Governance in F2P Multiplayer Games

[Music] Are you are you alive from all last night? No, I know that feeling. Um, this is governance in free-to-play multiplayer games, which is, I have to admit, is the single most boring title I have ever titled a talk of. However, the topic is actually uh rather exciting. Um, so this is me. I'm Daniel Cook. I'm a game designer. So, this is a game design focused talk. So, I apologize if that surprises anyone. Uh, the, uh, games I've worked on, I've worked on Triple Town, I've worked on Steamirds, uh, Realm of the Mad God. U,, we've got a recent one coming out now called, uh, Road Not Taken and All Your Planets. And I've done a lot with like trying to do original game mechanics that have long-term retention and are evergreen and last for a very long period of time, but still managed to have this sort of

service-based um free-to-play component to them. And so my perspective is a little odd because I'm trying to do original games, but also in this free-to-play space that is very much about optimization and pipelines and content management and that type of thing. So, this is a game we worked on called um Realm of the Mad God. Has anyone here played this? Bunch of folks. Great. Um it's a uh it's a free-to-play um basically dual stick shooter. Uh and it's also an MMO. And the strange thing about it is you build up these little pixelated characters and you run around and you shoot things together and it's cooperative. And then you die and you realize, oh, I died. Where's the respawn button? And there isn't one because there's perma death. And it ends up being a very, very brutal game. Yet, the retention is not

that bad on it. It's uh um it's this game that shouldn't really work as a um high retention free-to-play title that monetizes well, yet people were sticking around for a relatively long term. And the people that got invested in it ended up um building these crazy sort of like societies and governance structures and systems of of interacting with each other that were much more than your typical um your typical uh shooter. You know, you're basically going pew pew pew pew the entire time and suddenly you have culture emerging from that. So this made me question some of my big assumptions about how I was designing these these services. We have traditionally been a contentdriven uh media. We tend to focus on content. We tend to say if we make new content or we

make new events or we make new items or we make new levels then we will get people to pay for that content and we will have delivered value to them and then they will pay for us pay pay for that value. And if you look at a lot of the gating structures that you see in um free-to-play games, they're often based off this fundamental philosophy. You have a certain amount of game, you have a certain amount of puzzles, and then you say you you make it really really difficult or you frustrate people or you gate people and you say you can't get to that. And then if they pay you money, they can finally get to that content that you've you've uh kept from them. and we use metrics and we do all these techniques to slow the decline of the game, but ultimately there's diminishing returns on this on this strategy. Um, people burn out on content. There's a psychological phenomenon where when you give people essentially more of the same, they start

to become acclimated to it. There's sort of this hedonic treadmill that happens and they say, "Oh, I've seen this before. Maybe I'll move on to another another title." And so there's only so much content you can add before your evergreen lifetime 10-year free-to-play service turns into a one or two year or maybe even a few weeks um activity for most players. So what was interesting about Realm of the Mad God, it was a very very different dynamic. The amount of content that we put into it initially was quite low. Um, it's almost like what we were doing was we were seeding these laws for this universe and then the players ended up creating immense drama out of that. And I I have to admit that we didn't do this intentionally. It was it was very much, you know, we stumbled upon the right initial conditions and then this all this this

lightning in a bottle happened. And as a designer, I don't believe that lightning in a bottle is always the reality. I think I think it's this uh that the random chance is less likely. There's actually real reproduc reproducible human dynamics at play here. And if we could understand those and spread those to our developers and our designers, maybe we could have this sort of dynamic where the players are resting on a foundation that we've created, but they create this immense explosion of player dynamics and and content. Um if we we could perhaps replicate that. So that's is sort of what this talk is about. It's like what are the what are the foundational elements that we can put in our games that result in this strong emergence. And these are some of the things we saw. We saw like groups forming, hierarchies forming, trade popped up. People were creating laws. There were punishments like there were mafiaesque organizations who had like

political affiliations with each other and they would they would blacklist certain people and they'd whitelist other certain people and then they'd fight over it and then they were like spreading propaganda to each other about this. And this existed in the game and it existed outside of the game. Like people would go and have entire forums dedicated to smearing other people, but then the people who were being smeared would gain a uh people who would feel for them and that would create an entirely new group and then that group would turn would gather enough entities around it that they'd form this sort of like policing structure within the game. And it was there were orders of magnitude more interesting systems and dynamics that came out of it than what we put into it. Very very very different than the content model. Um now anyone who's lived through this realizes that this is messy. This is extraordinarily messy. Um there were

all sorts of design challenges like in terms of making rules for this group. Um instead of having clear pipelines where we knew exactly oh here's the pipeline and then the person goes and experiences this event and then 92.3% then experience this event and there's the little clear pipelines down through it. The players were creating most of the rules themselves. the players were creating these like crazy systems and they were organic and unofficial. Um, and there was this very clear dichotomy between sort of the local power of the players and the type of of governmental systems that they were creating and us as developers who it was a very small team and we didn't have the ability to even read half the stuff that was going on. And so what could we do as official rules versus what the players were doing as local rules was a giant giant issue. Um and you start to get into all the fundamentals of governance um at this

point like how do you make rules? How do you revise rules? How do you enforce those rules? There were of course hackers and cheaters and griefers um policing judgment of edge cases. Um there were things that were allowed by the rules but seemed immoral. And so the players would become upset about them and then the unofficial norms would try to police the people who are doing like, "Oh, you really shouldn't use that item in this particular way to prevent people from getting into this dungeon. That's that's that's an immoral activity even though mechanically it's allowed." Um, and then certain groups became extraordinarily powerful and they started doing power brokering between different players. Really quite fascinating stuff. But all these are massive design challenges that essentially are do do not exist in the same state in these more controlled content drip style games. So I call this governance and

that's the closest term that I could find to describe what was happening here. And governance is these systems of rules that enable groups to accomplish more together. And this starts to get at the heart of like why this is happening. These groups are in this situation in this particular type of mechanical environment where they feel the urge to work together. There's not individuals who are empowered to conquer the world by themselves. They're forced to work with other people and somehow like improve their lot in life. And in order to do that they adopt these systems of these systems of rules. This this governance um structures um it's almost like you can think of it as technology. Now some people ask like why why governance and not government? The uh government tends to be official and governance is rarely 100% official. A lot of this stuff is organic and coming from the player base itself.

So, of course, anyone who's been watching games knows this is not new. Um, Eve is sort of the the classic service-based game that has been doing this for years quite successfully. Um, and there's all sorts of crazy wonderful stories that come out of a game like Eve where they've set up these sort of physics of the world and then the players take that on and and create giant corporations and scams that last for, you know, months upon months. Um, however, we're also seeing this as the root of massive successes outside the free-to-play world. If you look at games like um Day-Z also has this element of immense u player dynamics that emerge from a relatively simple rule set. Rust is another one that's become quite popular recently and of course Minecraft is the one that's been doing it for a while now. So we see reproducible success over and over

again. Now a lot of the rules are of course unofficial at this point. People don't really know why these games are a success. And so this is about like digging in and finding out why is this happening. So why do this? This seems like a lot of work. It is a lot of work. The contents based models like I I I love Google Trends. I don't know if anyone else here uses Google Trends. What tends to happen is you get a few good years out of a game and then it drops off. People think they have this evergreen hit, but people burn out on them inevitably. And it's really fascinating to watch. You have companies who are really really excited about their their hit, and then you watch how it it like even though they update it, even though they have all the money in the universe to go and like make the best game possible, the the fundamental design structure of their game by releasing on by relying on content means

that it will die out. That's just an elemental aspect of of that architecture, that play architecture that they've uh they've built into their game. And then if you look at um something like Eve, you see that it also has a little bit of a drop off, but the drop off is far less steep. Um it it goes and it sort of bumps along for a very long period of time, and you start to get that that fat tail of retention and revenue for a very long time. And that's almost entirely driven by the player dynamics that are keeping the game alive. They add an immense amount of content to the game, but the content isn't what is driving that retention. It is the fact that you've got this um I think of it as um tying back into motivation theory. You've built a game where there's high agency. People can like set up dictatorships. They can set up trade organizations. They can work with their

friends. They can do whatever they want and they because of that they feel immense amount of like ownership over their lives within this this environment because the dynamics of politics and social organizations are impossible to master. There's always an immensely high mastery level for the for these systems. You can spend your entire life trying to master the politics of one of these games and never quite succeed. you'll just, you know, you might rise up a little bit above the surface and then sink back down and learn a bit and make some new friends, make some new relationships. And so, we've got agency, we've got mastery. And because one of the big aspects of that drives these things is big group goals and working together, you also end up getting shared purpose. And in a lot of free-to-play games, that's an element that's absolutely missing from them is this idea of shared purpose as a group. And of course, Minecraft just seems to

be keep going up and up and up. So, these are very different trend lines. These are very different um evolutions of a product than we see in the typical uh free-to-play space. So, all right, how the heck do we do this? So, what I did is um I talked to a whole bunch of other designers. I looked at our own games. I also went and I there's an immense amount of research out there about how um relationships form, how governance forms um and people have been studying this for an extraordinarily long time and we need to bring some of that understanding and knowledge into our industry so that we can use it in our games. So the the most useful perspective that I found was an economic perspective and um there's a fellow named Ed Castanova who's done a lot of studying on um economies in uh online games and monetary systems and how the

how they all work and fit together. And talking to him about it um he says look it's really simple it's incredibly simple. Governance happens when one thing occurs. When if you can answer yes to this one question, you're going to get governance. And it's really super simple. Is the group more efficient than the individual? And we have this sort of like techno utopia sort of world view of of how players should be treated in our games. We tend to believe that the empowered player is a happy player and the empowered player is someone who is capable of soloing the game. They're capable of playing through the whole game by themselves. And it turns out that that sort of idyllic utopic idea of how the empowered player should be is actually an antithesis to the creation of all these dynamics. You actually want players that are in situations where they want to work together with other

people and they can get more done together than they can by themselves. So use this as a lens as we go through the rest of it. just use that as a lens to think of the think of um um to put put the other things into context because the other things are almost supporting details to this fundamental idea. Uh now a great example of that has anyone watched uh Twitch plays Pokemon? Um there's the game Pokemon Red. Um there's a chat channel. In the chat channel you can type commands, move left, go up, go down, uh press the button, things like that. The game will randomly take whoever happens to get in the queue at the right time, it will randomly take one of those commands and execute it on the stream on the screen. And thousands of people, actually millions of people with millions more watching ended up beating the entire game of Pokemon with this absolutely crazy group control scheme. And one of

the interesting things about it is individuals working alone end up generating pure noise. There's almost no, it's it's impossible to beat the game just by entering in what you think are reasonable commands, especially in the puzzle rooms and puzzle situations. And what ended up happening is all these groups formed outside of the game to talk strategy, to talk like how are we working together? How do we want to work together? Um, Reddit had giant forums full of propaganda posters and it would became this like crazy almost religious phenomenon about how people were desperately trying to work in this situation where the individual was massively unempowered but as a group they could accomplish stuff together. And they ended up having these, you know, anarchy versus democracy like rule systems for how they ended up u determining what was the right activity to do. And we saw this again. We saw this explosion of emergence just because that one simple principle was at

play. Now there's foundations to all this. Now the first one is comes from economic game theory and it's the idea that relationships are an iterated game. So this is a reciprocation loop. We have them in a lot of our games already. There's a a person A and a person B. Person A gives something of economic value to person B. Um, it's fascinating. If you ever read about friendships and the actual like dynamics that go on when you study friendships, we often have these romantic ideas about how friendships work. Two people, they meet, they see each other, they happen to have a connection and that everything's wonderful and then they become friends and they play chess together and then they take their dogs for walks, right? And if you actually dig into it, it's it uh the economic model is much more predictive than sort of the romantic mutual attraction model. Um almost always the there's there's strong evidence that um friendships are

started when someone of lower economic status or uh socioeconomic status goes and inter and gives something to someone of higher economic socioeconomic status. and the person of higher e socioeconomic status only responds to the one who's lower if they see an advantage in doing so. Um so a lot it like it ends up being this very much this back and forth and at any point that that reciprocation loop can break down. So Skyrim not a multiplayer game doesn't have that reciprocation loop. NPCs do not count. They have no interesting dynamics that emerge out of them in this way. So, that doesn't work obviously. Let's just for a moment. Um, try a little experiment cuz I know it's it's almost lunchtime and uh just turn to the person next to you. There's there's this is these are this is an example of diads. So, two people at once. So, just turn to the person next

to you. I I want who whoever whoever is on the one of you raise your hand. One of you raise your hand. Okay. Um, I want you to look the other person deeply. This is really important. Look them deeply in the eyes and say to them, "I am a thunder bunny." All right, at the count of three. One, two, three. All right. Now, what you've just done is you've made an overture. um you may have felt a little nervous. There may have been some awkwardness there. But there what's interesting is when we talk about these economic theories of of relationships, it's not necessarily about money. The costs are things like rejection, loss of social status, time, like you took a little bit of your time, you you you risked like the other person rejecting you. Um, and so when we build

in these loops, we have to consider like as game designers, we have to think about like what is that token that you're giving? What social value does it have as opposed to pure economic value? What is the cost if that's lost? You know, things like identities, um, like relative status to one another. Those are important factors to consider. Now since I don't want anyone to be left hanging there without the complete reciprocation. Now the other person look your fellow DIAD member in the eye and you want to say respond back to them. I too am a thunder bunny. All right. One, two, three. I too am a thunder bunny. And what you've just done is actually about 10 times more interesting and complex than what happens in most um free-to-play games right there. You've set up you've set up a reciprocation loop

between that person. You've looked them in the eye. You had high bandwidth interaction with them because it was like in person and you could look you could see their facial expressions. There were smiles. There was actually a lot going back and forth between the two of you that started the seed of a relationship. And that's that's a fundamental piece to how we start getting governance into games. Now, potentially one of you was a jerk um and did not reciprocate like it was not worth your time or your or your social status in order to do so. And spoilers are super common in this. I would say that the vast majority of reciprocation loops they start with an initial overture and that overture is not returned. So that happens a lot. It's a fundamental part of that. You need to watch for it and it's almost this is actually something you can you can build into your um into your metrics to understand like where where the reciprocation is breaking

down. Um now this is one of my favorite techniques for getting around this. A lot of game design is zero sum game design. If you've got it, I don't have it. What instead we want to do is nonzero sum transactions. A common issue in MMOs is loot stealing. So I start beating on a monster. I'm almost dead, but he's about to die. And then Steve Moretsky comes along and he he gets that last hit in and he kills the monster and he gets all the XP and then he's playing the damned rogue and he's faster than I am and he runs over and he steals all the loot from me. And now at this point I don't want to play with anybody. I want to play by myself. I don't want to deal with other human beings because by working together I was actually worse off. So that's that's a zero sum interaction. What happened in Realm of the Mad God,

and this was a brilliant design piece that was in the early prototype, is when you kill something, everyone on the screen also gets XP. And that's just that's a wonderful, wonderful thing. Right now, I actually benefit from other people being around me and helping me out. And we get this free rider problem. the free rider problem comes up where it's like lots of people working together and one guy's just sitting back and sucking in all the rewards of everybody else. But what we found is that that was actually probably a positive because people would form these huge groups and there's just a pleasure of having groups of people doing something with a shared purpose together and they would run around in these giant trains and there'd be a giant train running down this road together for no other per purpose than like being together and sucking up XP as they killed things together. And so these nonzero sum interactions, these non-zero sum transactions help avoid some of those spoiler and free rider

issues because it doesn't matter if there's if there's lots of free riders if uh everyone's benefiting. The other thing we see in this is escalating costs. Um the Thunderbunny interaction was a relatively lowcost interaction. Um, and as relationships build up, the cost of your interactions become greater. I recently went to dinner with a friend. He was in town for a short period of time, and I I paid $30 for a parking spot downtown. And I was like, ah, $30, you know, that's a lot of money. But then I thought to myself, this is a good friend, and I've I've wanted to see them for for ages. And $30 is meaningless in the context of that friendship. And I get so much more out of that friendship in sort of in terms of social support and shared language and understanding and the ability to like interact with each other and build up our our social lives that that $30 is completely

whatever doesn't matter at all. So you start a relationship without out uh with coffee and then you sort of go from there, right? And it can get so you're owning houses together, you know, it can get uh quite the relationships can go quite far. So this escalating cost is a pattern that you can build into your games as well. Um this is actually a failing of free-to-play games. It's a huge huge failing of free-to-play games. Um in something like Candy Crush Saga, you can indeed ask a friend for a favor and you can ask them for lives and they can return lives and there is a reciprocation loop there, but there's no real way to escalate that reciprocation. there's no way to invest more in the relationship and make the relationship this dense meaningful rich item between the two of you. So there's there's a huge opportunity there for for these richer relationships within those reciprocation loops. We kind of deal with the most shallow shallow version of

it many times. Um and one of the things that that helps that is of course uh rich communication. Um, if you look at the factors that are involved in building relationships in games, there's sort of four big ones. Um, proximity. Are we near each other? You know, do we do we do are are are like, is there a high likelihood that we're going to bump into each other? And that sort of ties into this other one, which is repeated informal encounters. You don't really need to make plans. When we look at our matchmaking systems, do our matchmaking systems enable repeated informal informal encounters? Uh, a lot of times they don't. there's a very limited set of friends that we we interact with or the um there's really one one of the the things that happened in Realm of the Mad God was there was a hub and you could wander through the hub

and like bump into people and see them and see their name and it it ended up you saw the same people over and over and over again and eventually some people some of those people said hi to each other. So there was a social space that was a architectural design decision for the game that ended up causing these opportunities uh these these repeat informal encounters. Um similarity. This is actually a fascinating and scary one. They've they've done a lot of studies on children and how they make friends and who they make friends with. And essentially it ends up being that white rich kids make friends with white rich kids and poor black kids make friends with poor black kids and the the overlap between them is almost nothing. Um like you basically have to like force people together like in for college students you have to put them in the same dorm room in order to race some of these like

uh tendencies for similar similar groups to work to uh create friendships with each other. And this is not a theoretical study. This is like measuring what is and there's all sorts of factors involved in that. Um but it all ties back into some of these economic theories. So similarity is an interesting one because you want groups to cross-pollinate but at the same time one of the strongest ways of getting people to form groups in the first place is to create some sort of similarity in inside that group. Um so I I I find that uh you know just a fascinating topic. Um, and then there's there's one last piece, one last piece that um, we tend to fail on immensely. Just we're just not very good at it. And this sounds so touchyfey, but it seems to be so important in the literature, which is the opportunity to share ideas and feelings. Because one of the ways that we build up relationships is we say,

"Here's who I am. Here's what I think. Here's what I believe." And in doing so, the other person can reciprocate and you can start building up a shared language and understanding of one another. You start building up empathy. A lot of our systems are incredibly mechanical and so we never actually get to the ideas and feelings portion of the reciprocation. Journey, great example of this Journey. uh PlayStation 3 uh two-player um adventure game essentially I guess you would call it. Uh it has reciprocation loops. It is multiplayer. It does a lot of this stuff right. It puts people together in these opportunities to interact with one another. Um however, essentially what you can do is you can beep at one another. And because you can only beep at one another, you can never really explore those like ideas and feelings

like that never really comes across. So we don't see the same level of emergence in something like Journey even though it's an extraordinarily well-designed multiplayer game that we do in some of the other things. Um, interestingly, like everything that I've seen when when the designers talk about it, there was a lot of these negative human emotions that they were trying to remove from the experience because they were going for this very specific type of interactive emotional experience when they built it. And so, a lot of the negative griefing and such was eradicated from the design because it added this certain messiness and unpleasantness to the whole thing. Which brings us to League of Legends. Um, League of Legends has extraordinarily rich feedback if you screw up. Um it is a in highly competitive matches if you're doing something wrong like you step away from your keyboard for a bit or you're in the wrong lane or

whatever it may be there is everyone is immediately on your case and say no you did this wrong or this is the wrong way of doing things and you should do this instead and what are you doing and why are you playing that way and uh it's been a huge huge point of emphasis for the team to try to address that but it's deeply fundamental to the to the game structure. Now, I have a tendency to say this sort of negativity is bad and we want everyone to be peaceful and happy. Yet when I dig into like what makes rich strong communities, this type of drama is actually seems to be essential. This sort of storming that happens when groups come together and create social norms is an essential part of creating a group that works together well. It's almost like the individual has to give up a little something in order to become part of the group, but the trade-off is that the group is stronger because of it.

And I look at uh you know our lovely Google trends for League of Legends and even though it has an immense amount of this these sort of negative emotions in it, it seems to be a very strong rich community and that's a fascinating thing like sometimes the drama is perhaps a good thing and it comes to this idea that norms are violently negotiated social contracts. Humanity interacting with humanity is this messy ugly operation. It is not clean. However, that's how we build civilization. That's how we do it. And when you take language, when you go from the beep which is safe and censored or the like gift in a um in a social game or free-to-play game which is safe and censored and you open it up to chat or conversation, then people will yell at each other. They'll scream at each other. They'll insult each other, but

they'll also negotiate. They'll share feelings. They'll share ideas. They'll share these complex things that end up making the relationships that actually survive much much stronger, which gets us into persistence. Um, Day Z is a fascinating case study if anyone's played that. Um, it has so many of these elements for emergence and people really, really like it and it's explosive because of these elements of emergence. Um yet the forms of governance that come up in it are short-lived and and fragile and they kind of break apart a lot and I I was like why is that so? And in dayc persistence is actually pretty crappy. Um, it's a perma death game and unlike Realm of the Mad God where you keep the same account and it's there's a certain element of like even though you die like people still know your name and that type of thing. There's a social

persistence um in Daisy because the mo you have to move so far in the game. You have to walk around a lot. You don't see people very often. There's low frequency interactions and then once you die you're spawned in another location. So the chances of you encountering the same exact people that you've encountered last time are actually very low. So you're you're the persistence of your relationships is actually surprisingly weak and as a result it damages the emergence of of uh these social structures. This is a geeky one. Um, so once you have all these systems of building relationships, that's great and all, like you still have to answer the question, well, why the heck would you work together? Like what's it why what what in the game system would make you work together? And so there's this idea of collective goods. There's they're very close to public goods as an economic good that increases a benefit to the

community. So, it's like it's almost a tautology, you know, it's it's a it's a good that does exactly what we're looking for it to do. Um, but these are these are reasonably well studied and uh they're things like safety. They're kind of a little abstract. An unabract one would be air. The air in this room is a a collective good. Like we all breathe it, we all benefit from it. If someone starts polluting the air in this room, I don't know what you ate for lunch, but if you start doing that, then it hurts everybody. And we might decide maybe that person can't be part of our group anymore. They're a little too much of a thunder bunny. And we would put them outside of the group because that's not good. They're they're harming our collective good. Um, we can't really it's actually difficult to exclude people from using these these goods. You can create a club good. We can say we are the Thunder Bunny club here and we own the air in

this room. We own the safety in this room. Um and we can say anyone who's not in this room doesn't get to participate in our use of of these these items. But um that's that's sort of a that's an interesting like that's that again we're applying a a structure in order to manipulate that collective good. So, examples of these things in uh or or of of games. Farmville is something that has strong persistence. It has really weak collective goods. You exist in these isolated little shards. You're basically in these pocket universes shut off from the rest of your humanity and there's a very very little reason to interact with the other people. There's these very weak interaction loops between between folks, but in general there you you've got your pocket universe and there's there's very few collective goods involved. Compare that to something like

Kingdoms of Camellet, which are these MMO RTS's. Um, I I go back to like Travian and some of the games like that where you've got an army and you're building up this army and you're investing huge amounts of your real time life in building up these armies and and and cities and then anyone else can come along and just beat the crap out of you. And if they're bigger and stronger than you, they're going to destroy you or knock you down to the stone age and you're like, "Oh, this is horrible." So safety is this thing that ends up becoming the good that people trade in. And how they do that is by creating coalitions or alliances with one another. And by creating an alliance, I can then say, "Hey, look, if you hit me, my alliance is going to hit you back 10 times as hard." And there becomes the safety as a good becomes the thing that people organize around. So these type of games actually have really fascinating social structures because of that that

reason to to work together. Minecraft the blocks in Minecraft are a form of collective good. If you look at some of the largecale building operations like uh um replicating uh Middle Earth in Minecraft, they have these complex rules of how you can edit, where you can edit because there's only so many blocks. And if people start editing the wrong thing in the wrong place, it's it it breaks the whole whole aesthetic system that they're trying to build. And they have verification systems and like timeouts. If you place the block in the wrong spot, you get timeouts and you're not allowed to place blocks on that server for a period of time. um the uh the groups that form have if you look search for um government system and Minecraft or communism and Minecraft or socialist Minecraft any of these systems out there and there's these dozens of servers where like this is communist

Minecraft and this is our Cuba land and here's the little voting systems that we've put in place in order to structure these worlds and It's the this proliferation of many villages, each with their own unique form of government. Most of them are dictatorships, especially that Cuba one. He's like, "Yeah, it's communism and we really believe in communism and I'm the leader of it all." So, what we've talked about is a lot of foundational elements that lead this. All this stuff is necessary in order to get to governance, but we need to stress the system in order to actually cause these dynamics to emerge. Um, words with friends, multiplayer, small groups, it has most of the attributes. We don't see the level of emergence that works because we tend to rely on unofficial norm building when we when we can. And uh, in Words with

Friends, you you go back and forth with someone, you build friendship. There's really no need for anything more complex than that. It's a very intimate, friendly game in that way. But then we start hitting the limits of how many relationships we can hold in our head. Uh Robin Dunar was a fellow who looked at apes and he said, "Look, our brains are only so big. We can only hold so much stuff in our heads." And these little monkeys, they can hold, you know, like 50 50 relationships, grooming relationships in their head. And if we measure the same portions of the human brain, it's like he says 150. We can we can maybe maybe keep 150 relationships in our head at once. And this idea that like these informal reciprocation loops only scale so far is hugely important. Um these group dynamics they break down when groups get large. um below certain numbers and certain complexity uh informal norms work. Above

that we need to augment them. We need to use formal formal rules and like official rules to say look I can't spend the time negotiating with you. I just need you to do it this way. Um so it tends to max out like if you get your group up to 150 you'll start to see this stuff. This happens a lot lower. Like I've seen it happen with groups of like 15 or 20. So it really depends on like how tight the interactions are and how important the task is. So as the task gets complex, the need for organization and cooperation between various group members becomes becomes higher and you start to stress the group more and those informal uh informal uh relationships start breaking down. Um so this is this is a key point and it's it's a it's um government is a social technology. It is a technology

that we're implementing in order to solve broken things. We want relationships. We enjoy them. We enjoy working together. We have these big goals that we want to accomplish and we implement the techniques of government and governance in order to get there when the the the group is too big or our organization is too noisy or the task is too complex. So governance emerges as solutions to problems. So you have to have that foundation of building relationships and people wanting to work together before you can actually get governance. So it's not a magical thing that this stuff happens. You seed the relationship rule uh systems and then you stress them and then the community reacts by forming by adopting these these social technologies. So this is how government patches our

broken um social relationships. When relationships start to break down, we start to create group identities. Uh these are things that help us um identify better. Like if I know that you're someone who is uh you know, someone who's into freeto play or someone who uh maybe you have a big shirt that says Thunder Bunny on it and I'm a Thunder Bunny as well. and I look at you and I see that label, then I know there's a certain amount of expectation in the cost of our interaction. The risk of being rejected is less and I can actually start to remove some of those costs and scale my group better, bigger. So, you often see this at GDC. It's amazing how many people come into GDC and fracture into these smaller and smaller subgroups in order to find their their own group identity. So group groups are this official groups are this sort of way of of getting around like I don't know who you are and

by giving you a label I can know who you are at a low cost. There's an Onion article that was um stereotypes make my lives my life easier and it was a horribly funny Onion article and it was all about all the ways that stereotypes are beneficial and it was very tongue andcheek. Um, but at the same time, there's a reason why we adopt these labels. Um, they are a reaction to the messiness and complexity and the lack of intimate relationships in the world. Um, in the Twitch versus uh Twitch plays Pokemon, you saw people splitting up into uh these near religious groups of people who supported um anarchy versus democracy as the way to play the game. and they had icons and they would like slur the other people and they would get into these long battles about which of their icons was more was better than the other and they would try to recruit each other and it was it was fascinating but it was it was very much watching this group emerge

this group labeling and emergence in in uh in the wild. So uh I want you all to label yourself at the moment. This is a horrible thing. Um and also I just want to wake you up. Uh can you can you just help me out here and say like let's feel that we're all part of the same group. I know we come from different places. So at the count of three, we are all thunder bunnies. One, two, three. We are all thunder bunnies. Oh god. Once more. We are all thunder bunnies. Do you feel it? Do you feel like like the person next to you is a thunder bunny. Everyone in this room is a thunder bunny. And that's that's something that no one else outside of this room shares. There's there's boundaries that we formed. And the boundaries that we formed are because we share this experience. We have a shared experience. We have a shared expectation of like, hey, these were all people in this free-to-play talk and they all have a common common

understanding of some of the terms. And when I say Thunderbunny, you can all recall that same shared experience. And if you were to talk about it later, if you were talk to talk about this experience with a stranger, you might be it might be awkward. It might be strange. You might hesitate. But if you saw some the person next to you and you recognized them and they and you said, "Oh, remember wasn't that strange this thunder bunny thing that we just went through?" Then now you've got this now you you've eased the communication between the two of you because of that group relationship. Um we also create um official norms, official rules. So the social the implicit stuff fails to work. So we say no no no this is how you should interact with each other. Um a lot of games use code to enforce Lawrence Les code is law. We tend to um we tend to use really harsh dictatorial means of enforcing laws. players don't necessarily have that, so they rely on these more like

guild rules and things like that. The the classic dragon kill points for dividing up loot in MMOs was an example of an official rule that the groups created. Uh we then execute rules. Um we're really good at um you know making these things happen and forcing people to do stuff. Um, players don't necessarily have the ability to enforce rules with law. So, they tend to create these I think of them as protection rackets or mafias or they're these they're these things that there's these groups that exercise power in these ways that are extraordinarily grassroots. Um uh in World of Tanks they have um thousands of these clans and they participate in something called uh clan wars which is the strategic end game. And in general it works quite well. However, they end up doing these um almost like um these these mafiaesque ps with one

another where they'll say like, "Hey, we'll go and agree to attack this person together or if you do that, maybe you're going to get hurt." You know, that type of thing. And uh it it actually um it works quite well as as a way of exercising power. And often just the threat of power in these systems is enough to enforce the rules. So, what they've had to do is they've had to create a council for people to um report these things and negotiate these things ahead of time. They they have this giant Skype channel where what they do is they take one representative from every clan and they just have this mad chat room of people talking about issues and and comments and complaints and that allows the leaders of the clan to create this smaller social group on top of this thing. So we've got hierarchy now to sort out some of these issues in a way that the little guys can actually have some control over the bigger mafias. Wonderful

stuff. Um we can enforce official rules, we can detect rule breakers, we can punish them. Um we can like as developers we often are looking for certain patterns and then we go and we change the code. uh um othering and ostracization are the are the big things that tend to happen with unofficial groups. They tend to say uh we are going to shun this person. Shunning is a very very common technique for uh enforcing official rules. Uh they they'll go and say you are not part of this group. I don't want you to be part of this group. That type of thing. And these are age-old techniques and we see them happening in you know you can see them happening in tribal villages. You can see them happening, you know, in unofficial groups in modern society. You can also see them happening in our games. Uh, rebellions. Uh, this was a game called um, Night Online. And, uh, the developers did something that people

didn't like. So, um, hundreds upon hundreds of, uh, of people found that they could turn into a um, this dinosaur creature. And there they they found this super rare item that allowed them to turn into a dinosaur. And uh it ended up having a larger hitbox than their normal player character. So entire cities would be filled with these giant 30- foot tall dinosaur creatures. Um and it would block everything. And uh they were it was it was their pacifist pro protest their peaceful protest against the uh the changes that were happening in the game. And uh so what they did is they acted like a dictator. The developers said look we have absolute power over this world. What we're going to do is we're going to essentially re unleash the hounds. And so they created these um

attackers and spawned them in the center of these peaceful protesters and a stampede occurred. So the players were desperately trying to run out of the way, but they were these giant 30 foot tall dinosaurs and they couldn't fit through arches and they couldn't do anything and it was just a complete massacre and there was bloodshed everywhere. virtual blood. Um, and then what was interesting about this is uh, everyone loved it because it meant that the gods were listening and they actually like ended up addressing the issues and everyone had a good time. Um, and it was actually a reciprocation loop between the players and the developers. even though you can put it in these very violent terms, that's not how everyone saw it. So quite

quite interesting. Um the overall thing that I want to talk emphasize here is that this is Lord of the Flies. This is unofficial governance. You notice I haven't gone into long laborious discussions of voting systems or legislative structures or moderation policies. Each of those is a talk onto itself and is super specialized. A lot of these games get by by setting out this bare minimum and then letting the players build this civilization and society on top of that. Um players when we talk about social technologies um players are the grassroots adopters of social technologies. This stuff comes from the players and that is that magical generative element that ends up making these games last as long as they do and becoming as engaging as they do. So then the question for me is well

what the heck does that leave the designers where where where what's my role in all of this? And uh Richard Bartle, who was the co-creator of the first online MUD, um he had this idea that um designers create this physical reality that then players then build these complex social structures upon. And because we're operating at a level of he had these this idea of levels of physics like the players operate within the context of the physical world of the game but developers operate outside of that that reality of the game and they can actually change the physical laws of the universe. Um, and so we need to consider ourselves we're not really in the business of making governments like in some to some degree like that's something that's actually better left to the players because we that gives them their agency and their their mastery and and this large shared purpose by

building these giant social edififices. and they do a really good job of it and it's messy and it's awkward yet it's very effective and evolving and rich. And maybe our job as game designers are more to be these gods that set the physics. Yet, I am a god that likes poking things. I like messing with stuff. And what I've noticed is that it's really you see the physics and the players make the society, but there's a relationship between the players and the gods, the developers of the world that's equally important. And what we can do is we can sort of we can we can create that foundation and then poke it periodically and adjust and garden and and enhance whatever is there. Um,

so to summarize, this is the formula. This is the simple formula for creating governance in your games, for creating these emergent systems that are more than a content treadmill that slowly declines over time as you try to squeeze more out of that rock. So, first embrace and encourage the messy reality of people building society with other people. Like, it's going to be awkward and bad things are going to happen, but a lot of good things are going to happen as well. As developers, we create a foundation for human interaction. We need to build into our games rich, fast, strong, highfrequency reciprocation. We also need to build into our games rich communication so that players can actually share ideas and feelings and thoughts about theelves

and build social relationships with one another. Next, we need to create a reason to work together. So shared resources are hugely important. Um opportunities to be more efficient as a group. The utopia of the empowered individual is in some ways a negative if you're trying to build this type of system. It's there's different play styles. So like a mix is is is usually good. But often we tend to as we have this technol libertarian bent. So we tend to focus way too much on the empowered individual and not enough on on the group that is better together. Um we need to then the last piece is even if you have all these elements in here you need to stress those social structures so that the players end up spawning rules and reacting to this

stress social structure in order to form stronger groups. There's this wonderful study about what is the biggest factor in the formation of civilization when they they there's these scientists and what they do is they basically take these little tiny bots and they add rules to the bots and then they run thousands of iterations of the tiny little bots until civilization emerges which I think is a wonderful way to spend your time. and they were getting these things that replicated European civilization about 20%. They'd put in all these factors and stuff and they'd say, "Well, how do we can we simulate European the rise the rise of European civilization given all these factors in our tiny little bots and they'd get to the point where it replicates maybe 20% of it and then they added in war. They added in violence and war and then just that one change the correlation between their model and what happened in reality

rose to 80%. And it's this idea of stressing stressing those social groups and bonds. Stressing that iterated game that people are playing when they create relationships in order to force them to how do we create these technologies, these social technologies to work better together. And so once you have all those pieces, um, you get this sort of explosion of dynamics and that's exciting and that's something we can engineer. That's not it's not lightning in a bottle that you can never really capture. It's something we can actually engineer and build our games around. Um, and as as game developers, maybe we prune, we cultivate, but ultimately the players in the communities like are this massive generative source of value. And that is what keeps our games alive long term. That's all I

got. Um there's there's evaluations that are going to be emailed to you. Fill them out. Um, I don't have a system where if you have less than a five, you just email me. But if you want to email me, you can do that. Um, if you have any questions, uh, come up to the mic and, uh, be clear to, uh, be sure to state your, uh, question clearly. Um, and I'm happy to, uh, mumble on. uh connectivity is there anything that it can do to create some form of governance? The question is can a single player game

create this this sort of emergence? Um I have seen zero evidence of that. Um, now that's perhaps too strong. Um, I have seen communities who treat like the Spalunky community would be an example of this where they've treated a single player game as this shared experience and the reciprocation loops happen outside of the game. They happen either in Twitter or in small local communities. So once you can if you end up pushing the community outside of the game um then you can get some of these dynamics but you tend not to get super strong governance in the same way. It's a different it's a different style of of of thing. It's not they're not reacting directly to the game. They're sort of uh building their own experience outside of that. You see this in forums actually. So, a lot of times forums are

a a system of of social governance that has almost nothing to do with the actual game that you're you're playing. Um, it's a game that people are playing outside of the game. So, that that can happen. Thank you. Uh, hi. Uh, thanks for the talk. Uh I also have a qu question about uh do you think you need the group the strong groups need to be hostile and exclusive to other groups in order to have this strong community or is there some way to I don't I don't think groups need to be hostile. I think throughout human civilization there's strong evidence that we tend to be hostile. So there is a tendency due to limited economic resources tend to tend tendency due to the way we exercise power that it is the blunt tool that is highly effective and you see it over and over and over again. I think there's an art to creating games that don't have that. I personally tend

to focus on zero sum cooperative games. So by doing so you actually remove a lot of that tension that tends to cause those hugely hostile uh groups. So I think you can I think you can tune and engineer and create social norms that prevent some of the more egregious uh instances of that but it's a lot of work and we're actually we seem to be operating against human nature. Okay. Thanks. Yeah. Um thanks for the talk. Uh at what point do you as a designer like would set up restrictions or rules for limit limiting governance in order to maintain structure and reduce like a barrier to entry? Seems like a lot of governance it's hard to get into the game once they've built up because you feel like an outsider. Um so how would you go about dealing with those issues? And how did you come up with the Thunder Bunny? The Thunder Bunny came to me in the shower one day. I'm like Thunder Bunny

is the coolest word ever. That's the best place for ideas. Yeah. Um the um so so could you what was the first part of the question? Like yeah sorry. Yeah. Um so at what point as a designer do you set up restrictions or rules for limiting governance? Yes. Um so so both this is this is a fundamental issue. The the big the big question here is like when you when this explosion of of dynamic behaviors start happening and governments start happening and people start forming these these groups, groups tend to have boundaries and walls and walls tend to keep people out. So um it's actually like a big problem. So what you do is you you encourage groups to recruit. Um, and by doing so, you can actually put the the onus on the groups to bring people into them because all these groups have churn and people leave

and and they want to become still stay strong. And you see this like a lot of these games even though they're incredibly complex, what ends up happening is that the way that people are brought into them is through organic channels as opposed to say something like paid advertising or so. because people are so invested in the game that they're willing to either sit down with a friend or or bring them into it. Um, and so the the big spikes you don't necessarily always get or like people tend to if they come into it without an introduction, they may bounce off of it, but the people who come in through those organic channels end up being a very very strong participant in the community. Hey, thanks. That's good. All right. Thank you so much.