The talks I attended on Thursday; all were good, but the Walking Dead one was particularly enlightening.
10 people in the audience as the talk kicks off.
This talk is the result of a conversation with himself. (He tends to win most of those arguments!) And it’s the result of years of frustration and disappointment within the industry.
We join our story after a crazy couple of years. Established franchises aren’t doing well, over a hundred studios have been shuttered in the last five years, social gaming didn’t turn out so well, piracy, etc. And there’s new hotness on the horizon. But of course this is all part of a cycle, we’ve seen this before.
And a big part of this cycle involves mistreatment. (I didn’t copy down the examples, but he had good ones.) So: why is this a cycle? Because the developers remain, we persist, we’ll ALWAYS be here, we ALWAYS TAKE IT. (Caps in the original.)
Aspects of this are like the worst relationship you can imagine. You always tell yourself that you’ll leave, but you don’t. Because when it’s great, it’s great, but the awful keeps on coming back.
Thinking about this more: the common thread in all of this is himself. What can he do to extract himself from these loops? Where to begin? He saw two paths: a personal/cultural path, and a community path.
Game developers come out of nerd culture, which is a fringe culture. This, in turn, leads to an abdication of responsibility: bad stuff is happening, but it’s not our fault, it’s stuff that other things are doing to us. In other words, blame culture: you’re part of a large organization, but you don’t feel obligated to do anything, instead you suffer and go along with it.
You might go along with it while loudly complaining, pointing out the dysfunctions around you. But that’s still blame culture, that’s abdicating responsibility. (Hmm, I wonder if Dustin is familiar with Christopher Avery’s way of thinking? I asked him about that afterwards, and the answer is no.) So: don’t use “it’s my job” as an excuse for going along with something bad. It’s unhealthy, and you’ll burn out after a few years anyways if you do. If things are so bad, don’t bitch and moan, do something about it. (Which quite possibly involves “go fucking find another job”.)
Triple Town / Yeti Town cloning example. This is bad for so many reasons. But also: what are they saying about you when asking you to clone a game? They’re showing a complete lack of respect for your own creative abilities. And it’s well below any reasonable ethical threshold.
Work against passive aggression. Where did we forget that being honest is actually something that we can do? One of the rules of his studio that you don’t talk about people that aren’t there: that tamps down on passive aggression, and leads to developing a skillset that lets you be honest without being an asshole.
There is a determinedly irresponsible perspective of game developers about their role in the community. We need to stop waiting for someone else to stand up for us. The ESA does not represent developers. We are not as protected in the US as we might think. Are we going to be the next medium to be smooshed into a box, to be forced to be plain vanilla? Stop being a punching bag, blamed for society’s ills.
What can you do? Be vigilant on the local laws that impact your area. Go on record at public meetings to discuss the benefits of games to entertainment and education. Don’t expect someone else to do it. And also: understand how the game is played. It’s not fair that money and connections influence politics, but wishing won’t make that go away. Get involved in local school; mentor; donate to charities.
And leave things better than they were when you found them. Organize: creative people are not necessarily good business people, you need to work together. Yay unions; would an organized labor union work in this industry? (The IGDA is a trade organization, it’s not a union.)
Founding a union is not easy: you’ll be blacklisted, you’ll be pushed away, you’ll be made to feel like you’ll never work in the industry again. Up above, he said: if you’re asked to clone a game, you should quit. Are these costs worth it? Also, are game developer attitudes compatible with union bureaucracy? He doesn’t know if we can do this, but it’s a question worth asking.
The game you’re working on might be your last game. Who takes care of people who have been driven out? (E.g. by health issues, carpal tunnel.) Should we?
Nothing is normal right now. For a few years, it has felt like something is wrong, but he’s excited: we’re about to reinvent ourselves in a very serious way. So let’s do that in a thoughtful way, and leave the industry in better shape than we found it.
(Audience question about young game developers who already feel shut out. Dustin says: maybe the indie game movement is a union, of sorts.)
The Walking Dead is an adventure game, but one that focuses on dialogue and choices, not on inventory-style puzzles. (The inventory-style puzzles are some of the easiest in the studio’s history.) They took a lot of the traditional game elements out; what’s left? In The Walking Dead, game design is character design.
How do you play a story? Here, character and story are the actual game. Are games at a disadvantage here compared to movies?
A question that motivated The Walking Dead: what does a person do with the time they have? This led to lots of choices. And, to make the choices matter, the game never gives you qualitative feedback about your choices: no light/dark meter, etc. “The game players played in their heads was more dynamic than any system we could have built.”
In most games, you poke at the game, to see what effect your actions have. The Walking Dead has a much more limited set of systems, so it ends up flipping this around: the game pokes you.
Proaction versus Reaction. Minecraft is an example of the former: the player has access to a huge toolbox that lets the player make stories happen. (And who the characters are is almost irrelevant.) Spec Ops is an example of the latter: you have a predefined story that happens to you. The Walking Dead is also on the reactive end of the spectrum; to help players react, context is important. What is happening, whom is it happening to?
What is happening?
They needed to give people time to absorb the context. So they left a lot of space before the important choices happen. This took a while to learn: in particular, more branching choices does not equal more meaningful choices. In an initial version of Episode 2, they had tons of choices in the first half of the story. But it didn’t work with The Walking Dead: all of those choices meant that the meaning actually fell away. Instead, what’s important is spending enough time with the world and characters until you feel that you know them well; only then do you add choices into the mix.
They prioritize emotional outcomes over victory states. And players generate their own meaning and their own emotional goals. There’s a context / reflection loop going on here.
They knew they wanted a big choice in the first episode; having somebody die seemed like a good candidate for that. So: Doug and Carley. You meet them in the drugstore; the idea was to get to know them both as people, and then be forced to choose. Carley is a very competent journalist; good with guns, but she also knows stuff, in particular Lee’s back story. Doug is a web developer, more of a tinkerer type, who had already saved Carley.
You work with both of them: with Carley to go to the lodge on a rescue mission, with Doug to get keys to break into the pharmacy. So you get to know both of them. But what turns out is: 75% of people saved Carley, and this felt off-kilter.
On surface: Carley is an attractive girl with a gun, while Doug is a dorky guy who knows controls to remote controls? This seems like a pretty obvious choice.
Going in deeper, though: they’re both people. As developers, they liked both of those characters as people. Maybe that didn’t come out?
When you first meet Carley, she risks her life to say a boy. Later on, she tells you she knows your secret and will keep quiet. And the two of you talk about keeping a child alive in an apocalypse. Also, she has your back in a mission; at the end of that mission, you share a horrific experience (the woman in the motel who was bitten and wants to kill herself), and talk about its effect on both of you.
Doug, in contrast, is a passive observer during the initial rescue of you and your friends. Then you make a plan to get the keys; but it’s predicated on you passively lying to Doug, because otherwise he’d know you’re a murderer. And there’s very little shared context with Doug beyond that.
So, regardless of the details between the two, there are many fewer opportunities for context building with Doug; given that, players will naturally gravitate towards Carley. The experiences the player has with other characters in the story is what shapes the choices they make.
Who is it happening to?
Who is our player character? A blank canvas, they needed to put some stakes in the ground. Not the camp leader: that nullifies too much of the potential tension. (You’d be making decisions instead of asking why a decision is correct.) He’s old enough to want a family but not have one. (They weren’t prepared as designers to make you care about a pre-existing family in a way that a parent would.) He’s physically able but not imposing. And he has a difficult past the player could build their own relationship with. Someone you root for, but who has major personal flaws.
Lee’s race is not on that list of attributes. It’s informed by the social facts of the region they’re in, though. White is the default choice for video game characters; that creates a context, and they needed to figure out what they wanted to do with that context. Also, because of the number of mechanics that they stripped out, everything about the context has outsized weight. (A phrase they use when thinking about new characters: “what rocks do they have in their backpack?”)
With Clementine, you pick up a fair amount of context before really meeting her. The treehouse, her family’s house, the fact that she was staying with a babysitter.
A basic question that she helps with is: why don’t you leave this group of assholes? Because, without Clementine, it might be a lot more natural for Lee to want to leave the group. But with Clementine there, staying with the group seems more natural.
In general, being around tense/annoying people is stressful. So they needed a way to manage that in a way that helped the player to stick with the game. Their main tool was empathy: helping you feel other people’s experience at a conscious and subconscious level. So they needed to build empathy, even towards people who are acting like assholes at a given moment.
E.g. Larry. He never says a nice thing about Lee in the game; most players did not like him at all. Players don’t really end up empathizing with him too strongly, but at least they can see Larry and Lilly’s relationship as a mirror of Lee’s and Clementine’s.
Players get frustrated with the nature of their relationship with Larry (there are no good solutions for players in the game, really), but this doesn’t get frustrated with their progress towards game goals.
Graph of food: 42% of people feed Larry, which is fourth on the list, behind the two kids (96% and 95%) and Mark (58%, a new character in episode 2). So people are more likely to feed Larry than any adult you knew from episode 1.
The job of the game: empower as many different player stories as possible. Lee can’t decide to negate the player’s story, though re-examination and re-articulation is okay.
If games can expand how we feel about other people and do good business, that could mean something pretty special to us as an industry.
He’s actually not talking about sex: he’s talking about how BioWare, in its approach to romance, has had to approach questions of sex and sexuality. They first approached romance in Baldur’s Gate 2 – at the time the idea of romancing your party members and embedding that into the plot was new. (He wrote three of the four romanceable characters.) Only one of the four romanceable characters was male; female players complained that this was unfair.
At the time, he thought that those complaints were unreasonable – romance options are expensive to implement, they didn’t think they had many female players. Still, he had that in mind. For their next game, Neverwinter Nights, he’d gotten tagged as the romance guy; he wanted to write a male romanceable character, but wasn’t sure how to do that best. He went to one forum that was home to many female BioWare fans; by Knights of the Old Republic, Carth was relatively successful, and BioWare was known for their romances.
Jade Empire had a same-sex romance (and apparently there was one hinted at in KotOR, I didn’t catch the details?); he didn’t work on that game, but his next game, Dragon Age: Origins, had four romance options, two of whom could be romanced by players of either gender, and included the first BioWare sex scene between two men.
For Dragon Age 2, all four romance options could be romanced by either gender. After all, why not? Each romance option was relatively expensive (lots of dialogue), but once you’ve gone to that expense, having the option romanceable by either gender is cheap. There was a surprising number of people who were bothered by this, but also a lot of people who greatly appreciated this.
So that’s where we are: BioWare has sex in its games, they’re not the only company who does that. What’s the problem?
Well, people who don’t play games think: games are for children. And, sadly, the industry itself thinks that games are for men ages 18-25, or maybe a bit lower. Both of those are very out of date: these days, the gender split is close to 50-50, and the average age is 30. You can hypothesize that the older, female players are playing “casual” games; in Dragon Age, their metrics suggest that 30% of the players are female, and that, depending on the character, up to 24% of the romances are same-sex.
One reaction to this: great! Lots of women playing games, lots of women playing our games, we don’t have do to anything! But development costs keep on rising, so we need more and more sales. Is there an untapped market out there? Quite possibly: the industry isn’t actively trying to appeal to large swathes of the market; women, adults, minorities still play, but maybe they’d play more.
Sexy may be good; but sexy isn’t the same as sexualized. And there’s a lot of bad sexualization out there. BioWare isn’t immune to this: you can make a good argument for Isabella being positive sexuality, but even there there’s room for both sides, and moving over to Mass Effect, Miranda’s ass shows up in ways that aren’t so defensible.
And, honestly, what the “right” answer is isn’t the point: we’re coming under increased scrutiny, we’re a maturing industry, and people are going to form opinions about us based on what we see.
So why isn’t this a bigger deal? One claim you hear: “This is what sells.” Which conflicts with the earlier desire to appeal to untapped markets; but conventional marketing wisdom is that proven techniques are what matters. But there are so few big-label titles out there with female protagonists, and no clear bar to overcome here. And conventional industry wisdom has been wrong over and over again.
Conventional wisdom is a manifestation of privilege. If you’re part of a group that has always been catered to, then this catering is fine and is unnoticed. For people outside that group, though, it’s not so great.
And when some people see their privilege exposed, they get really annoyed. Because some straight male fans really didn’t like it that Anders made a mild pass at their characters. They still have the same number of romance options, but the fact that other people now got as many romance options as straight men really bothered them. And those fans aren’t alone: see the abuse that Anita Sarkeesian had to put up with. The game industry is complicit in this.
So what is he suggesting we do about it? As a first step, don’t even worry about attracting women: worry about not actively repelling women. And worry about the influence you currently have: the game industry’s choices has an effect on how people think.
Ask yourself: could this character be female, black, gay? If they already are, how are you using them: are they awesome on their own, or are they there only for the male fan, to forward a male-centric plot?
If you don’t know how to make female characters awesome, or just not male-centric, then ask for help. Maybe through consultants, but also hire people with different backgrounds on your team, and listen to what they say! The Dragon Age writing team is majority female; even so, the men on the team have blind spots. He brought up a specific example from a discussion among their writing team, of content that was offensive in a way that wasn’t intended by the person who wrote it; offense by itself isn’t inherently bad, but if you’re going to do that, do it intentionally, for an artistic purpose.
Different viewpoints on a team are an asset. When you’re hiring, assets are what you want; this is one to keep in mind.
What parts of the emotional palette do games explore? Early games, arcade-style, explored achievement. Then games for teenagers explored empowerment: sports, action, shooting, driving. More recent games: social aspects: Sims, Wii Sports, Rock Band.
In college, he played a lot of World of Warcraft. But players he encountered on it were interested in different things from him, so that made him feel lonelier. He wanted a game without violence, and where everybody is the same (no distinctions of age, gender).
A vision of standing next to somebody on a bridge, overlooking a waterfall. Another vision of standing quietly with people in a snowy field, with colossal figures walking around, trying to avoid their attention. (He’d just finished playing Shadow of the Colossus.)
They formed thatgamecompany, with Sony’s help, and made Flow and Flower. He felt it was time to try online gaming now. Social games were on the scene, but they were about numbers going up, more like arcade games than his social vision.
He met many astronauts; they felt that being on the moon makes you mystical. On the moon, there’s nothing, and the earth is so small that you can cover it up with your thumb. So: why are we here?
These days, they’re so much we can do, we’re so empowered. And games reflect this concern with empowerment. So, if in a game, you have power, then rather than being about survival or connection, the game is about power. (In Left 4 Dead, he only feels a connection when somebody is patching him up.)
So: reverse the relation between the player and the world. Make the player less powerful, the monster more powerful. Or, in multiplayer, make the other player more powerful, more important to you, the monster and world less important. (I missed some of the argument here, so some of the above is probably wrong.)
How can we keep the focus on other people? Get rid of noise; get rid of weapons. So that’s like in a lobby waiting around. To heighten the focus shrink down to two people, in a hazardous environment: if you’re in a desert, a lone person on the horizon is very interesting to you.
That’s the emotion he’s trying to evoke; how do we get there? There aren’t a lot of example to look at, after all.
So prototype; they started with music. He showed a concept trailer they made four months in; the music is there.
But there were problem; the trailer is the only thing that worked well! He showed a rope-based prototype; it only worked in multiplayer, though. Also, their 2D prototypes ended up not translating to 3D – e.g. a 2D prototype that depended on trails and on strategic movement/collaboration didn’t work nearly as well with 3D cameras.
They think of graphics as gameplay. The distance of the mountain posed a problem with that; so they added trails and dunes for local variations, and then sliding down dunes to make that traversal more fun.
How will the mechanics of setting up multiplayer work? Usernames take you out of the world. Voice chat adds its own set of problems. Sony recommended friend invites; but playing with friends whom you can’t chat with is frustrating, while allowing chat interferes with the integrity of the game. So: no friend invites, and usernames only show up in the list at the end.
They also experimented with four-player groups; but that set up 2 vs 2 and 3 vs 1 narratives. (A single player who wanted to explore slowly caused particular problems.)
Flow for collaboration: too much me is alienation, too much we is conformity, flow is “coliberation”. Give the player a choice for how to navigate that; the result is Journey‘s seamless online lobbies. The first few people you match up with might not work, but eventually you’ll find somebody who is a good fit, and you’ll have a stronger connection.
What about resource consumption? At first, they had the player taking resources, but dropping the ones they’ve used behind them. Psychologically, this feels like somebody else taking advantage of you: stalking behind you, picking up your stuff. So, to eliminate this problem, they got rid of possession: infinite resources, everybody can pick up from that pool, but limited pocket size.
The next question is physics. They wanted to let people help each other; but the physics necessary for that had darker uses. (“For quite a while, I was disappointed by humanity.”) Morality doesn’t carry over to game worlds, and players want feedback: killing somebody gives much stronger feedback than helping them over a rock. So eventually they gave up on collision, but standing next to the other player gives them money. That removes feedback for trying to harm people, increases feedback for being together and helping.
Once the mechanics are figured out, the next step is: pushing for catharsis. For Journey, they’re pushing for feelings of awe and mystery; he read some Joseph Campbell to give him ideas about how to heighten this. He liked this, so he created areas that mapped to Campbell’s monomyth.
When working with musicians or visual artists, it’s easy to talk about emotions like sadness; but we don’t have experience mapping that to mechanics. At the end of the first year, they actually had the world mapped out, but the emotions were flat.
By the end of the second year, it was working a lot better: they thought they were matching the emotion curve pretty well. But playtesters didn’t like it, they said the game was bad. In particular, they didn’t like the end, and recommended cutting it. (That’s where the emotion curve didn’t match what was desired.) There was one interesting experience, though: a game crashed, the playtester didn’t realize what was going on, and spent a couple of minutes staring at a white screen figuring out what was going on.
So they worked on the last level, to heighten the effect there. There’s a wind requiring struggling, the scarf can freeze, monsters seek you out and can crush structures. They added a rest area, a fortress wind area, and the walk of death; the player’s movement slows down more and more there. (Finding the right length for that walk was a challenge.)
And then the summit: you have to make it much more exhilarating. Initially, the final area looked great but was on rails; they opened that up a lot more, and added a surfing area to make it more fun. The final area was very free; even at the end, walking into the light is your choice.
At the end of the third year, three out of twenty-five playtesters had tears in their eyes. That’s a success. (Playtesting for emotion, not just usability, is crucial, but you don’t get it until the very end.)
He was nervous about the launch. But in the forum, he saw people thanking and apologizing to each other; he couldn’t imagine a Call of Duty player doing that, but it’s the same person!
Fan art reflects the most intense emotion: he saw drawing hearts at the end of the game, the joy of surfing, the fear of being hunted, and struggling together through the snow. And the fan mail: they knew they’d changed peoples lives for the better.