Sponsored By

Unsocial 'Social' Games

In this feature, veteran game designer Greg Costikyan unpacks whether social games are truly social or they are not -- and having dissected the form, then leaps in with some suggestions on how to make the games more rewarding for players and developers.

Greg Costikyan, Blogger

May 24, 2011

22 Min Read

[Veteran game designer Greg Costikyan unpacks whether social games are truly social or they are not -- and having dissected the form, then leaps in with some suggestions on how to make the games more rewarding for players and developers.]

Three years ago, my friend Eric Goldberg called me up and asked me if I would be interested in working on a "social game" for a little Web 2.0 company in California he was consulting to.

A social game. Hm. That did sound interesting. I had fantasies of limited-duration, closed room live-action role playing, or perhaps multiplayer boardgames that fostered intense player interaction, or maybe something like an urban game, suitable for the Come Out and Play Festival, with players engaging with each other for extended periods of time.

Or perhaps some merger of online forum and gameplay, some elaboration of the competitive wordgames we used to play on Genie and The Well and Echo, showing off for one's peers and preening in a social environment.

Yes, pushing the social element of gameplay could be very fruitful, an obvious and exciting extension of the capabilities of the ars ludorum. And someone might actually pay me for this?

Well, no. Suspiciously, I asked Eric what he meant by a "social" game.

After quite a bit of blather, I realized what he was talking about, and cut to the essence. "Ah," I said. "I see what you mean. A game that is played on a social network. Is there anything actually social about it?"

You could hear the shrug on the phone. Would I like the introduction?

Sure. I needed the work.

When I was 13, and an enthusiastic fan of board wargames such as those published by SPI and Avalon Hill, I spent some time looking over a list of wargames divided into different categories -- Napoleonic and World War II, and so on -- thinking to myself that I wanted to find a style of game to make my one, to study more seriously and become expert on. One category popped out at me, and was what I decided to specialize in: multiplayer games.

In digital, we think of "multiplayer" as meaning anything that isn't soloplay; in wargaming, however, multiplayer meant "a game for more than two people," since most wargames are struggles between two opposing sides. I chose the category that encompassed games like Diplomacy and Kingmaker. They struck me as far more interesting than two-player wargames, because the complexity of interplayer dynamics produces far more variability than in head-to-head games. Negotiation, alliances, trades, and simply reading other players became important; it wasn't all about system and the mastery thereof.

I spent many long hours negotiating, allying, backstabbing, and learning to deal with others.

When I was 14, Dungeons & Dragons appeared, an inherently multiplayer but cooperative game set in a fantasy world not unlike those of the novels I loved to read. Its appearance spawned many long hours learning how to coordinate, cooperate, persuade, do "improv" in an almost theatrical sense, and work to shape stories cooperatively with others.

Dungeons & Dragons, in all likelihood, saved me from being a studious, depressed loner, ultimately making a somewhat charming adult out of a shy adolescent. It taught me to be a social being -- not surely from any intent of Gygax & Arneson's, but from the nature of its gameplay.

For many years, I dismissed digital games as devoid of merit, partly because of their lack of intellectual and narrative seriousness, but more importantly because of their inherently solitary nature. It was M.U.L.E. -- Dani Bunten's landmark multiplayer game for the Atari 800 -- that showed me that digital games, too, could be highly social.

I moved early into designing and playing games online -- even before the internet was opened to non-academic users, on the commercial online services -- because online games redressed the greatest flaw of digital games: their inherently single-player nature.

And in recent years, I've become fascinated with the rise of LARPs, indie RPGs, and story games, because they place socialization among the players, improv, and the assumption of character, front and center in play.

Social games -- correctly defined -- are important; and games that have hooks for socialization are, I think, our best bet for the creation of true art in this form.

It's a pity that "social games" are so unsocial.

One of the fundamental contradictions of the human condition is that we are simultaneously individuals and social beings. We are each lost in our own heads, with only the limited bandwidth afforded by expression and language to allow us to understand each other. We typically strive to achieve our own needs and desires, often at the expense of others; we are atomistic individuals, striving for our own benefit in a Randian sense.

And yet we cannot grow without the nurture provided by our parents, are driven by our needs to find sexual and emotional partners, have built civilizations that depend utterly on cooperation and exchange, and are most often happy in the companionship of friends.

We are both free and members of societies, dispassionate observers of our surroundings and passionate members of groups, freethinkers and partisans, competitors and cooperators.

We prize individual freedom, and also community. We solo, and join guilds. Americans, in particular, make much of the importance of individual liberty -- and yet to accomplish almost anything, we need to enlist our fellows in a common endeavor.

Though ultimately we live, and die, alone, imprisoned in our individual skulls, we are inherently social beings.

And just as we most often find happiness through others -- our partners and children, our friends, and the extended praise and approbation of the many -- so too the best and most affecting game experiences we can have are those that involve others.

The fiero of a World of Warcraft raid successfully accomplished with your guildies; the laughter produced by sardonic gameplay in an RPG; the emotional impact of a jeepform or story game produced through unexpected improvisation with others; the banter and play of power dynamics in a Eurogame; even the table talk and insights into character you gain through the play of a conventional card game with friends -- these are experiences to be prized.

Even for single-player PC or console games, if you try to recall the most compelling experiences you've had playing them, I think you'll find that those experiences derive not so much from triumph over mastery of a system -- but from a moment of insight into the mind of the designer, a grasp of what the game is trying to achieve, an insight into the game's subtext.

That moment of epiphany, when you finally understand; that sense of engaging with the creative product of others; that sense of being part of a cultural conversation, of something that you can discuss and debate with your friends -- that, perhaps, is at least as important and earning a score. In these cases, the social nature of your experience is indirect, since you experience it alone; but the experience is part of a larger, and continuing, social conversation among the players and creators of games.

If you were the last person on Earth, would you play games? And if so, would it do anything other than to make you sad with the realization of what has been lost?

The social nature of games -- indeed, of any form of art -- is part of what makes them compelling; and games that strike deep into social connections are often the most compelling of all.

"Social games," then -- that is, games that strive particularly to make and exploit social connections between players -- have enormous potential, as forms of art.

It's too bad, then, that few so-called social games are remotely social.

The first commercially successful "social games" were social network role playing games (SNRPGs), like the various Mafia and Vampire-themed games. Like more conventional digital RPGs, they are games in which you control a single character, with the primarily objective to level up by completing tasks of one kind or another. They strip this down to a bare minimum -- missions are accomplished by a single click, leveling up opens up new ones, and "energy," which recharges slowly, is the main constraint on advancement, since you may accomplish only so many missions in a period of time.

In addition to mission and level advancement, SNRPGs allow you to attack other players. In an attack, a character stat (that can be improved with level) is added to the attack value of equipment, and compared to that of the defender; the winner gains EP and money, the loser loses money and health.

An additional fillip -- critical to the game's virality -- is that players may use network invites to ask friends to join their clan (or mob, or what have you), and when in combat, your value is increased by the values of your friends. Thus, the more clanmates you have, the more powerful you are, the faster you can advance -- and the less likely you are to suffer from the attacks of others.

SNRPGs are, in fact, completely solitaire in nature, except for the ability to attack others, and the ability to have clan mates. The idea of the "clan," however, has no real meaning; each player's clan list is entirely separate from every other player's.

If I join "your clan," this does not mean that I join also with other members of your clan; I can belong to any number of players' clans simultaneously, and the only relevance this has is to increase of combat power of these people (and my own). The "clan" is merely a game conceit; it has no organization, no mechanism for interaction or planning, no common assets. It is nothing like an MMO Guild; it's just a list.

Games of this type do allow you to type in messages that are seen by your clan mates, which sometimes produces weird conversational lacunae, since your clan mate may be responding to a message from his clan mate who is not your clan mate, so you see only one side of a conversation.

The main way players use this feature is to list friends of theirs who are looking for more clan mates, so you can increase your power by friending and adding them to your clan; I now have more than three hundred social network "friends" who I do not, in fact, know, and have scant interest in knowing. Their only purpose is to make me competitive in SNRPG combat.

You can argue, in fact, that SNRPGs are antisocial in nature, since the only real interaction with other players is attacking them.

At present, most social network games are, in essence, light sim/tycoon style games. You control a map on which you may build structures, place decorations, plant crops, and so on; some of the items you place produce game money, which you accumulate by clicking on the item in question after some period of time.

Sometimes, but not always, you have an avatar, who moves around, performing the "work" you request when you click on an item. You level up over time, and this gives you access to more placeable items and other content. Typically, there are badges, collectibles, and quests you may accomplish.

In other words, the core gameplay is inherently solitaire in nature; you do not cooperate with others in building your business/city/what have you, but are doing so on your own, each player in his own atomistic world.

However, social tycoon games provide many hooks for inter-player interaction, of a kind. One critical system is "gifting"; players are urged to send others free gifts, which cost the recipient nothing but provide some modest game benefit to the recipient.

In particular, the construction of critical items in a game are often gated by the requirement to have some number of a particular item that can only be received as a gift, giving players a motivation to request such gifts -- and, of course, to spend real money to eliminate the restriction if they have a small number of friends or are simply impatient.

Players may, of course, brag about their accomplishments, which serves the purpose of spamming social network communication channels with posts about the game, thereby helping to attract new players (a practice still encouraged by social tycoon games, despite the fact that Facebook now makes these posts invisible to non-players, thereby nerfing their viral function).

And players may "visit" each other's maps, performing some limited number of tasks on each friend's map. This provides some sense that other players exist in the same game world, although in truth each map is wholly independent and solitary.

If you look at the interplayer communication fostered by social tycoon games, you will see that every possible communication, every game action that a player may take relative to another player, exists solely to serve the purposes of the developers. Each communication action is designed to do one of three things: attract new players (virality), encourage players to return (retention), or encourage purchase (monetization).

Player interaction has modest, if any, impact on game progress; no impact on game outcomes; no or virtually no consequences for the players involved.

If SNRPGs are, truthfully, antisocial, social tycoon games are, truthfully, asocial; the interplayer interactions they foster are not, in fact, social in nature, do nothing to build friendships or enmity, and provide scant, if any, sense of connection to other people. They're like soloing in an MMO; you're aware of the presence of others, but they are largely irrelevant to your play.

Developers of social games have clearly given great thought to using the social graph to foster player acquisition, retention, and monetization; but as far as I can see, no thought whatsoever has given to the use of player connections to foster interesting gameplay. It's all about the money, and not at all about the socialization.

The peculiarity of this is that social networks are actually far better suited than most online environments to fostering social gameplay. Messaging and chat are built into the system, and need not be separately implemented by developers; but more importantly, the social graph allows players to interact with people who are their actual friends.

Most other online environments make that difficult. For example, MMOs make it hard to play with true friends, because of their segmented nature. I may learn that you are a WoW player, and want to play with you, but find that you play on a different server from me; I'd have to start over at level 1 to play with you.

Even if I and a group of friends all decide to start out on the same server, the reality is that we play on different schedules and with different levels of intensity, so even all starting at level 1, we will soon be of divergent power, and MMOs gate content by level. If you are level 5 and I am level 20, there's no easy way for us to play together, since progress for my character requires me to be in areas of the world that will kill your character quickly.

On a social network, it is easy to discover what games your friends play, and it would be trivial to implement systems to allow people to play with each other -- reserving games for groups of friends, say, or designing systems to accommodate players of diverse power who are network friends.

Online sites that allow quick, pick-up and play games are problematic as well; it's easy to go to a site like Pogo.com or Days of Wonder and play with strangers, but this is far less interesting and enjoyable than playing with friends. The messaging provided by social networks, and the ability to share content with others, means it should be far easier to implement games that allow play with genuine friends than on other sites.

In short, developers have learned how to use the social graph to rake in the bucks, but not how to use it to foster gameplay that is actually social.

What does "actually social" gameplay look like? It's not very mysterious. The following is not exhaustive, but here are at least a few of the means by which social play can be encouraged.


In a team sport, you must coordinate with your fellow team members. It is certainly beneficial if you are an excellent athlete yourself, but if, say, in basketball, you do not remain aware of where your teammates are, and be prepared to pass to them (or receive the ball from them), you will not be a great contributor to your team.

While sport is sufficiently faced-paced that communication beyond hand signals is rare, the team that coordinates effectively will, more often than not, triumph over the one that does not.

Much the same is true in any sort of team-based activity; in a team shooter such as Counter-Strike, team coordination counts at least as much as individual FPS skill. In a World of Warcraft instance, the boss can only be overcome by excellent coordination -- and the ability to confer before a battle, together with voice chat, makes interplayer communication vital.

In a social game context, teams have great value as well; they are potentially a strong player-retention mechanism, because players will return to the game, not wishing to let their teammates down. Why SNRPG "clans" are implemented in such a meaningless, atomistic way is a mystery.


In any multiplayer game where players may form alliances, diplomacy becomes vital. In the classic boardgame Diplomacy, for example, all players are of equal strength and can rarely overcome a single opponent alone; alliance formation is critical. Moreover, there are no joint victories in the game (though ties are possible), so there is a strong incentive to backstab your allies at just the right moment.

The result is that players talk constantly; you need to persuade your allies to remain on board, you need to lull them into faith in your loyalty as you plot their demise, you need to try to gauge the likelihood that they are about to betray you -- and you need to try to persuade your enemy's allies that their best interest lies in switching sides. Diplomacy is a game with fair strategic depth, and some degree of tactical finesse -- but its heart likes in negotiation, and the silver-tongued will more often win than the player who has memorized opening strategies.

To allow diplomacy, however, you do need a game in which players can provide material assistance to each other, and also damage their foes; you cannot have a system of "solitaire games played concurrently," as in social tycoon games. It is true that some (such as City of Wonder) allow "attacks" against others (as, of course, do SNRPGs) -- but they do not enable the formation of alliances, the alignment of strategy, or joint actions of any kind.

Negotiated Trade

For trading to be important in a game, it needs to contain a degree of asymmetry. In Edgar Cayce's classic cardgame Pit, for instance, the random distribution of cards encourages players to attempt to assemble monopolies in different types of cards. In most games that foster trade, there are several different kinds of resources, and players are likely to have access to some but not all, providing an incentive to trade with those who have the resources you lack.

Trade, like diplomacy, gives players a reason to engage with each other; to discuss strategy, to find those who have complementary things to trade, to negotiate price, perhaps to establish enduring trade relationships. Social tycoon games have an element of this -- for instance, it is common to allow one set of players to gift certain items and another set a different group of items, encouraging swaps between them. But this is ad hoc, and pointed mainly at promoting Facebook requests, without any real attempt to foster market behavior or player discussion.

Resource Competition

In many games, resource competition is a key element of player interaction. In an RTS played in multiplayer mode, for instance, battles often arise around critical resource extraction sites.

But "resource competition" can be over things more abstract than literal game resources. For instance, in "action selection" boardgames such as Agricola or Puerto Rico, a limited number of action types may occur each turn, with some actions passed over.

In Agricola, one player's choice of an action means others must choose different ones; in Puerto Rico, all players may perform selected actions, but must select the ones they think will benefit them most and others least. In Steam, there is a very strong first-mover advantage, which the game balances by auctioning off the right to go first each turn, forcing players to make a difficult calculation over the costs and benefits of what they bid.

Whenever resources, opportunities, or other aspects are constrained, with not all available to all players at optimal levels, players are forced to engage with each other to try to obtain what they need, while denying critical aspects to others.


Games with explicit hierarchical systems are rare, but the concept is particularly well suited to a social graph environment. As one example, in the play-by-mail game Renaissance, a new player begins with few resources, and can progress much more quickly in the game if he swears allegiance to a more powerful liegelord, or joins an existing trading house as a subordinate. In Slobbovia, a Diplomacy variant, players who control more than a few provinces are required to turn control of some over to a subcommander -- who can declare independence if not satisfied with his commander's play.

Hierarchies are useful in acculturating newbies, in fostering continued communication up and down the hierarchy, and in fostering team play, but with the additional advantage that 'teams' can be more fluid than in games with explicit teams, since players may change hierarchies, and move up or down within their existing hierarchy.


In a non-digital role playing game, much of the pleasure of play derives from playing a role: from making decisions in the spirit of your character, from a vicarious experience of what it is like to be a person living in an imaginary universe, from speaking and acting "in character."

In some types of role playing games -- in LARPs, in many indie "narrativist" RPGs, in story games, and in jeepforms -- the line between 'game' and 'theater' is blurred to the point that the distinction between between "role playing" and "acting" is moot.

Role playing, in this case, is not merely about vicarious experience, however; it's also about showing off for, and entertaining your friends. Saying interesting things, acting in unpredictable (but character-consistent) ways, and improvising with your friends becomes critical.

Performative play like this need not be as freeform as in tabletop RPGs, however; as an example, one of the most popular games on the old Sierra Network (a pre-Internet commercial online service) was called Acrophobia.

The game is very simple; a group of players are served a random string of characters. Each must devise a sentence in which the first characters of the words of the sentence together produce the random string they were served.

When all players have entered a sentence, they are revealed, and the players then vote on which they like best; after several rounds, the player with the most cumulative votes is the winner.

In this case, the way to win is to come up with the sentences your fellow players will think are wittiest and most amusing; despite the restrictive, formalist nature of the game, it is performative at its essence.

Performative play is utterly social in nature, and possible only in multiplayer games; it is also something to which online games have always been well suited. The existence of a social graph, making it easier to play with actual friends instead of strangers, should foster performative play more effectively than other online environments. And yet not only do social games not foster performance of any kind, they eschew the kinds of communication -- messages (other than the preformatted) and chat -- that could support performance.


In recent months, the number of monthly active users for social games has dropped a bit, and there's been some discussion of the idea that the social game boom may have passed its peak. No good thing lasts forever, and unquestionably the success of social games owes some debt to their novelty, and novelty wears off quickly. In my opinion, this is a temporary matter, largely resulting from the sameness of available games; one new game exploiting a play pattern social gamers haven't seen before will be sufficient to restore interest in the field.

Where that play pattern will come from is a matter of conjecture, but to my mine one place to look is obvious: in fostering gameplay that is itself social, in a meaningful sense.

As social media natives, social games need to become truly social. This means more than exploiting the social graph for business model purposes; it means figuring out how to use the social graph to provide unique and compelling gameplay.

Social games must become actually social.

Read more about:


About the Author(s)

Greg Costikyan


Greg Costikyan has designed more than 30 commercially published board, role playing, computer, online, social, and mobile games, including five Origins Awards winners (ludography at www.costik.com/ludograf.html); is an inductee into the Adventure Gaming Hall of Fame; and is the recipient of the GDC Maverick Award for his tireless promotion of independent games. At present, he is a freelance game designer, and also runs Play This Thing!, a review site for indie games. He is also the author of numerous articles on games, game design, game industry business issues, and of four published science fiction novels.

Daily news, dev blogs, and stories from Game Developer straight to your inbox

You May Also Like