Friday, December 7, 2007

"Design Reboot"

This came out on November 29, and it's one of the best lectures on game design I've listened to. It's called "Design Reboot" and it was given by Jonathan Blow at the Montreal International Game Summit 2007. If you go to the page on his blog about the lecture, there's a link to the Zip file that contains the MP3 and the PowerPoint presentation.

He talks about what games really are, what they can teach you, why game designers lack discernment, and architecting vs. exploring in game design. I highly recommend this talk because it's a surprisingly truthful analysis of game design. Jonathan said everything that I've been wanting to tell the industry over the past 6 months, was in a position to do so, and told me some things that I didn't know.

I'll do a summary of his talk and interject a few things of my own towards the end of this very long post, but first I should start this out with some of research I've been gathering on the subject prior to when Jonathan's talk was given.

Research
I've been doing a lot of research on game design over the past year. Recently I've been very careful about my sources of research once I realized how many lectures at GDC and similar conferences are simply the ego strokings of the developers; no fundamental principles or laws of game design, just a bunch of, "Here are some of the cool design decisions that made our game so very successful." Unfortunately, they often don't explain their reasoning for those design decisions, just that it was cool and it worked. No theory. No teaching. Just showing off.

Even when the lecturer is trying their best to teach actual principles of game design, it's usually just a regurgitated list of game mechanics from previous games that have worked in the past and have proven to be fun. I believe this is largely due to how young the medium is. We really don't know what games are capable of yet, much like the film makers in the first 30 years of motion pictures when cinematography was barely in it's infancy.

However, I know that video games are going to turn into something significant to our culture in the same way other media have before it. So what are video games really and what are they capable of?

From my own personal observations, I've learned this: games are instructional design.

While working with the Academic Technology department I studied instructional design as part of my job, and one day it dawned on me that video games are in fact instructional design. A good game teaches you a series of tasks that you'll need to be good at in order to beat the boss at the end of the dungeon (look at the Zelda series for very obvious examples of this). It's surprisingly similar to the instructional design process of organizing complex tasks into simple tasks and then teaching those tasks with some kind of reward system to encourage the learner.

The great thing about games is that they are surprisingly good at doing this, and you don't even realize that you are learning something. All you know is that you are being challenged and entertained.

Playing games like Dance Dance Revolution and Trauma Center (all very non-conventional video games), and seeing how fun and easy it was to convince myself to put the time into being good at them, I can't help but wonder if the skills that I learned could actually transfer over to more constructive areas. What if Dance Dance Revolution actually had dance patterns that resembled swing? What if Trauma Center taught first aid? Sure, I can defeat all 8 strains of the GUILT virus, but all that game has done for me is made me really really good at the pointing portions of most Wii games (it's an ongoing joke with my friends that my prowess with the Wii Remote is because of all those "surgery games" I've been playing).

And then you have games like Guitar Hero. What if by playing that game you could actually get better at playing a real guitar?

Because games are instructional design, what are we teaching people, either consciously or subconsciously? If you think about how literature entertains as well as teaches us some kind of truth about the world we live in, what can video games, a wonderfully interactive medium, do for humanity if used it in much the same way?

Valve's Cabal Process
I was really impressed with Half-Life 2. After playing every single game in the Orange Box this past October, I was stunned that every single new game that Valve put in there was golden. Half-Life 2: Episode 2 was amazing, Team Fortress 2 is hilariously fun to play, and Portal was just--wow, the best game I've played in a long time. The dialog was witty and funny, the puzzles were mind-bending and very rewarding, and I learned that "the cake is a lie." Seriously, you need to play that game if you haven't already (it's only $20 on Steam).

How did Valve manage to pull off a hat-trick of stunningly innovative games?

I was determined to reverse engineer Valve's design process. I played through the developer commentary, took notes, read their articles, and actually bought the audio recording of "Valve's Design Process for Creating Half-Life 2," given at GDC 2006, where I learned a lot more about this thing called the Cabal Process.

What was interesting is that myself and a few others are working on starting an design and entertainment studio that is run more like a graphic design studio (as opposed to a production studio), which stimulates cross pollination of creative ideas amongst different trades of design. This was patterned after IDEO's methods, but applied to entertainment in addition to design. IDEO uses small teams comprising of one person from each discipline. Usually that means there's a graphic designer, an engineer, a psychologist or interaction designer, fine artist, architect, or whatever else the projects requires.

I didn't know this until recently, but what Valve refers to as the Cabal Process is exactly what I just described. That's why their games are so good. Now, depending on the size of the project, they'll have three of these agile teams working on the same game.

In addition to this, they also prototype and play test the game with outsiders during the entire process, not just alpha stage at the end. According to Valve, the problem with making fun games is that you don't know if a concept is actually going to be fun or not until you've prototype it and tested it. Play testing helps the game designers see how people are going to play their game and spot problems early. It also helps them decide which problems are the most important, and prevent heated discussions about which idea is the best due to personal preference, because the play testing as already verified what idea works better with real people playing it.

Compare that kind of process to that of colossal production studios, and you'll see the difference in both the kinds of work created and the quality.

Valve uses an engineering approach to game design: "Define your goals and constraints, come up with an idea of how to meet them, perform an experiment to test the idea, evaluate the quality of the experiment, evaluate the quality of the idea, evaluate the quality of your goals, repeat." The way they quantify the quality of their product, in this case the video game, is on the reaction of their play testers. They said that the game is ready when the play testing sessions are not painful to watch.

When I realized that Valve ran like a design studio by using the Cabal Process and had the production philosophy of an engineering firm, a lot of what they do makes sense. Half-Life 2 was really well done, but it was just a shooter. It wasn't until Valve hired a new team of graduates to work on the spiritual successor of their student project Narbacular Drop, that another truly innovative gave came out of Valve. Portal.

So, Valve has a great method for making games, in particular the gameplay aspects of it. So, what do they know about the principles and elements of gameplay that the rest of us don't? Honestly, I haven't found much about that in my research with Valve.

Listening to their talks and their audio commentaries, it seems that any design principle they have for visual design, writing, and interactive design are taken directly from their respective fields.

However, for level design they use terms such as vistas, gates and arenas to describe areas in each level and their function (this appears to only be useful terms when describing areas in a linear story game, which all Valve games are, even Portal). However, for the interactive parts of the game they don't seem to have any firm understanding of what will make a gameplay mechanic work, or established theories as to why. No one in the industry knows what will be fun in a game if it hasn't already been done before. That is why they test gameplay concepts before adding all the art because they don't know if an idea will be fun or not.

It's very possible that using the engineering process is the only successful way to come up with successful gameplay mechanics. We do have a field called interaction design, which is based mostly on human observation, not so much principles and laws. The whole point of interaction design is to come up with ways to decrease the learning curve of complex devices. Something that does this well we refer to as being intuitive. Interaction designers will talk about things such as visual metaphors, affordance, and human psychology to anticipate how a human will interact with a new object, but it really boils down to testing it to see if it works.

So that might be the answer to gameplay design. You just have to test it out.

Games Are Wish Fulfillment
Another part of the design process that I believe exists in game design is deciding on what type of game you want to make. Tim Schaefer (creator of The Secret of Monkey Island, Full Throttle, and Psychonauts) says that games are wish fulfillment. You want to be able to do things that you can't do in real life. So that seems like a good place to start when coming up with the concept for a game. My favorite games have been the Descent and Freespace series because I'd love to be able to pilot vehicles in zero gravity. That's wish fulfillment.

Tim Schaefer's process is to create a world that you'd love to explore and is interesting enough to spend some good vacation time in, then create the coolest character in that world; that is going to be the person you are going to play. This is really great for world creation and developing interesting characters, but that idea also lends itself well to come up with gameplay ideas that people would be interested in.

Unfortunately, even though Tim Schafer's games are very well written, have a wonderful milieu and great art direction, they often don't play very well. He started out doing adventure games and moved on to platformers like Psychonauts, but his games are just not as playable as Valve's. That's what made Portal so important is because it had the wittiness of a Tim Schafer game with the gameplay polish of a Valve game. Interestingly, the co-writer for Psychonauts actually wrote the dialog for Portal.

That's where a large part of my focus of research has been in; coming up with creative and potentially successful ideas for games. Tim Schafer's lectures on the subject helped me develop a process for coming up with new game ideas--and just about everything else entertainment related--by answering the question, "What do people want to do in real life but can't."

So, at this point I know a lot about how to make games, as least as far as the Cabal Process is concerned, I know how to create interesting worlds and characters, I've come up with a designerly method of coming up with creative game concepts. But what about the elements and principles of game design theory?

Do we have ways to describe that kind of thing, or are we just going to have to experiment with different ideas and test it to see if it's successful? Do we have ways to describe pacing in a game? How about controlling the difficulty curve effectively or even dynamically? What about NPC interaction, effectively telling stories with games (people do have a lot of conflicting theories about this one), visual information hierarchy and why do certain control schemes work better than others?

"Design Reboot"
That's where this talk came in because it helped answer some of my questions. He had a really good explanation as to what a video game actually is:

You are trying to achieve a goal, with some rules governing your actions and the game-world's response.

Games create a meaning of life in a temporary, low-stakes subdomain.
He said that when he started thinking of games in these terms, the medium of video games made more sense. It's a world with it's own set of rules that you can experiment and live in for a while, and has no real affects outside of the game.

But that's what's interesting; it does have an affect outside of the game, like all forms of entertainment do. As you play the game, watch the movie, or read a book, you learn something about those fictional characters and the world they live in: what they've done, how they live, what choices they've make and what they believe in.

Then he talks about what games can provide:
1. Entertainment / Fantasy / Escapism
2. Meaningful Artistic Expression
3. A means of exploring the universe
He expresses his frustration with games in that they focus mostly on the first point: entertainment. There's more to life than entertainment, and he's starving for games that offer more than that.

He then talks about how game designers lack discernment, that they only care about whether a lot of people play their game. They don't care what makes people want to play the game, just so long as people play it because the success of a game these days is measured by how many people buy the game. They do this at the expense of the player's quality of life.

Video games provide scheduled rewards to keep you playing. These include collectibles, unlockables, advancing the story and achievement awards. In MMOs, he says that these rewards are fake. The rewards are lies. In an MMO you work hard to get a better weapon to give you an advantage over the monsters you are currently fighting. After you've achieved that edge, they are no longer worth the experience points, and you have to fight stronger monsters that are matched to your current level. You never get an edge. It's like playing Heroin Hero, where all you do is shoot up on simulated heroin and chase a dragon. "Catch me! Common. This way! Shoot up more, then you can catch me." But you never catch the dragon, you just keep on chasing it.

Funny allusions aside, this is exactly what MMOs are like. The reward system is a trick to get you hooked on the game, chasing a reward that you can never catch.

Jonathan then makes the challenge that if you removed all of those scheduled rewards, would people still want to play your game? In the case of current-generation MMOs, I don't think anyone would.

What I found interesting is that releasing games for Xbox LIVE encourages the use of achievements, collectibles, and unlockables. Also, you sometimes have to pay additional money to unlock certain things in the game through the Xbox LIVE Marketplace. When I first watched the Microsoft keynote address at E3 about the Xbox LIVE Marketplace a few years ago, I actually felt sick because I knew what the goal was. Making micropayments to get additional content for a game isn't designed to improve the quality of a gamer's life.

He categorizing game rewards into two categories: food (naturally beneficial) and drugs (artificial stumuli), he says that game designers focus too much on using drugs in their game designs because they don't understand food. What rewards can we get that are naturally beneficial? Remember my comments earlier about Dance Dance Revolution and Trauma Center?

He later uses Portal as a good example for natural rewards, quoting Manveer Heir:
The brilliance in Portal lies not only in its simplicity (and excellent humor),
but also in the moments of realization when you figure out a puzzle.

No puzzle stumped me for more than five minutes in that game,
yet I went from being completely dumbfounded one moment
to feeling like a genius the next, as I realized what I was supposed to do.
One of my favorite parts of his talk is when he takes a jab at World of Warcraft. The rewards are fake in World of Warcraft as well.

He didn't discuss this, but it's part of my personal experience with the game that the gameplay changes the longer you play it. The focus of the game slowly changes from being a casual game of exploration and quest completion, to having to join a guild in order to progress in the game. Once you have joined an active guild the game has begun to be a commitment. Now you have to be sure that you are playing at 7pm every Tuesday and Thursday night so you can be present at scheduled raids with your guildmates.

What does Jonathan say World of Warcraft teaches people?
You are a schlub who has nothing better to do than sit around performing repetitive, mindless actions.

Skill and shrewdness do not count for much; what matters is how much time you sink in.

You don't need to do anything exceptional, because to feel good you just need to run the treadmill like everyone else.
That's what frustrated me about World of Warcraft. I was getting killed by players who only had an edge over me because they had played the game more and had less of a life than I did. I played it because I liked the exploration, but the game wasn't about that anymore after level 30. At that point the game was about survival. All of the quests at that point are in contested territory, and the opposing faction on that server, played by real people, can kill you at any time without notice or consent to a duel.

At level 36 I stopped playing. When I canceled my account they asked why I decided to cancel my subscription. I had no doubt that it was so they could find ways to keep people hooked. So I wrote, "The game is just no substitute for real life." What I really meant to say was, "The Alliance is constantly ganking me because their faction outnumbers us 3 to 1, and they have nothing else better to do an level 60. Give the Horde some hot races so those testosterone-poisoned teenagers can run around in their Horde panties. Then they'll play on our side and I'll have a chance of survival after level 30."

Well, they took my advice.

Jonathan then talks about architecting the game design as opposed to exploration it. He uses Bioshock as an example of poorly-architeched design. The designers wanted to give the player moral choice so they implemented a mechanic where you can kill--sorry, harvest--Little Sisters to obtain a mutagen that gives you special abilities in the game. You can either be merciful or obtain power to give you an edge in the game.

However, since they believed that First Person Shooters need to be well-balanced, no matter what your choices are, the game still has a steady difficulty curve. That idea conflicted with the moral choice in the game. It didn't really matter what you did in the game universe. The moral dilemma of altruism over power was fake. In addition, the game still encouraged you to shoot anything that moved without warning, including the Big Daddies (the Little Sisters protectors who are actually the most caring and protective characters in the game).

He says Bioshock is a weird game and can't be used an example to the rest of the world of teaching moral choice in games. The designers were trying to manipulate your feelings in a very clumsy way.

Comparing Bioshock to Portal, where the development of the game was explorational using the Cabal Process, there's a big difference in the designers decision in Portal to include the same themes in the game.

In Portal, there's this stage where you have to carry a metal crate with you to deflect plasma balls, activate switches, and step on to reach high places. You had to carry it with you throughout the entire stage. In play testing, they realized that players had a certain fondness for the crate, which I'll remind you is an inanimate object, asking of they really had to leave their crate behind.

So, they decided to play on that idea. With a little bit of supplemental architecting, they painted a heart on the crate, had the AI computer that is putting you through all of these puzzles call it the "Weighted Companion Cube," and at the end of the stage the AI then makes the player "euthanize" their Weighted Companion Cube (dumping it into an incinerator) in order to continue on to the next stage.

At the end of the game the AI accuses you of murdering the Weighted Companion Cube, your best friend.

Similar theme, but the choice to include it was done by exploration, being a careful observer to what their playtesters were experiencing, and accentuating it. The focus of the game was to solve portal-based puzzles, but noticing an opportunity and then adding that theme using cleverly written dialog and a little texture work made the game better, and in turn made the Weighted Companion Cube an Internet phenomenon.

Jonathan continues with architecture vs. exploration, and what he refers to as listening skills. "It's hard to listen if you are shouting all the time," he says. Architected games, where you are imposing your original vision of the game on the gameplay design process, means you are shouting. You need to be able to listen. Try out new ideas, pay attention to how people play your game, and let the design of the game evolve. Most modern big-budget games are architected. They have the concept art, the story and the gameplay figured out well in advance. That doesn't give you many opportunities to explore.

Anything Else?
That's about it for now. I'm going to encourage people to write comments about this article, so please leave yours.

I've been keeping lots of notes on this topic, and I didn't cover all of it here. Although writing about it has been very therapeutic, so I'll write more on it later.

Also, if you listen to Jonathan Blow's talk he mentions two books on games-as-teaching called "A Theory of Fun for Game Design," and "Persuasive Games: The Expressive Power of Videogames." I haven't read those books, but I might at some point.

2 comments:

Matt.C said...

w00t! first post!

Seriously though, I really enjoyed this post, and I agree wholeheartedly with all of it. It doesn't matter how many advertising dollars get piled up behind a game (bioshock, world of warcraft), if it isn't truly fun I lose interest pretty quick. You can take a tried and true method of gaming and spruce it up a little (Halo 3, UT3) but if it doesn't stay relevant then it's merely a nice nod to what has gone before.

I imagine that when the first platformers came out people though, "How can this game-style be better than the top down style we've come to know and love? Pong, asteroids, and space invaders are awesome... pitfall looks like trash." And then platformers dominated the market for nearly two decades.

A good engineered system that comes to mind is the early resident evil camera system. I didn't play the games much (never had a system they were on), but the stationary corner camera angle was a nightmare from what I saw. The developers had a vision. Too bad game players weren't that impressed. Now, apparently the games weren't all that bad seeing how well they've done in their niche.

I do a little work here there for fun on story repackaging via video editing (anime music videos). When I'm working on a concept, I like to think i use a bit of the Cabal process mentioned (at least to my understanding). Some people just take their favorite linear video segment from a show and paste their favorite song on it and think, "This is great!" The problem with this method is that usually it isn't great. It's meaningless. I try to think about the "psychological" effects of visual and auditory combinations. Though I do have an envisioned architecture in mind, I constantly gauge people's reactions to what I have, what works and what doesn't. To me, this makes sense.

Why don't the large production companies follow this exploratory process? Is it just habit to pay developers, writers, and designers to churn out another hyped market winner? Do they feel someone might find out about their idea from the early production test and make a better game? Do they hire less than competent people and not want to admit it? Likely, they just want to churn out another cash cow in a short period of time and exploration can take a while (Team Fortress 2, StarCraft Ghost...)

That's what I have to say for now. I apologize that it's not that organized. Keep me updated!

Glen Moyes said...

I haven't played the Resident Evil games yet (although RE4 for the Wii is really tempting in the near future), but it worked well enough that Cliff Bleszinski used that over-the-shoulder viewpoint when he designed Gears of War. He referred to the game as "Resident Kill.switch" because the gameplay of Gears was designed to be a mix between the two games.

Third-person games are more difficult to do well, but it gives you the opportunity to see exactly what your character is doing, which can be important to gameplay. It's also more fun to watch.

As for why big studios don't use the exploration method of game development, a lot of people say that it has to do with risk. Most publishers are publicly owned, and risk is a big issue for stock holders.

The reason why so many WWII, Alien vs. Humans and Tolkienesque fantasy games are made, is because if you advertise those games just right they will sell. It's kind of like what we see with bad movies that are marketed to death before release; they are meant to sell big on opening day before they tell their friends not to go see the movie. Alternately, an under-marketed movie (such as Ratatouille) will do poorly the first week but do much better once the word has gotten out.

As a general rule, if a game is over-marketed it's probably not worth it.

That's what disappoints me; if a game is over-marketed and it sells well, it is seen as successful game. If the success of a game is in fact measured by how many people bought it--using that as a metric of how fun a game is--and if that number is skewed by marketing to begin with, then the entire basis of what publishing companies decide to make is built upon a negative reinforcement loop. In other words, they are getting it all wrong and it's spiraling out of control.

As far as the entertainment industry is concerned, I always tell people that the pendulum is going to swing back. When the market becomes unbearable people start to react. Gamers have complained about the quality and creativity in video games for a while, but now video game studios are starting to do something about it.

I don't think Valve tried to make the gaming world a better place with their methodology, but people who do want to change the world will be looking at their work as a template on how to create better games.