Our initial mission statement at the outset of the project workgroup was to “Generate a list of mechanics and/or dynamics which induce more civilized play in individuals.”  In practice, this mission evolved during the process into something more closely resembling “Generate a list of principles and best practices which enable designers to create game systems that better facilitate and incent civil play.”

The following are summaries of thought emerging from our discussion and research, as well as some design takeaways which might be applied to a game system in order to achieve an elevated level of civil play.

What then do we mean by “more civilized play”?

From the outset, we chose to avoid spiraling into debates about the nature or essence of civility.  We instead established the scope and tone of the discussion in such a way as to define civility (quite informally) as “the extent to which each player’s expectation about fairness and positivity of experience is met or exceeded.”  This might include anything from a player’s initial sense of welcome into the game, to her sense of support from other players, to her own efforts (consciously or subconsciously) to “give back” to the game by herself espousing the positive principles and behaviors that were exemplified for her in-game.

Historically, game design's approach to dealing with the "soft" issues of player culture and civilized (or not) behavior has been largely hands-off. There seems to be general agreement that the responsibilities of the game designer does not necessarily extend into the role of social supervisor.

We disagree, profoundly. Games are now spawning communities that dwarf many historical civilizations. If the designer is not responsible for the community that develops from the rules they have created, then who is? Recent innovation in social responsibility systems has demonstrated conclusively that game design can affect player culture in a positive way.

If we designers can have this effect, then it seems to us that we must.

Perhaps the most poisonous counter to this proposal is the idea that player culture as it is today is simply "how it is", and therefore "how it always will be", or "how it should be". This refusal to embrace any attempt to improve the life of others is dismissible purely on moral grounds, but there are strong, practical reasons that, we believe, will drive designers into this role of systemic anthropologists. In simple terms, history has demonstrated that games that engender strong social communities around them are generally more long-lasting and more successful.

A civilized player culture will likely result in a stronger social community for its game. Therefore, it is the designer's role to exert every effort to produce systems that will result in such a culture.

Our mission, as described above, is to “generate a list of principles and best practices which enable designers to create game systems that better facilitate and incent civil play.”

Within the broad scope of that goal, we want to identify principles and best practices that:

1. Apply broadly to all game cultures. This is in recognition of the fact that there is no "best" culture globally. A competitive combat game would not be best served by the same type of culture as a cooperative village sim.
2. Apply to all players, not just to one specific type. We want to lift the top, middle, and bottom simultaneously, empowering community heroes as well as making the average player’s interactions better.
3. Are not 'prescriptive solutions' of the form "put this mechanic in your game in order to improve your player culture". This would be dangerous and foolhardy, as the effect of any particular mechanic on a player culture is highly sensitive to its surrounding context.

We set out to not address the following issues:

1. Punitive techniques. We acknowledge the shithead troll problem. However, we feel this is a largely known-and-solvable problem, despite the fact that known solutions aren’t employed nearly enough.
2. Real life identity. We felt that the debate between real life and anonymous community identities was a distraction. Anonymity may not encourage negative behavior given the permanence of online data. Additionally, real life identities seem most effective when employed within a strong, real-life social graph such as Facebook.
3. Psychological lenses. We went into this using alpha personalities as an example of something to design for. After numerous social experiments on the team itself, they decided it was dangerous to characterize players that way, since their roles often change based on social context.
4. Reinventing the wheel. There are a lot of games and teams doing good work towards this goal already, even if the industry at large is not. The team has tried to distill those best practices into this document.
5. Gamifying society. While real world techniques can inform these best practices, we cannot assume they are interchangeable. The social structure of a game is an opt-in society, not an opt-out society.

While exploring this topic, we arrived at several principles--a set of cultural expectations--we felt designers can use as a springboard toward promoting civilized play.

Like it or not, game designers share more than a few traits with dictators. They have almost god-like powers over their creation, they benefit from a predominantly one-way relationship with a large mass of people, and they set the rules by which others have to behave. Within the process of creating a multiplayer game designers are confronted with unavoidable questions that, no matter how small, always contain at least traces of moral and political value judgments. The designer is, therefore, a sort of benevolent dictator who sets the rules that constrains the behaviors of the citizens in his digital metropolis.

Perhaps it is better to think of the relationship between designer and player as similar to the relationship between parent and child, and that the designer always exists on some point on a spectrum of paternalism between himself (or herself) and the player. The question is not whether there is a paternalistic stance involved, but to what degree the designer is paternalistic. For a useful distinction along that spectrum, we can turn to political theorists, who make a common distinction between "hard" and "soft" paternalism (e.g. Sunstein and Thaler’s 2009 book Nudge, which has popularized the discussion recently).

In general, hard paternalism is the position that the paternalist has a very robust or comprehensive conception of how the other person should behave or live his life. Most of the world’s major religions adopt this sort of hard paternalism, defining in often minute detail how the faithful should behave, speak, think or feel. These sorts of attitudes are usually highly codified, containing lists of explicit rules that are sometimes so irrefutable that they have to be etched into stone tablets.

In contrast, soft paternalism does not seek to proscribe a full account to how to live life, but instead defers to the individual's own stated set of goals and desires and begins from there. The soft paternalist does not seek to constrain an individual's ultimate goals, but instead looks at the individual's stated desires and seeks to constrain their behavior in a way that will most likely help achieve those desires. We think it is this more constrained version of paternalism that the designer can safely operate in. That is, the when determining the range of choices or the amount of freedom to give to the player, the designer ought to adopt this more thin or soft paternalism and encourage those sorts of behaviors that the player himself would like identify as desirable within the game or game's community.

Additionally, careful use of propaganda can communicate intended social norms without restricting choice. While propaganda can be seen as something negative and jingoistic, it needs to be thought of as just another tool. One real world example of propaganda is a musician seeding her tip jar. When setting the jar out, the musician often drop a few of her own dollars into the jar to help establish the expected norms of behavior. Passersby see the money in the jar and understand that they’re intended to tip the musician. Even if they don’t tip, the norms have been communicated to them.

Designers must take care when using propaganda to influence a game’s culture. A player might think well of a game with an active forum filled with friendly people, but that opinion would sour if she finds out 90% of the posts are from the developers posing as players. Any established norms may be thrown out the window, and the players may be less likely to trust any other information the game provides.

To that end, propaganda is best (and safest) used when it is based in truth and more about subtly nudging a player as opposed to completely misleading them. Consider the musician and her tip jar, and how a couple dollars can say a great deal.

Empathy is a cornerstone of civility. It is also a natural feeling that people have. We don't have to make people into something else, we can expect some if not most people to be helpful if an opportunity arises. Moreover, empathy breeds empathy. If most people believe that a community is "about" helping out people whose game experience is sucky at the moment, then indeed most people will do just that.

Empathy only works, though, if you know how the other person feels. In meatspace, body language tells us. How often have we run into someone crying and asked "what's wrong" and they say "nothing." In current games, though, only the "nothing" is visible. So other players miss all this otherwise humanizing, empathy-inducing information about the Other.

Creating civil communities therefore requires better insight into players’ emotional states, especially when it is negative so that they can share any frustrations or bad experiences they may be having. This information should naturally trigger empathy, which breeds help, which breeds a general tone of civility.

You can’t empathize without information that humanizes.

Altruistic acts within a game community have the potential to offer a reciprocal benefit: firstly, the player in need of assistance has their experience of the game and their perception of the environment lifted, and secondly, the player performing the altruistic act leaves the interaction knowing that their effort was appreciated and (ideally) that it addressed a real need.

We believe that altruistic acts represent a facet of civil game behavior, and that, if a player is provided both the visibility of opportunities for altruism and the empowerment to act on that information in a meaningful way, then the overall rate of altruistic acts will increase, reinforcing the positive perception of the community. In essence, this system would be designed to encourage players to bring out their “inner hero”, even if they might normally assume a more passive role.

Furthermore, we believe that we can incentivize civil play acts (including, but not limited to altruism) by designing rewards for positive social behavior at the culture-formative moments of the game. These rewards would, by design, taper off over time so as to not be a gameable system in and of themselves.

When players join an online gaming community, frequently they must agree to follow a Code of Conduct drafted by the publisher. As in a civil law system, these documents make explicit what constitutes unacceptable behavior, provide guidelines for conditionally allowed behaviors and, perhaps most importantly, spell out terms of enforcement. Also, similar to civil laws, Codes of Conduct are seldom engaged with by most of those the rules apply to, but their existence makes for a better environment as a whole.

What Codes of Conduct fail to do is inspire civic responsibility. This is where we as game designers can make a positive impact. Just as game systems can train players for empathy by spotlighting individuals emotional states, systems can plant the seeds of citizenship and stewardship. Games can give players challenges larger than the individual or the guild to foster the notion of civic duty. Moderated communities can recognize and celebrate the exemplars of citizenship outside of the context of measured achievements. While not every player will aspire to give back to her gaming community, when players believe that they can be more than a list of stats--that they have the opportunity to make a difference--their bonds with the gameworld and fellow citizens are strengthened.

In competition, player interaction will often devolve into negative social behaviors if left unchecked. Players stake their ego on an outcome that reflects their skill, and when paired with a social environment like a competition, there is a natural tendency to evaluate both the winner’s and the loser’s social worth based on the results. The typical “sore loser” tries to save face by passing off their responsibility for the loss by making excuses, blaming external conditions or lashing out. Conversely, the “bad winner” tends gloat over their victory and belittle their opponent’s performance.

Sportsmanship is a social ethos that attempts to manage the social situation around skill-based competition by giving both the winner and the loser clear procedures to minimize the impact to their egos and maintain civility. Values such as fairness, mutual respect, and fellowship with competitors are usually stated explicitly in codes of conduct. Those values are then reinforced through sportsmanship rituals such as sumo rituals bowing before a bout or little league baseball teams shaking hands after a game.

In the interest of inducing sportsmanship through mechanics rather than explicitly stating a code of conduct to players, modeling those social rituals from real-world sports is a good start.

We brainstormed a range of mechanics & dynamics--including choice architecture, communication methods, and even propaganda--with these principles in mind. The most actionable practices appear first. Promising practices that require further design investigation appear afterward.

Actionable Practices

Onboard Social Norms

Introduce new players to the desired social norms through gameplay and onboarding, rather than relegating that to a EULA.

For example, completing events in Guild Wars 2 often requires groups of players working together. The game broadcasts the events to the players, creating the context of playing together without explicitly telling them they must play together.

If you intend to use extrinsic rewards as part of the onboarding process, dial them back after onboarding the player.

Allowing other players to reward good behavior reinforces the social norms, reinforces the bond between those players, and makes that reward more intrinsic to the social interaction.

For example, players can bestow “honor” upon another player in League of Legends for a making the match a positive experience. Categories of honor include “helpful, friendly, teamwork,” and “honorable opponent.”

Calculate a player’s reputation, display it to others, and imbue it with value.

For example, a game can make unique status symbols such as titles or items available to users with a high reputation.

Prime players with examples of expected (and desired) behavior. While a direct message works well, a counter works better.

For example, you can display a counter in a matchmaking interface that states “5000 players were resurrected by their teammates today.”

Additionally, you can set a minimum display number to give players the impression of strong participation in this behavior. Just be sure that the player doesn’t find out.

Give players the ability to express their needs or emotional state.

Examples include the “Medic!” voice over call for help in Team Fortress 2, as well as emotes common in many MMOs.

In a system with discrete choices, frame those choices by making the positive choice the default. That forces players to opt-out of making the positive choice.

The classic non-game example of this is making organ donation registration opt-out instead of opt-in. In other words, agreeing to donate is the default choice.

Make positive choices easier to perform in the user interface.

For example, the “Medic!” button in Team Fortress 2 is bound to a primary key on the keyboard, while Guild Wars 2 displays downed players on the game map, making it easier to find them.

Likewise, undesired choices can have increased friction, such as sorting negative emotes to the bottom of an emote list.

Simply limiting the types of social interactions users can have with one another.

The most recent example of this is Journey. It only allows one method of communication: a musical tone or ping that blends seamlessly into the soundscape.

On its own, this practice isn’t consistent with our approach to paternalism. However, if the interaction constraint is consistent with the game’s fantasy, it won’t seem paternalistic to the player.

Another approach that’s been bandied about the industry is enabling social interactions based on a player’s reputation and experience within the game. This seems more heavy-handed, however.

Create opportunities during or after gameplay within which players can interact with one another. This seems especially useful after competitive matches.

This can be accomplished with simple text macros (“Good Game!”) or something more complex. In a poker match, for example, the winning player often buys a round of drinks to the losing players. Zynga Poker ties this ritual into in-game rewards. A player who has just won a huge pot can buy a round of virtual gifts for the other players at the table as a show of fellowship.

Empowering all players to perform desired behaviors without having to plan for it in advance.

For example, all player classes in Guild Wars 2 have the ability to heal other players. This stands in contrast to other MMOs that offer healing abilities only to specific classes.

Add persistence to a player’s sum total of social interactions, while highlighting civil interactions from notable players within the game’s community. This fosters a culture where “heroes” arise, providing inspiration for other players and validation of civil behavior put into practice.

For example, the “Trusted Medics of the Wasteland” organization provides free medical assistance to players facing permadeath. They have grown to hero-like status in that game. The controversial Shard Vigil in Asheron’s Call is another example. In this case, the developers created a memorial to the participants within the game itself!

Enable more experienced players to directly assist and onboard new players, while earning rewards for doing so.

This topic requires further discussion. MMOs have done this extensively in the past. DayZ’s Trusted Mentors of the Wasteland is another example to research.

The following ideas aren’t fully formed. They have a lot of potential however.

Enable players to shift between different or orthogonal social interaction modes to give them greater opportunity become invested in the game.  By supporting two or more different modes of interaction, players are more likely to find a medium or forum for expression and therefore less likely to feel disenfranchised. 

In real life, this is analogous to a meeting in which participants may talk over each other, but yield to a raised hand. Imagine then a cooperative board game that moves between an “open” mode (in which players kibitz freely) and a more “closed” mode (in which each player in turn states her opinion or decision).

Enable players to “freeze” a multiplayer gameplay scenario if they’re not having a good time. Upon freezing the game, they could seek a hint or moderation for a poor social interaction. This idea obviously carries with it technical challenges that require further thought.

Display a map of the social dynamics in your world. Other players or moderators could see areas where folks aren’t having fun and go help them out.

A close analog of this is the social media monitoring tool Radian6. It generates a word/tag cloud so that moderators know what the trending forum topics are at any point.

A lot of these techniques are easy to integrate. Get cracking!

In order to focus our efforts, we declined to address certain topics in this paper, such as punitive techniques or prescriptive solutions per genre. That didn’t prevent us from discussing these topics, nor does it prevent us from sharing these practices in the Appendix.

Punishment for negative behavior can encompass many techniques, ranging from a change in permissions to banishment from the game.

For example, if a certain number of people mute or ignore someone, they default to being muted or ignored.  Alternately, they could be restricted to only communicating with friends.

Making an example of players with negative behavior illustrates that there are real consequences, while simultaneously reinforcing the social norms.

For example, publicly calling out and defusing troll-like behavior, or announcing player banishment from the community.

Bring transparency to moderation and even invite community participation.

This is exemplified by the Tribunal System in League of Legends. Players are allowed to review flagged chat logs and vote on whether the interaction was consistent with that game’s social norms. They are rewarded for participation only if they vote with the majority.

Foster player bonding by giving all players seemingly epic challenges larger than the individual or guild/clan could achieve on their own. Leverage these to provide opportunities for players to “rise to the occasion” by performing positive civil acts for the rest of the community.

            Numerous examples from MMOs.

Question whether your game needs leaderboards. The very nature of scorekeeping makes a game more competitive. Furthermore, it sends a message to all players that the game values this behavior, even if it wasn’t designed with that in mind.

Consider the relationship between new and veteran players in your game. New players often make mistakes in a pressured situation. And if that situation leads to a more experienced player losing, the more experienced player may grow to resent the new players. This is especially an issue when the gameplay has one solution.

Tune your game’s resource economy so that there aren’t simply winners and losers in any given interaction between players. True scarcity will drive players to hoard, cheat, and so on. Instead, allow for shared or orthogonal rewards to mitigate the issue.

section 4

next section

select a section:
1. Introduction
2. Workgroup Reports Overview
3. Training for the Game Design Olympics
4. Toward a More Civilized Teabagging
5. Opening the Kimono
6. Divine Design
7. Crux
8. Design Feedback Loops for Successful Independent Productions
9. Game Jam 2012
10. Schedule & Sponsors