How to Design a Social Game

Back in the day, I was fascinated by Nintendo Power's game ratings system.  Sure, the overall game was awesome or terrible, but the sub-categories on what made it game gold or bargain bin fodder was what really made it interesting.  Here is an example from the now defunct retro magazine:

The magazine broke their ratings into five categories ("Sat." was "Satisfaction"), but each held an equal weight, with all on an equal level.  Clearly this was better than a simple thumbs up or thumbs down, but was fairly arbitrary, and far from scientific or following any logical paradigm.

Another point of reference was the also now defunct retro GamePro, which riffed on the Nintendo Power system. Here is an example from them:

Fast forward to 2006.  I was working for a casual and indie game publisher, evaluating 10-20 games per day, every day, for candidates to publish, help improve then publish, or to fund.  In addition, I was evaluating our own internally developed games, and needing to compare them to games we were funding and publishing.  I needed a system.

I had just read Ken Wilber's seminal "A Brief History of Everything" which included the following "four quadrants" model upon which all of life, science, psychology, art, and culture can be neatly tucked into:

The bare essential of the model, and the book, is that all of life and ideas can fit into hierarchies.  And the key concept about these hierarchies, is that an upper level transcends and includes the lower levels, not transcends and rejects.  To transcend and reject is to dominate.  To transcend and include is to evolve.

It was at this publishing desk that I developed the following four tiered hierarchical system of evaluation for games:

The publisher I worked for was also a game engine developer, and I was contacted by the IGDA Game Accessibility Special Interest Group (est. 2003) about promoting accessibility options for game developers within our game engine.  I grew up with two parents who were social workers that assisted persons with developmental disabilities, so the SIG's mission had special meaning to me, and I promoted their advice for our game engine within the company.  

Developers used our game engine to create experimental games for people who were blind, for example, by utilizing 3D audio to create space to move within.  One button games for persons who were quadriplegic to play using their head or breath, for example, were also created.

Around that time also, the casual games movement just started to take off, with the forming of the Casual Game Association, the Casual Game Conference, and publishers like PopCap and Nexon raking in huge sums.  I immediately pivoted to casual games, drawn by the promise of "games for everyone," which I immediately connected to improved accessibility in order to bring the previously hardcore domain of gaming to the masses (though ironically, having started out as casual arcade games like Pong and Pac-Man).

The intersection of these two meta-industry events led me to make accessibility the foundation of my evaluation pyramid.  Nothing is more important than accessibility.  Nothing.

Accessibility means more than one button or three buttons instead of twelve or 101.  Accessibility means a language the player can understand (what if they are international?).  It means a platform the player can use (sorry Linux, but Grandma is no root sysadmin).  It means a price point the player can access - which hinted at the rise of Free to Play games at a time of $60.00 MSRP's during a long economic downturn.  It means a skill level the player can engage with.  It means a tutorial, and instructions the player can understand.  It means not requiring an over-clocked $2,000 PC rig to play at a decent frame-rate, and using audio drivers that work on all hardware systems.  And yes, it also means having subtitles during voice-overs for people who are deaf and have hearing disabilities, font sizes that are legible to 40 year old eyes with glasses, and buttons that can be clicked that forgive you if you are a few pixels off or "fat finger" the button.  Lastly, it also means basic technical stability.  Accessibility will make or break your game.  If players cannot play your game, they cannot play your game.  You can quote me on that.

Next on the hierarchy we have theme.  After accessibility, the theme is the next most foundational piece of the game, upon which all later art and gameplay are built.  If the theme is one players don't understand, then even if the later art and gameplay are great, players can't get past the theme and overall narrative. Two of the most important aspects of the theme are the story and characters. The theme should at best be a hook with gravity that draws the player in, at worst be neutral as something the player is willing to accept but can comprehend.  

One example of a game with a poor theme is Total Distortion, where the player is a music producer who uses an inter-dimensional transporter to jump to the Distortion Dimension with a million dollars in their pocket while guitar robots attempt to kill them in order to try to produce a new music video - kind of cool, but a concept lost on most, and one that is clearly trying too hard and is over-produced.  By contrast, Guitar Hero's premise is, play a guitar well to become a hero - a theme easily understood by all.

Bad themes can be not just over-developed, but under-developed.  As another classic example, we have Irritating Stick - what is this "stick," and why is its' "irritating" quality fun?  Clearly this title didn't try hard enough with its theme, and the fiction and narrative are under-developed - and no amount of high production quality 2.5D art with explosions of procedural particle effects can save it.

As anecdote, I once asked a game design job candidate to pitch me a sample game he thought would have top 10 appeal as a design test.  His pitch?  Alpaca Farm.  My first question - what the hell is an alpaca?  I had to look it up.  Of course, indies have the luxury of making far-out themes, because they are not necessarily attempting to create a mass market product.  The majority of us in the game industry don't have that luxury.  But even if you're designing for a niche market, if that niche market doesn't "get" your theme, you are lost.

Next on the hierarchy is audio and visual, which builds upon the theme framework with a fleshed-in layer of graphics and sound.

Why is art more foundational than gameplay?  Simply put, if your audio grates, or your art clashes (and not with clans), players' ears/eyes will bleed, and they'll never get far enough along in your level to have that euphoric "ah-ha!" moment where they "get" and become hooked on your gameplay, and subsequently your game.  

In fact, if your loading screen is bad enough, or uses stock models from Poser, players may rage-quit on the spot, knowing it can only go downhill from there.  At its most extreme, if the box art on your game is sufficiently terrible (or your marketing screens), players will never buy the cart (or install the game) to experience the next higher level on the hierarchy which might be the best in the world, the gameplay.  If the art is bad, it reliably follows that the gameplay was crafted with just as little consideration.

Low budget indies are an exception, though infrequently so.  However, with them too there is delineation between art style and production quality, and even low production quality art (low budget art) can have excellent style and reach mass appeal (Osmos, Darwinia, Flappy Bird).

Lastly, at the top of the pyramid, the most complex aspect of the overall product, the one that transcends but includes all prior levels, is game design.  The game design relies on all other foundational levels in order to be scribed - it depends on choices in accessibility, theme, and art, and is built upon those choices.  The game design is the one aspect that can only be felt internally by the player, that must be played to experience.  It cannot be measured by senses like accessibility, it cannot be viewed and heard like art and sound, and it cannot be contemplated and understood intellectually like the theme.  It must be played, and it is only in playing that all the pieces of the multi-faceted jigsaw puzzle intermix into an orchestra of brilliance, or coalesce into a cacophony of chaos.

The multitude of near-infinite game design pieces include the player controls (interaction), goals, moment to moment gameplay, meta-gameplay, pacing, "game feel," game elements such as hero, enemies, power-ups, items to collect, rules of of the game universe (gravity, physics, bullet logic), timing, game world truths, internal game world mechanics, puzzle design, level design, multiplayer design, game economies, element simulations, and all the ephemeral "right brain" things that define a game world but are themselves difficult to define categorically, such as Nintendo Power's variable of "Satisfaction" and GamePro's parallel "Fun Factor." Just as love and faith are impossible to define but real (just ask someone in love or someone with faith), so too are satisfaction and fun - and it is within that nebulous realm that the game design lives and breathes, and subsequently, makes or breaks your game.  Each of these elements must be gotten "just right."  The task of this is nearly impossible, and it is why there are so few "perfect 10" games.  The gameplay must be perfect.  If players don't like playing your game, they won't play your game.  You can quote me on that.

We have reached a hierarchy of categories to evaluate and develop games based upon.  Let's now pivot to social games.

In 2009, when I began designing social games for a major publisher, I learned new areas of design that changed my self-established model.  I modified my model to accommodate social games specifically, which produced the following:

Gameplay is no longer the item of top transcendence.  Now social gameplay has transcended but included mere gameplay.  Social gameplay uses non-social gameplay as its foundation, but goes beyond.

Social gameplay comes in four varieties, as defined by Mildred Parten in the late 1920's at the Institute of Child Development in Minnesota:

Most social games utilize parallel play as their strongest social feature.  Each player has their own identical play space, progressing through the game objectives and levels individually, using the same set pieces and game rules, and there is little to no interaction with others aside from the visiting of other player spaces.

The meta-awareness of the oddity and limitation of using this method as the primary social play style is spoofed in The Simpsons Tapped Out, where friend spaces are identical but different in progression from the player's own, and are explained as being in parallel dimensions (literal parallel play narratively!).  Typically in these social games, players can only visit another player, then leave a small bonus or message.  The interaction is a bare minimum.  This can evolve.

There are huge opportunities in social game design to strengthen parallel play, and to introduce and create new mechanics supporting onlooker play (spectating, as in poker and some sports games), associative play (gifting and item trading), and cooperative play (guild and clan play, most often found in MMORPG's, but which has been distilled in social game designs, notably Clash of Clans' clan play).

The next tier on the hierarchy, which transcends but includes social play and depends upon it, is that of viral features, essential for the spread of the game to friends, family, co-workers, real world neighbors, social network "friends," water cooler break buddies, and even cats, if the game is accessible enough for them to play.  Simply put, a social game without a viral feature is like a virus without a viral feature - it won't spread. You can quote me on that.

Getting viral features right means players aren't even aware of them as viral features, just as game features.  This isn't because it's posting to Facebook without the user's permission.  It's because the player should feel like the game is enabling them to take screenshots of their accomplishments and their story through the game world for them, to then brag to their friends about, and/or is allowing them to share bonuses and gifts and buffs with friends naturally.

Done well, viral features are totally transparent, and completely integrated into the full game experience, making them a value-add, not a cheap marketing trick. Done wrong, they're a last minute spam addition, tacked on with joyless and superglue-like adhesion.  At their most horrific, they feel like a digital stalker, watching and following your every move in the game, insisting that everything will be better if you just listen and post what it wants on your social networks, and give permission to whatever it asks for.  Overzealous viral features don't transcend and include lower categories on the hierarchy, they transcend and reject, or dominate them.  In other words, they take you out of the immersion of the game theme and narrative; they disturb the core fun of the gameplay; they create ugly 2D popups in an otherwise beautifully rendered artistic scene; they relegate the person being marketed to with only onlooker social play and nothing more, no deep offer to "jump into the ball game" and "help save the game."

When executed in this domineering fashion, these viral features may have short term positive metrics "proving" their flawlessness, but their impact on long term retention will always be that of the impact of a stalker on a stalking victim.  And it is always a sad thing, when a game turns into a thing that a player must put up their defenses and "guard" against, to defend against an unwitting unintended answer on a prompt that will then turn them into a spam-zombie for their real life relationships, when a game should be a vehicle for escapism and fun.

At one GDC talk years ago, I heard a ballsy commentator cocksurely declare, "Quality is our viral."  Though part of a speech and written to be quotable, I loved it and whole-heartedly agree.  Although additional discrete and discreet viral features are necessary and not necessarily evil, do not overlook the organic installs that come from one player simply telling another player about how awesome a given social game is, and asking them verbally to play and friend them.  This is constantly underestimated in social game design.  This is how Rovio rose to then-prominence with Angry Birds taking over the world, without even heavy or deep viral features. This is also how Tetris took over Russia - by being passed, person to person, physical disk copy to physical disk copy.  The pieces just fell into place.

The final category to design and evaluate social games on is that of monetization. Why is monetization at the top of the social game hierarchy?  To repeat the theme, because monetization transcends but includes all previous elements on the social game design pyramid.  Monetization relies on all previous elements of the pyramid.  There is no product aspect, from accessibility to art, from social features to gameplay, that monetization design does not include, and yet it has its own set of complex rules and best practices onto itself.  This is also why those managing the monetization in a social game, the Product Managers (PM's), need to be the most literate, experienced, and multi-disciplinary team members on the game team.  

If the prices of IAP's are inaccessible (accessibility, the foundation of the pyramid), even for just the economy of a given geographic region, then even if all other aspects of the game are Grade A, the game will fail (in that region).  If the fiction behind why you must buy things in the game is convoluted (theme), the game will fail.  If the art for premium IAP items in the game is of no better visual or sound quality than free items in the game, those items will fail (audio visual).  If the gameplay loops were designed independently of monetization considerations, with monetization tacked on later instead of deeply integrated in parallel with the product's development (game design), the game will fail.  If the game's monetization abuses social play with too obvious of pay to win, instead of those same paid pieces used to share with friends or guild members, the game will fail. If the game doesn't socially surface purchases as raising the status of the one who made the purchase, ideally by offering to share a portion of the purchased bonus with one or more friends, the game will fail to take advantage of a substantial viral opportunity in the intersection of monetization and viral spread.  And lastly, if the monetization design, independent of the game design and having its own laws and rules, economy and pacing, quirks and exceptions, science and art, is poorly designed, poorly monitored, or poorly nurtured into a slow long term optimization and growth, the game will fail.  If that happens, the road you took led you there, it was no accident but a direct and correct byproduct of the incorrect path you took. Similarly, if your monetization design is immaculate, but any lower level on the hierarchy is substantively broken by itself (accessibility, art, gameplay), the game will fail.

This is why there are not only so few "Perfect 10" games, but so few break-out top grossing social games that hit the top lists and stay there, and stay there, and stay there.  When the hierarchy is done right, a game can hit exponential success, and camp there for the long term.  But when even one level on the hierarchy is broken, the game will never hit the top 10, and will be lucky to hit the top 100, and will struggle to stay even there.

Lastly, I want to highlight this view of the hierarchy of social game design as seen through the eyes a new user funnel, in this case mobile social and a major app store like Apple's:

You can see how the progression of each social game design hierarchy element not only comes right after the other in serial time progression from app awareness to in-app purchase, but how each social game design hierarchy element transcends but includes all prior foundational elements before it.

CONCLUSION

This hierarchy gives us not only the categories of importance, but their order and relationship, as well as relative weight.

Nothing is more important than accessibility.

And yet, if you get the top of the pyramid wrong, you can also make nothing.

How to design a social game then, is to methodically consider each of these elements, in their order of importance, and design them from both a product design and game design standpoint, in this order.

When weighing a decision on whether to include Feature A or Feature B, consider how each feature will affect this hierarchy, and which element on the hierarchy is more important.  For example, something that improves art but hurts the client engine performance (accessibility) is not a good trade, if the hit on the client performance is significant.  Or if a monetization feature guts a game design feature instead of complimenting it, that is also not a good trade.  It is, however, ok for an accessibility decision to limit art further up the hierarchy, or for a game design decision to limit a monetization feature further up.  This is how it should work.

Design your game using each of these steps, one at a time, and slowly work your way up the pyramid, remembering to transcend but include all lower steps each step of the way.  Any given higher step is not "more important" and therefore entitled to dominate over any lower steps.  Instead it should utilize but include lower steps, which are its foundation.

Although these steps are fashioned in a logical hierarchy, they are in "game team" terms a circle where all are of equal importance.  The best monetization will fail if QA is not exceptional to catch technical bugs before they arise; similarly the best QA means nothing if the monetization is not using the latest in metrics analyzation techniques and multi-variate testing to optimize IAP's and track sources and sinks.

I dislike too abstract and theoretical of game design articles - they are all too easy to write, and typically have little practical take-away.  The take-away intended here is to design your social game, using the pyramid model above, from the bottom consideration up, and to have each higher consideration transcend but include each of its lower foundational considerations.  Doing so will yield you an optimally designed product, which in turn will do the most important thing - entertain people and make them happy.  So happy that they will pay for your game, and do so willingly - voting with their hearts using their wallets.

I hope you have enjoyed reading my theory on social game design, and I hope it helps you design better games for more people to enjoy.

Josh