My perspective on video games has changed a bit. I used to care primarily about interesting game mechanics, fascinating challenges, and compelling stories. I still care about those things, but nowadays I tend to think of them as means to an end rather than as ends in and of themselves. Nowadays, I care most about the feelings, moods, and emotions that video games evoke, with a particular emphasis on wonder, awe, and mystery. I’ve invented a name for this perspective: I’m calling it evocativism.
I’ve appropriated the word “evocative” in the same way that expressionism appropriates the word “expressive”. All painting is expressive, so it seems presumptuous to label one particular style expressionist. Similarly, all video games—even Tetris—evoke emotions. However, games like Myst, Another World, Silent Hill, Ico, Shadow of the Colossus, Knytt, Cloud, flOw, Flower, Portal, Braid, and Osmos seem to put a primary emphasis on evoking certain kinds of feelings, moods, and emotions, and that distinctive focus is what I’m referring to as “evocativist”.
One of these days, before I starve to death, I’d like to try my hand at designing and implementing one of these evocativist games. My approach would exhibit:
- An emphasis on exploration and discovery.
- An emphasis on evoking a sense of wonder, awe, and mystery.
- An emphasis on natural environments and natural ambient sound effects.
- An emphasis on minimalism and non-verbalism.
- An emphasis on “artificial life” over “artificial intelligence”.
- A de-emphasis on pre-designed narrative.
- A de-emphasis on violence.
- A de-emphasis on frustration.
- An emphasis on procedurally generated content.
None of these characteristics is essential to my definition of “evocativist”. These just happen to be the characteristics I’m interested in focusing on. Each deserves multiple blog-posts, but here are a few introductory thoughts, just to get the ball rolling.
Exploration and Discovery
This is kind of a no-brainer. Computer games are better than any other artistic medium at creating experiences of exploration and discovery, and, arguably, they’re better at it than they are at creating most of the other kinds of experiences that games try to create. Although the most obvious type of exploration involves moving through a (virtual) physical space, you can also “explore” a game’s mechanics and their dynamic consequences. Portal involves the physical exploration of the Aperture Science lab, but it also involves an exploration of how the portal guns work, and what you can do with them. Even ignoring the “story”, the character of GlaDOS, and the subtext of escape, exploring the dynamics of the portals can evoke a sense of wonder.
Wonder, Awe, and Mystery
There’s a prejudice that states that games are not good at evoking emotion in general. However, games about exploration and discovery almost always evoke a sense of wonder for me, at least to some degree. Games are good at evoking a sense of wonder. In addition, I think they’re good at evoking a sense of loneliness, isolation, and desolation. And recent games like Flower show that they can be great at evoking a sense of exuberance and joy.
Video games have traditionally been bad at evoking complex social emotions. Paintings and symphonies are also bad at this, so our medium is in good company. Many designers are interested in cracking this problem. Although I’m interested in seeing what others come up with, I’m not interested (at the moment) in working on the problem myself. I’m willing to let novels and movies deal with complex social emotions, since that’s what they’re good at. If a computer game can evoke anything like the emotion I feel when looking at the ocean, or a tide-pool full of sea anemones, or the Grand Canyon, or the night sky, that fully justifies it as an artistic medium. (See Flower and Osmos.)
Nature
Wikipedia says: “Characteristics of Impressionist paintings include visible brush strokes, open composition, emphasis on light in its changing qualities (often accentuating the effects of the passage of time), ordinary subject matter, the inclusion of movement as a crucial element of human perception and experience, and unusual visual angles.”
“Ordinary subject matter” is not a logically necessary item on this list. There’s no technical reason why Monet couldn’t have used “visible brush strokes”, “open composition”, and “emphasis on light in its changing qualities” to paint pictures of rotting corpses. But practically speaking, artists who are inclined to use those techniques are probably also going to be inclined to paint scenes of natural beauty. Similarly, game designers with evocativist leanings are probably going to be inclined to focus on beautiful natural environments (Shadow of the Colossus, Flower), or game mechanics that exhibit nature-like dynamics (flOw, Osmos).
I don’t really need to elevate this into a matter of principle. I haven’t played Left 4 Dead, but it seems to exhibit some evocativist tendencies. The movie Blade Runner presents an evocative non-natural environment (although there the absence of nature is part of the point). I’m happy to chalk up my interest in natural beauty as a personal (and maybe even temporary) artistic preference.
Minimalism and Non-Verbalism
More artistic preference here. There’s no technical reason why an evocativist work couldn’t be wordy and baroque. But in practice, aside from completely text-based works of interactive fiction, virtually all of the evocative experiences I’ve had while playing video games have happened while the game was staying the hell out of my way and letting me wordlessly experience something. I enjoy silent protagonists, and many of my favorite games contain companions that either don’t speak the player’s language, or don’t speak at all. I hate explanations. I’m a big fan of silence in general—aurally, narratively, and expositionally. And when it comes to game mechanics, I’m fanatical about simplicity and elegance.
Artificial Life
This idea goes hand-in-hand with the previous one. I find it significantly easier to connect emotionally to characters in video games who mostly don’t speak but act in procedural and life-like ways, than to characters who do speak (whether or not they also act in life-like ways). Somehow, attempting to simulate dialog just shatters my suspension of disbelief. I think the technology just isn’t there yet. I view Yorda in Ico, and all of the beasts in Shadow of the Colossus (including the horse), as instances of artificial life rather than artificial intelligence, but maybe that’s a meaningless distinction in the game-development world.
No Pre-Designed Narrative
This is a tricky one. Years ago, when I loved text adventures and LucasArts games, I thought I loved story. Now I almost consider myself to be anti-story, although that doesn’t really capture the truth, either. I do know that in every one of the “evocativist games” I listed above (except probably Portal), I find the “plot” or “story” (if it exists) to be the weakest element of the game—at best irrelevant to my emotional experience, and at worst an eye-rolling distraction.
I do like the idea of creating a game environment in which events occur that players later spin into narratives. I don’t like the idea of drama-managers. I do like the idea of a game that provides a succession of relatively short (and probably procedurally generated) play experiences with beginning, middles, and ends, rather than one long slog through a pre-designed campaign. I like the idea of providing an overarching narrative context for such an experience, but I don’t like the idea of providing a (pre-designed) narrative spine along with lots and lots of (also pre-designed) side-quests.
I have a lot more thinking to do about this one.
No Violence
It’s hard to talk about this one without sounding self-righteous, so let me first say that I have no moral problems with violence in video games. Instead, I have two non-moral problems with it. First of all, it’s boring. It’s been done to death (no pun intended). The world doesn’t need yet another person (me) making yet another game about killing stuff. But secondly, and more importantly, killing stuff in video games evokes certain feelings. These feelings aren’t “wrong”. Maybe they’re even cathartic. But they’re also in conflict with the kinds of feelings that I want to evoke in my own hypothetical game (like joy, wonder, etc.), so they actively work against my primary purpose.
Once again, this isn’t a matter of principle. Shadow of the Colossus is one of my favorite games, and it’s pretty violent. The game evokes a sense of desolation, isolation, and despair in me, and that’s partially due to the game’s violent gameplay. I’m totally ok with that. But those aren’t the kinds of emotions I’m interested in evoking. If it turns out that the best way to evoke some particular emotion that I’m going for is to allow (or force) the player to kill something, I’ll do it. But I’m not interested in filling my world with things to shoot, stab, or head-stomp, just because I can’t think of anything else engaging for the players to do while they’re exploring.
No Frustration
The ubiquitous fail/redo paradigm is my least favorite aspect of video games. I can’t stand making two successful jumps, failing to make the third jump, and then redoing the first two jumps in order to try the third one again. I’ll admit, it’s hard to see how to avoid this in a game like Ico. The roguelikes represent a theoretically sound alternative, but permadeath is frustrating in a different way. My general take is to start with the Rogue model of playing in a procedurally generated world in which all player’s actions are irrevocable, but make it so that the gameplay doesn’t revolve around killing, and the player can never die (thus eliminating the frustration of permadeath). Of course, this leads to the central game-design challenge of coming up with something interesting for players to do in such a game.
Procedurally Generated Content
I’ve sort of talked about this one throughout, but to recap: I’d like to create games which procedurally generate worlds to explore and play in, and which provide relatively short (say, 1-2 hour) play experiences with beginning, middles, and ends, but which involve styles of gameplay which are different than those found in the typical Roguelike or side-scrolling platformer. It would require multiple blog posts to explain why I’m interested in this, but the short explanation is that I see a lot of potential for creating emotional and atmospheric experiences with proceedural “vignettes” rather than long, designed experiences.
Of course, that explanation hides a host of technological and conceptual difficulties. Not only is it very difficult to procedurally generate compelling game spaces, but the game-design challenge of coming up with compelling things for the player to do within these spaces may be even more difficult. Nevertheless, this is the territory I’m interested in exploring.
“My general take is to start with the Rogue model of playing in a procedurally generated world in which all player’s actions are irrevocable, but make it so that the gameplay doesn’t revolve around killing, and the player can never die”
But getting killed is not the only source of frustration… Blockhouse has proved that for me!
I wonder if you could just position everything as being opportunity costs. You made the wrong choice that meant your score will be lower, if you had chosen this other thing you would have been rewarded to a greater extent but made it more difficult to earn more later. That kind of thing.
My first reaction is that this sounds like it could be very compelling.
I then started thinking about what would make it very interesting for me personally. My immediate thought is collection. The collection could involve objects, skills, information and status/rewards.
Just off the top of my head, imagine you are wandering through a world in which there are a variety of contests each with a different judge and a different description (e.g. best still life display, most rubies, tackiest music collection, best bestiary, best collection of garden gnomes, most peaceful rock garden, etc). As you wander you will find objects which you can acquire in some way. You can also acquire information about the different judges tastes (e.g. loves gnomes with glasses, hates Elvis, etc.) There could also be other characters who would be doing the same thing who you would be competing with for these prizes and you could interact with and trade with them. You could also acquire skills that would make it easier for you to acquire other things or do other relevant things (eg. feng shui level 2 may increase the peacefulness of your garden, bedazzle level 3 may allow you to increase gnome tackiness levels, etc.) Periodically the contests would be held and rewards given and new contests announced.
Decisions would be based on time, resources and focus. Which prizes do you want to focus on? If you start finding a lot of items for one goal should you switch?
I’m not sure about other audiences, but that is something that I would enjoy, especially if you got to really collect the things and move them around in a virtual space to really create contest submission.
@Dave: To some extent, I’m just using “frustration” as a code-word for the fail / redo paradigm, even though I recognize that there are other ways that games can be frustrating. Getting stuck on puzzles is a different kind of frustration, and I’m a lot more ambivalent about it. I don’t enjoy “getting stuck” in games, but the possibility of it happening seems inextricably intertwined with what I *like* about games like Ico, Braid, and Portal (and Blockhouse). In contrast, the fail / redo paradigm isn’t inextricably intertwined with anything I like about any game.
I suspect that fanatical avoidance of fail / redo mechanics will lead down interesting video game design paths, in much the same way that fanatical avoidance of inelegant mechanics leads down interesting board game design paths. Your “opportunity costs” suggestion indicates one promising avenue of exploration.
Thanks for the article! In among “exploration and discovery” what are your thoughts of “creativity” or at least “a user’s perception of creating”? My fantasy video game is one where the player can have all three feelings: What’s around the corner? What do these clues tell me? What can I build with these widgets? In a procedurally generated world, opportunities for the user to create new plot devices or subplots might increase replay.
Re: Goals
Does there have to be an explicit goal? I would love to see an E&D video game that lets the player decide what he or she wants to accomplish, if anything.
The Flexible Goals mechanic can be seen in a very limited way in the board game Careers, where each player secretly selects his/her own winning condition that has to be a combination of some number of happiness, fame, and money points (totaling 60, if I recall).
The flex goal is even more obvious in RPGs like D&D. Sure, a GM can impose a goal of killing the goblin king or saving the kidnapped princess on a given session, but the rules themselves don’t impose a goal on the game. Instead they merely provide the “physics” of the world. Of course many players take level advancement or collection of money as implicit goals.
This could be used in a video game. Instead of one single “score”, track multiple scores. Strength, level, fame, happiness, beauty, health, cash, karma, etc. for arguably quantifiable properties. You could also add in some sort of “trophy case” for one-time yes/no achievements: visited Wunderland, won a flower arranging contest, etc. Then let the player decide for himself what, if anything, he is actually trying to accomplish. I could see one player bragging to another that she just broke the record for being the happiest person ever at Wunderland, while another is proud of having figured out how to accumulate karma points (whatever that would entail) PLUS a large bank account.
One other brainstorm: What if the thing you accumulate isn’t really just a linear quantity of some type but rather something more abstract? As you accomplish various goals (in whatever order), it somehow modifies you “mandala”. Something simple, like helping a senior citizen across the street, might change part of you mandala from stark black and white to a nice pastel set of colors. Something more involved, like collecting three types of dinosaurs, would change the symmetry from left-right “bookmatched” to four-part kaleidoscopic.
I’m curious as to what kind of environment there would be to explore and what the player’s avatar would be. Are you thinking something ultra realistic, something cartoony, or something abstract perhaps with completely different physics? I could see E&D of the physics of the environment as being the first part of the game. If you’re going after a nature environment, I could see the player characters themselves being different type of animals.
Ok, I think I’ve rambled on long enough with half-baked ideas. Maybe some of what I’ve thrown against the wall sticks.
@Kristin: I like your ideas about collection-based goals, although they’re probably more elaborate than what I’d start with. I think I’m better at designing systems and rules than I am at designing content, so I would feel a bit overwhelmed trying to design this hierarchy of stuff to collect and combining them into different “contests”, etc.
Have you every played the freeware indie game Knytt? The simplicity of that is more along the lines of what I’m imagining. You’re presented with a little world, there are a bunch of objects scattered throughout it, and you explore until you find them all. A combination of this and Rogue would totally float my boat. Just keep procedurally generating little worlds for me to explore, and I will keep playing this game for months.
@John: For whatever reason, I feel a lot more interested in “what’s around the corner?” and “what do these clues tell me?” than “what can I build with these widgets?”, so that’s probably not a design space I would explore any time soon. It’s perfectly compatible with “evocativism”, so chalk up my (current) lack of interest as a matter of personal taste.
The kind of player creativity I’m most interested in is the kind that’s expressed through game decisions. For instance, you and I once discussed a kind of “tribe simulation” game, where you are a shaman or tribal leader, and once a day (each “game turn”) people from the tribe come and ask you for advice or direction. After you play this game for hours, you’ll find that you’ve “created” a tribal situation because of the decisions you’ve made - you’re at war with your neighbors, you have large plots of farmland, there’s a well in the village center, etc. I’m interested in that kind of player creativity. I wouldn’t be interested at all in a tribe sim game where you drag units around, creating buildings on a map, etc.
Obviously that’s not exactly what you’re talking about, since you seem to be imagining moving through a physical space and maybe solving puzzles (gathering clues). So maybe you’re envisioning a kind of action-adventure game where as the player you have a lot of room for creativity, either in how you play it or how you combine various tools to give you various abilities or whatever. And if the game is procedurally generated, it’s more likely to present the player with systems to be played with rather than pre-designed creative “choices”. That sounds neat, but I don’t have a strong vision about how it would work.
@Ryan: I like the idea of letting players choose their own goals, or just lateral drift for a while, but I don’t necessarily know how to design that. Maybe the various Sim games are the closest examples in big-budget commercial games.
In the context of exploration-platformers like Knytt, I don’t have a strong vision about how to make a game like this compelling without some kind of explicit goal. And this kind of game, with its minimalism, doesn’t lend itself well to the idea of tracking lots of stats and accomplishments and letting the player decide which ones to care about.
On the other hand, my brief outline in the previous comment of a hypothetical “tribe sim” game does lend itself to the idea of non-explicit goals. I envision that the fun of this game would be just playing it and trying to steer the tribe in certain ways, and seeing what events unfold. An interesting design question would be, does the player get to explicitly set goals - tell the game “I want to win when I’m the strongest / most well-liked / richest tribe in the area” or whatever - or does the player just play the game, and personally decide when his or her goals are met? The former seems more “gamey”, and for some reason less attractive to me, but on the other hand there’s a game design (and technological) challenge there that I like.
I think you hit on something important. I stopped playing Starcraft for the same reason I uninstalled Warcraft III while my wife was playing it. The combination of realtime strategy violence with frustratingly difficult scenarios made both of us a lot more irritable and confrontational. The emotions invoked by games can be very negative, and this is nothing new.
Back in the days of Donkey Kong, Pac Man, and Zaxxon it was all about learning the patterns of gameplay. Getting good at those games was mesmerizing repetition.
Other games provided different psychological effects. I still vividly remember the extreme frustration of seemingly endless do-overs from dying in a single demo level of Rick Dangerous 2 on the Amiga platform, and then the feeling of achievement after completing the level. Before completing it I about wanted to throw the joystick out the window.
It’s interesting how games have evolved, especially RPGs and the motivations to play them. I often find the “leveling up” motivation to be an undesirable element of MMORPGs because it’s the opposite of how I always approached paper RPGs. It was supposed to be about the actual experience, not experience points.