, but more importantly, if the characters’ emotions and motivations have any bearing on gameplay at all, you’re giving the player crucial game information.
WALK THIS WAY
Games do not have the luxury of unique animation for every scene the way film does, but by cleverly utilising cycles, transitions and short actions, there’s not much game characters cannot express.
As shown in the examples above, all you need is different versions of the same basic building blocks. A player with full health walks energetically while an injured player limps and staggers and a tired character drags his feet…
A stealth game: Two guards stand outside separate entrances, each stuck in an idle cycle. The first is looking around scanning the environment for intruders… the second is trying hard not to fall asleep. Which entrance is the player going to pick?
How about if the player picks up something useful, he briefly goes into a happy walk?
A lot of this would only work in a third person game, as the player character would have to be on screen for you to learn anything from his animation… However, the first person view has the advantage that it allows NPCs to directly address the player. This leaves room for a lot more subtlety in their movement. You could have the same subtlety with NPCs in a third person game, but with a freely moving camera, chances are the player isn’t going to notice, unless the camera specifically focuses on those characters.
The question of clarity versus subtlety is a tricky one. Do you risk ruining the atmosphere with obvious over-acting, or do you keep it subtle in the hope that the player will notice anyway, if only subconsciously?
Probably, the solution is to have a bit of both.
PERSPECTIVE
If we’re going to figure out how to make characters move more tellingly, we need to look at body language and the psychology that drives it.
In his book Peoplewatching (2002) Desmond Morris talks about something called “Postural Echo”. Among friends, people seem to unconsciously match their postures to each other, creating a comfortable atmosphere of equality. When somebody adopts a noticeably different posture, it ruins the mood. This could be used to great effect in games…
If certain characters carry themselves similarly to the player, he will be more inclined to like them than he would those with more contrasting postures. You could factor this into the aforementioned popularity contest as well. As the ice melts, characters start adopting similar postures to you.
Another thing we could use is “Gaze Behaviour”; the length and frequency of eye-contact… Does the NPC meet the player’s gaze, or avoid it? And how long before eye-contact is broken? A subordinate does not hold the gaze of a superior. Instead he looks away, only briefly glancing at the other to check for reactions. A dominant individual could stare threateningly, or casually look away, ignoring his subordinate.
If the player looks at someone for too long it could well be perceived as a threat. If the player constantly looks away, an NPC might decide that he is bored… or perhaps shy.
Though probably not a feeling that games are likely to evoke, shyness would be an interesting emotion to roleplay.
Traditional animation too offers some interesting insights. If we go back to the “running to and from” example: The only real difference between the two is the so called “line of action”, the line that flows through spine, head and limbs to form a pose.
Running to something, the body leans forward, the head leading the way with intention and purpose… Running from the scary monster, the body leans back, the panicking head trying hard to keep up with the legs.
These are just some of the many fascinating behaviours that could be used to inform and immerse the player. From Peoplewatching to The Illusion of Life (1984), there’s a wealth of ideas out there just waiting to be used interactively.
The question then, in the end, is not how can we make animation look better -
it’s how can we use animation to make better games?
***
This is the unedited version of an article that appeared in Develop magazine in May 2005 (issue 50). You can also read it in PDF-form as it appeared in print. It should be noted that the above animation examples, while adequately illustrating the ideas expressed in this article, do not represent my current standard of work as they were made some time ago. If you have any thoughts or comments, let me know.
FURTHER READING etc.
HOOKS, E., 2003. Acting for Animators. Rev. ed.
ISBISTER, K. 2004. 10 Tricks from Psychology for Making Better Characters (GDC 2004)
JOHNSTONE, K., 1979. Impro, improvisation and the theatre.
MORRIS, D., 2002. Peoplewatching, the Desmond Morris guide to body language.
THOMAS, F. AND JOHNSTON, O., 1984. The Illusion of Life. Rev. ed.
Showreel - CV -
- portfolio -
© 2005 Christiaan Moleman. All modeling, texturing and animation by Christiaan Moleman.
A big thanks to my supervisor Simon McKeown for his excellent advice throughout the project. He helped me stay focused and challenged me to make the best project I could make.
Thanks as well to Ed Hooks, Adam Capone and Marque Pierre Sondergaard and all of the members of CGCHAR, the Adventure Gamers forum and CGTalk that were kind enough to offer their viewpoints on the project.