Zyngapocalypse Now (And What Comes Next?)

Hollow Men

Two and a half years ago I wrote an article for Gamasutra that kickstarted my career in consulting and caused a huge stir inside the walls of various social game companies. Its title was “Zynga and the End of the Beginning“.

I essentially said that the main problem in social games was that the product was almost identical across all providers, and that social game makers had trapped themselves into thinking that it had to be so. I said that this had led them to treat the market as akin to fast food, as a necessary commodity rather than a quality relationship, and that this meant they were heavily dependent on new users and their sense of novelty. New users would eventually decline and, if the products didn’t get start to become genuinely sexy, then they would eventually stall.

In fact my theory was that social games could slide into the same death spiral that Atari did in 1983. In this model, rather than rejecting one game in favour of another, the market declares a plague on all their houses and stops buying into the platform as a whole. The Atari example is the most dramatic, but the same thing has also happened in other markets such as interactive TV. It will also probably happen to handheld gaming soon.

Was I right? While some of my numbers proved too conservative, the essential points remain true. I thought Zynga would miss the opportunity to spend incoming investment on building better products and instead inflate the kind of product and business model that they already had. This is precisely what happened, with the company expanding headcount massively through acquisitions and new studios, but pretty much repeating the same type of game over and over.

Poker remains the steadiest Zynga game by far because it has a game dynamic that works. Almost all its other games are reliant on content, and pacing. While some social game designers continue to insist that this is a good thing, its the users who ultimately decide if it is. So without constant and heavy promotion, social games tend to fade away pretty quickly. For all of the development effort that gets put in to making new games, makers like Zynga seem to mostly just push existing users around (which leads to wondering whether they really need to make new games at all), stealing from Peter to pay Paul. And occasionally trying to buy a bulk of news users to add to the mix (which is what the Draw Something purchase was all about).

Long story short, the fundamental problem remains that the business model of social games is hollow because the value it provides is poor. Different providers compete against each other to look good (not unlike most other game sectors), but equivalent value is not the same thing as actual value. That’s the same trap that Atari fell into.

Thinking and acting equivalently leads to a case of “When Plan A doesn’t work, go to Plan A”, and Zynga is not alone in this. EA Playfish’s original games are all dying off, to be replaced by EA branded games which struggle to stay popular. Playdom is rarely heard of any more, and its once-vaunted Social City game has dropped to a mere 40k MAU. Digital Chocolate is in fast retreat, and Wooga is merely holding steady.

Even while Facebook is nearing on that elusive billionth user, social games as a share of overall network usage is way down and Zynga only accounts for 15% of Facebook revenue. Of course there are extenuating circumstances, from the lack of access to viral channels to platform considerations, available real estate and so on.

However those are always the reasons that the tactically-minded reach for. Yes, access and channels and funnels and all that stuff matter, but what matters more is strategy. With the right strategy they can all be overcome and success can be found, but with only a middling strategy based on formulaic product it’s tough to make a difference. As with the supermarket shelf, when your product looks the same as everyone else’s and tells the same marketing story, you are just another jar of pickles.

The obsession with tactics is directly related to the obsession with metrics, and this leads to a culture which devalues original thinking. Social games have the exact same problem as network television in that respect, in that there are far too many quants running the show, demanding numeric proof for decisions. Quants understand little to nothing of why players play games, and reduce everything to the kind of extrinsically motivated decisions that Daniel Pink described in Drive, and all of the problems that go with them. Day to day numbers govern everything, so the only permissible decisions are the ones which hurt or help those numbers on any given day. Timidity rules.

And timidity is precisely the problem.

Social Games, Generation Two

Zynga’s whole model (and that of its competitors) is pretty similar to operating a gaming portal. The model’s main differences are the game’s ability to insert itself into players’ social streams, to charge on a freemium basis, and to maintain the save states of players’ games on remote servers. These are the foundations of what we might call the first generation of social games, much as the basic console, cartridge and joystick comprised the first generation of home video games from Atari.

The sunset years are at hand for the first generation. Just as happened with many portals, now is the time when the revenue engine gets milked and executives cash out. Then come the mass layoffs, studio closures and trade sales (As a side note: If you work in a major social game studio, now is the time to think about moving on before the job market becomes flooded.) For Zynga there will be probably be calls to shutter many of its studios in the name of efficiency and refocusing. This will probably then be followed by talks of acquisition, perhaps by Marissa Mayer’s Yahoo, at 50 cents on the IPO dollar.

The second generation needs to be thinking like HBO, not network television. It takes research and prototyping time to develop good game dynamics, but more than that it takes the right technology, talent and faith. This last quality is perhaps most frightening because it pretty much means letting the inmates run the asylum.

Games are no different than any other creative outlet in this respect though, but it’s hard for some people (managers, investors, producers, quants) to accept that. They think that games should be much more like software: predictable, mappable and about process engineering but games and players disagree. You may wish that game design was a process, but it’s an art.

That’s why the second generation of social games is unlikely to come from any of the current big players. They think too small, just as network television executives tend to think too small and need to be shaken out of their equilibrium by an HBO. By which I mean investing in providing real value rather than only playing the equivalent-value game.

While early social game makers were canny in realising that Facebook was about visibility, ever since that tap was turned off their story lacked a second act. (“When Plan A doesn’t work…”) G1 games have become trapped by their own conventional wisdom, trying to pump up the model as much as possible in a classic red ocean frenzy. This has led to bad user experiences, weak track records of innovation and the fear of gameplay. It has also led to a lot of discussions over platform and process, but not product.

G2 cannot afford such obfuscation. It has to tackle the value problem head-on. Therefore the fundamental questions that should drive social games G2 (and if they can’t be answered there will be no G2) are these:

1. Where is the real value?
   
2. How to deepen that value?
3. What is the most immediate way to deliver that value?

Where Is The Value?

What is a social game is for? What do players get out of it? Why should they consider it valuable? In plenty of other arenas players exhibit tribal loyalties and spend thousands of dollars on their pastime. Why must social games remain comparatively unengaged, generic and timid?

Asking “Where is the value?” is not just about whether the game is any good. It’s asking why should the player play your social game as opposed to a downloaded PC game, a DS game, an Xbox game and so on. What’s the unique value that playing online with others in a social graph of some sort offers to players?

Resist the temptation to turn this into a question about you: I have often sat in meetings with well-intentioned designers and executives who have answered this question through projecting their ideas of how players might play in the future (socialising, bringing families together, multiplay), or confusing value for the investor or the company with value for the player (status objects, the opportunity to live a life, purchasing opportunities etcetera). Both are sidestepping the question.

Here are some pointers:

Forget the Multiplayer Future

In every game clans and tribes form, and so it can appear that the future of games is to do with community formation. It is often assumed that greater socialising, greater contact and more synchronous loops are what players want, but it’s not. While communities are loud, they almost never represent more than 1-in-20 of players of any game.

Social games do not bring people together. Most players in fact play them in a largely single-player fashion, making contact purely for reasons of necessity like trading, earning Energy and so on. Many have tried large multiplayer designs, and failed because the players were just not there.

It’s Not Really About Status

Much hay was made back in the day of the value of virtual goods and status, leading to large projects like PS3 Home and Second Life, as well as cosmetic items in many virtual worlds. There was a time, indeed, when many developers were thinking that status items were all they would sell. They missed the point on this by a large margin.

Status is a minority interest in games, and sales of status items tend to be small compared to sale of utility items (in cases where both are available). Most players do not spend that much time on their avatar, do not really care that much about how their virtual house is arranged and when they realise that the rest of the world does not give them social proof, stop buying status items altogether.

This can be confusing to understand because in games like pet simulators it seems as though much is spent on cosmetic items. However it’s important to note that a cosmetic item can also be a utility item (for example: to increase my pet’s happiness I must buy flowers etc). A status item is one that has no utility other than allowing the player to be a little creative or show off.

Living A Life

Virtual worlds are great as ideas, but boring as experiences for most users. They are often build around the notion of community chat, living a virtual life and otherwise just sort of being in a place and time, but most players want more than that. They want a game.

Players play to achieve, to do, to build, to create, to explore, to destroy and to win. They need the game to provide them with a fascinating system which enables them to do all of those things, and usually for the game to also provide an absorbing fiction. This is as true for The Sims as for Skyrim.

What games don’t do particularly well is the whole “living a virtual life” thing. They have boundaries, and players know it, and while they will build a lot time and investment into their Diablo 3 avatar to make it the biggest and toughest, all that activity is once again about more winning. The avatar is supposed to be of use, for the player to self express through play rather than only being pretty. So is their farm, city, amusement park, sports team and so on. You build to have something better to win with.

The game is always a world to have fun (the joy of winning while mastering fair game dynamics), and if it provides this well (comparative to the player’s expertise level) then synergistic effects like community and fan culture and so on gather around it. However you don’t get there by simply building an empty shell and avoiding the value question entirely.

The Real Value

G1 social games are, with only a couple of exceptions, single player games which are free to play and occasionally tax users into communicating or spending to be allowed to keep playing. So where is the value for players in being social?

Aside from being free to play, the answer is advancement. Social contact in the context of games provides real value to players when it substantively helps them to win without tying that up in synchronous loops. In other words, to be worth it the contact needs to get me where I’m going, but without obliging me to turn up to do likewise.

Players overwhelmingly prefer to play their games on their timetable for their fun, and this is why every single successful online game facilitates this. Whether single-player questing or drop-in/drop-out tables as in Poker (this is why poker is still Zynga’s most solid game), it’s all about the self-propelled, self-organised and self-successful player. “Social” simply helps that happen faster, in what we designers are increasingly calling “parallel design”.

So play Journey. Play Realm of the Mad God. Get into a multiplayer server onMinecraft. Notice how they are about cooperation toward advancement? Notice the lack of obligation? Study that.

Deepening Value?

95% of social games use the same game dynamic. That dynamic is a mix of roleplaying rules (levels, experience points, etc) with wait-or-pay rules (energy, building with clicks, paying to shortcut) and guided goals through a repetitive activity (digging, questing, item assemblage). It has had many linear improvements, from the basic zombie-chomping games of old through the sophisticated Ville-structured games of today.

Over the course of their evolution, social game makers have also tended to try and broaden that dynamic with different themes or surface differences in gameplay. So there are monster themes, fashion themes, mafia themes, racer themes, cities, vineyards, castles, sims and so on. Mostly they are the same game (or at least the same type of game) with a variety of different types of set dressing to try and appeal to various audiences.

In the casual boom a similar kind of explosion occurred in hidden object, plate-spinning and action-puzzle games. Likewise there has been a similar evolution path for gestural games, from the simple arm-waving of early Wii games through to bright lights of Just Dance, Zumba Fitness and Kinect Sports. Each is an example of linear improvement and broadening of subject matter, but each also shows a lack of depth. For reasons of audience, technology and interface, each proved limited – and so the sense of value for players declined from delight to linear to threshold features (as in the Kano Model of User Satisfaction).

There comes a point in all genres where linear improvement and broadening subject matter becomes obvious to the audience. Their play brains start to realise that they are seeing the same frames again and again, with the same actions and the same constraints. So they become instantly boring. At that point the entire genre either goes into its sunset years, or a game maker figures out new depth.

“Depth” in this context means that the maker figures out a radical and delightful reinvention of a known dynamic (Halo compared to its contemporaries), or invents/popularises an entirely new dynamic. Depth is sometimes enabled by new technology, sometimes by interface, or by clever design. However it almost never involves adding further depth to a long and broad structure. G2 social games will not simply mean interweaving more complexity into a game like CityVille because that depth will arguably be lost under the already-existing game. It also does not mean trying to find just one more theme, one more spin on the same idea.

Deepening usually requires throwing out a lot of what has come before to get back to what matters. Having identified the value, the G2 social game takes the idea of advancement through obligation-free collaboration… and throws out everything else. Social bars, energy, business model, metrics, even platform and technology go. They are all G1-era trappings, and irrelevant to the true value. Some of them may come back, but none should be automatically regarded as necessary.

Then ask yourself this: Freed of all the trappings and understanding the true value of social games, what could you build? How could you make it deeper before getting too scared by the fear of not being like everyone else. Just how far you could you push parallel advancement with new dynamics like action, strategy, resource management, puzzle solving or anything else?

Now, how to build that?

Delivering Value

There is no law which says that G2 has to happen on Facebook. Nor on iPad or any other platform. It could happen anywhere. Social networking itself has broadened far beyond Facebook in the last couple of years (Twitter, Google+, Path, Instagram, Pinterest…), reminding us that it is a networking function and save state, not a domain. So perhaps G2 social games could happen on Google+ using Google Native Client as its core technology. Why not?

Facebook is probably a bad place for G2 because of the weight of G1 competition. While there is no arguing that it was the platform that kicked the whole thing off, in recent years Facebook has become much heavier, and visibility within it is much tougher to maintain. Arguably for users this was a good thing (remember the days when your notifications were constantly full of Mafia Wars?), but the magic has long gone. So G2 will likely not start with explosive growth in the way that G1 did, however it will probably revolve around much more valuable interest graphs than generic social graphs.

The fact is that most players’ game graphs in their Facebook games are either empty or full of the orphaned accounts of those who stopped playing. The really dedicated players even go outside Facebook to form communities and add each other as friends in order to build more valuable game graphs. This, of course, violates the intent of Facebook, but they do it anyway. Players gonna play.

For G2 to be about true value, the game graph also has to be valuable. Connecting interested strangers produces much more game interaction than limiting it to just friends (such as Monstermind, which doubled its engagement rate in a day by connecting strangers). Players don’t really care about whether they are playing with their friends. They care about playing with others who can help them, and if that happens to be a friend then so much the better.

Roleplaying game fans want to play with other roleplaying game fans, and word game fans likewise. If the first round of social games tried to sift through graphs to find those few nuggets, then G2 needs to construct an interest graph of players (many of whom may be strangers) who like the same kinds of game that they like. Turn that into a low-cost, easily distributed, opt-in network that can plug into any game… And that’s how value delivery works.

Technology may also have a role to play here. The hard limits of Flash have really started to show, particularly in terms of fast responses, uses physics or complex work (such as believable AI). These limits mean that simple casual, sim and rpg games tend to be the mainstay of the platform because they are known to work. Few are the Flash games that really go beyond that.

However there is Unity, which every indie developer I know is busily trying to master. It’s designed for making games, and works on most platforms. Then there is Google Native Client, which enables developers to put games using much more powerful languages into the browser. There is also cloud gaming, which is generally not for games that require split-second responses, but could be very powerful for sim, strategy, massive multiplayer and other games.

In short, G2 social games will probably have very different delivery to G1, like the difference between “software” and “app”.

Conclusion

If Yahoo was “Search, Generation One” then Google was “Search, Generation Two”. The first generation was the one which became cluttered with all manner of complicated ambitions, poor performance and a whole load of “conventional wisdom” which often proved contradictory. Generation Two, on other hand, realised what mattered and delivered just that. A similar shift is what will make “Social Games, Generation Two” real.

If Minecraft were a little more friendly, free to play, a little prettier and a little more easily hooked into social networks maybe that would be it. Or perhaps if Triple Town were a little more parallel. These are examples of games with different dynamics that are simpler than many G1 games, but experimental. One of them has been very successful while the other is more a cult hit. Both show inklings of the future.

It’s from these kinds of roots that a second generation of social games will eventually grow because they are about value for the player. Be ready for the shift.

About Tadhg Kelly