Saturday, August 13, 2005

User Content: Working with players to efficiently create profitable games

I recently have run across a couple of articles that have piqued my interest. The first is title 'Future Imperfect' by Jim Rossignol and discusses how user created content in games results in a deeper, more meaningful player experience. Most importantly, he coins the term 'the wikification of games', which is a money catch phrase if I have ever heard one.

The second article was in "Act of Mod: Building Sid Meier's Civilization IV for Customization" by Mustafa Thamer in the August issue of Game Developer Magazine. The article discusses how Civ IV was built from the ground up using XML, Lua and Python to make the core of the game accessible and modifiable by players.

Games are slowly emerging from the land of custom coding into the world of data driven, standards-based application development. I love this trend for a number of reasons including faster development, more artist involvement in the design process and increased opportunities for innovation. What is the future of user created content and how will it effect our profession?

Ultimately I’m interested in how user content can be leveraged to more efficiently create highly profitable games. (I’m a pirate at heart…arrh.)

The slippery slope of data-driven development
The opportunity for user created content was born due to the solely selfish activities of smart developers. Successfull games are about content, not code. They were looking at ways of speeding up their development process and began creating data-driven game engines.

The classic data-driven game starts with a small engineering team who builds out a solid infrastructure. Often the engine is based on middleware and most of the initial programming work involves ensuring the correct data driven architecture and interfaces are in place.
A hand off occurs and the designers and the scripters begin developing the game play in a modular fashion using scripts and easy to manipulate data objects. With high level scripting capabilities found in Python or Lua, teams can build interactive elements several times more quickly than if they were developing the same functionality in C or C++.

Prototyping and play balancing, both of which have historically been programming heavy activities, can now happen at a much more rapid pace. Iterative development becomes highly effective and in the best situation you end up with more polished games with shorter development schedules

Artist friendly tools
Using a standard format such as XML means that it is easier to create simple tools that give artists access to your data. At the most basic level, an artist can edit human readable text. With a bit of .NET mojo, simple editing tools can be quickly built for almost any task. Tools are certainly additional effort, but in today’s production oriented world where artists outnumber programmers 3 to 1 on many teams, the benefits are economically obvious.

The result is that game design becomes more like creating an HTML page than programming in assembly or C++. The wikification of games isn’t there quite yet, but it isn’t far off.

With these new authoring options comes a shift in the power of who can create games. Historically small games creation has been restricted to people with both a good sense of game design and the ability to program. Such creatures exist, but they are quite rare. If one out of 100 game developers is a competent designer and 1 out of 50 is a genius programmer, we’ve got a delightful 1 out of 5000 chance of stumbling across a renaissance man who can do both. These are bogus numbers, but you get the point.

I like to think of this as disqualifying 90% of potential Olympic athletes because they can’t calculate differential surfaces in their head.

By reducing the hardcore programming to an infrastructure support role and shifting the central creative capabilities into tools that can be used by almost anyone, you start to end the industry’s implicit discrimination based on technological skill. Make no mistake; you still a solid theoretical grounding that is necessary to be a competent designer. However by removing arbitrary technical barriers we increase the chances of stumbling upon that one person who was born to design great games.

Once you are able to provide tools for technically incompetent artists, it is a relatively small step to toss the tools over the wall to the consumers. A method for improving the game developer's efficiency becomes the means of empowering players to create user content.

A wonderful explosion of crap
When you reduce the cost of production and open up a highly technical creative activity to a relatively unskilled base, what happens? Let me propose a law:

If the cost of authoring traditionally highly technical creative content is reduced by an order of magnitude, there will be at least an order of magnitude increase in the amount of crap produced.
We saw it with desk top publishing. We saw it with early websites. Goodness knows we are seeing it with blogs. And we are seeing it with games. Have you honestly downloaded a game mod or flash game lately? Sewage collected from the festering depths of a bulimic convention’s septic system is often more appealing. Some of the worst manages to be both morally offensive and artistically bankrupt. Most of it simply demonstrates an utter lack of talent or imagination. And the shear amount of the bad stuff is growing at an impressive rate.

I take great delight in the early rumblings of the democratization of game design. If you can wade through the muck, there are some great designs to be experienced. The crème de la crème of this outpouring has already produced some of my favorite game play experiences (Counter Strike and More Strange Adventures in Infinite Space come to mind). In the long run, more game designs ultimately mean more good game designs even if they are just a small percentage of the total.

The good news is that the authoring tools are only going to become more friendly and the engines more conducive to modification. There are three big forces, two of which we’ve discussed that point towards the commoditization of game development technology

  • Improved production efficiency
  • High quality product gained from leveraging a broader population of design talent
  • And last, but certainly not least, strong economic incentives
Leveraging this creative explosion for profit
Let's dig into this last force in a bit more detail. Early supporters of mod-friendly data driven games were a bit like the underpants gnomes from South Park. Their business model was the epitome of good natured benevolence but of dubious financial benefit.
  1. Allow mods
  2. ???
  3. Profit!
History has proven many of these early pioneers correct and their apparently generous gesture to the fans resulted in substantial financial rewards. The following are some of the reasons why this tactic worked.

  • Player created content equals more content. More content means that players keep playing for a longer period of time and makes the game a better value.
  • In many cases, the appeal of the game was extended into market niches beyond the original intent of the designers. Even if the game developer gets their design wrong, it is likely the community will fix many of their mistakes. (Some of the balancing efforts being performed by the Age of Wonders community comes to mind)
  • A strong mod community acts as a persistent word-of-mouth marketing organization for the title.
  • The final result was more sales and a longer shelf life.
All this becomes justification enough for the more Machiavellian game developers out there to follow a similar path for the sake of profit. If making modable games reduces risk and increases profit, then smart developers will make modable games.

The first generation of user content
All these benefits are potential benefits dependent on quality execution. Poor implementation of a user content system results in fewer benefits to the game developer. The first generation of mods in particular suffered from many of the problems that plague current indie games:

  • Poor distribution channels: Often the distribution channels were fan websites that only reached a small fraction of the primarily single player, offline player base.
  • Technically demanding installation procedures: It took more than one or two clicks to use most mods. This was too much for many players.
  • Lack of common filtering systems: With so many mods available, it was difficult to know which ones were worth a player’s time. The rating and review bodies were often fragmented and held little authority with the target audience.
One of the titles that I have personal interest in is Age of Wonder. They have a robust mod and mapmaking archive online, a strong community site and great tools. Recently, they polled the online users and asked if they used mods. 64% had never tried to install a mod. Only 20% successfully used mods with any frequency. This sample is heavily biased towards mod usage since we know that 80 to 90% of single player users do not participate in online community sites. This patterns exists a number of other games.

So with first generation games that support user content, a small fraction of the total user base actually takes advantage of new user content. There are better ways.

The second generation of user content
We are beginning to see the second generation of games that attempt to facilitate the financial benefits of open, data-driven engines. They are building in services that simplify access to additional user created content.

  • Official stores that allow the purchase of user content: Steam allows the ability to download and purchase of commercial mods whenever you play Half Life 2 or Counter Strike: Source.
  • Portals for easier trading of user content: Sims 2 includes a content portal for the download of user content.
  • In game systems for managing mods: Unreal offers mini-mods called mutators, that can be easily turned on or off inside a user friendly interface.
The third generation of user content
These are wonderful steps in the right direction, but they are just the start. I suspect that we’ll see a third generation of games that promote player created content taking a hint from titles like Spore and Second Life.

In this third generation, the review system is built into the core game mechanics in the form of a standard purchasing system. Installation is transparent and happens in the background. Distribution is also transparent and automated such that you never even notice that you are getting new content. As an added bonus, the authoring of new content becomes a fundamental activity of the game play.

The third generation offers an important advantage over 1st and 2nd generation titles. In earlier attempts at promoting user content, the content was always something outside of the core game experience. The user was required to step outside the game environment, enter into a radically different user interface and context. That transition, as simple as it might sound to the tech heads of the world, is enough to prevent over 90% of users from ever partaking in user content.

The third generation of titles treats user content as an integral part of the game experience. User content isn’t an extra; it is how you play the game. With this shift, I would easily expect 100% of players to partake of user content.

Consider the benefits of this third generation of mod friendly games.

  • The game developer provides an inspirational sketch of a game and a well stocked tool box.
  • Then, over many months of player activity, the game becomes fleshed out with content that the players desire.
  • Players build the majority of game content and the game developers monetize their results.
Go underpants gnomes, go.

The implications of third generation user content systems on game services
The more that game developers support and encourage the creation of user content, the more they play the role of a service provider and not a packaged goods producer. We need to stop thinking about a game as simply a game engine plus a fixed set of content. Instead, the developer is hosting and maintains a common game development platform that supports an ever growing library of gaming content, much of which is user generated.

The game developer sits in a unique position in the value chain. They are the only group with the power to provide proper integration between player-created content and the end user experience. A third generation design like Spore is only possible if the game developers owns, maintains and controls the financial, mechanical and social aspects of content distribution.

In fact, providing a robust service for the creation, distribution and filtering of user created content becomes an important design activity in its own right.

Creation Mechanics
Earlier I covered the concept that in order for a game to make money, the game designer must consider the financial mechanics of the title. For a developer to create a fully realized 3rd generation user content friendly game, they must also consider the creation mechanics of the title.

Creation mechanics are similar to traditional game mechanics in that they can be described using a traditional risk / reward sequence
  • Action: The player creates a desirable piece of new content. The content is rated by existing filtering systems.
  • Reward: The player is lavished with either social kudos or in-game rewards
  • Risk: The player builds something that no one likes and their efforts are wasted.
Creation mechanics, like financial mechanics, are optional. Only a small percentage of the player community will take advantage of the design tools that are available. The majority of what is created will be derivative and will add approximately zero to overall addiction level of the game. A much smaller percentage of the user content will transform the basic game set forth by the game developer into a more widely appealing experience.

One of my favorite stories is Stone Soup. In this tale, a starving soldier arrives at a small town in the middle of a famine. No one will give him food, so he pulls out a large kettle, fills it with boiling water and drops in a magical stone. The villagers, being quite curious, ask him what he is doing. He is making a delectable meal of “Stone Soup”, he replies confidently and mentions that it tastes best with a carrot of two. A carrot is found and one by one the other villagers add their hidden food to the pot. By the end of the evening, the pot contains a rich stew that provides the entire village (and the wily soldier) with a sumptuous meal.

The creation mechanics for a creation oriented game are the initial kettle, the water and the magic stone from the tale. They are the incentives that get the community of strangers to contribute and build something great that can be shared by all. The game designer is the soldier, the charming and crafty persuader who sets the system in motion.

More than just tools!
It is worth mentioning that 3rd generation creation mechanics are not merely tools. This is a common mistake made by many mod-crazed designers. They provide a set of verbs and call it good. The content needs to be filtered, made available to other players in a painless fashion and finally good efforts need to be rewarded.

When a game developer merely supplies the tools, they are abdicating their position as the leader of their game’s creative community. They are saying “Look, we let you experience our creative vision and we gave you a few tools. Now figure it out. We have more important things to do.”

I’m convinced that this path leads to investing 80% of the effort and only getting 20% of the benefit. Developers that properly service their creative game players have a greater chance of achieving a longer sales life. It is really that simple.

Rewarding the creative class
The ultimate drivers of a 3rd generation user content system are the mod makers, map makers, and other creative folks that play your game. They deserve special attention. Suppose that only 1% of your players are actually contributing interesting new content to your game. The main fellow is a chap by the name of Willy S. and he has single handedly produced a mod that is played avidly by 20% of your player population. You could A) Do nothing or B) Give him a reward and publicly shout his success from the highest rooftop.

I’m all for intrinsic rewards, but a making a welcoming environment that encourages artistic experimentation in is always a good thing. I’m willing to lavish in-game currency on the 5% of players who make the 80% of new content that drives the continued subscriptions. I’m willing to make them in-game rock stars. Game design naturally blends into designing systems that promote a desired social structure.

Sure, the creative class will be a bunch of whiney bastards when the fame goes to their heads, but that’s okay as long as they keep producing hit content. Yes, there is indeed a certain irony here. In the new world of user content, game developers are meta-publishers and the creative users are the new game designers.

Some thoughts on making money through user content
So we have a great user content system in place. The content is making its way to the vast majority of players and adding substantial value to the title. The creative class is cooking up new content and feels happy and well rewarded for their efforts. Wonderful, now how the heck do we make money?

There are two distinct scenarios that come to mind
  • Retail Titles
  • Online Titles
Making money off user content in retail
The goal of retail titles is to sell more titles. You only make money from selling boxes, so every tactic relates to this goal.
  • Word of mouth: Have a strong community (an outgrowth of a strong user content strategy) will often cause players to promote the title to their friends. Developer support for the community can enhance this marketing force.
  • Perceived value: A large archive of user content can increase the perceive value for money of a title. Campaigns can use this as a benefit point.
  • Repackaging user content: A common technique is to bundle user content together into an expansion pack and sell it as a new SKU. This extends the shelf life of the title at a much lower production cost than if the game developer was required to create the expansion pack content from scratch.
Ultimately, user content ends up being a ‘nice to have’ in the retail world due to several factors.
First is that fact that users generally only purchase once. At a certain point in the game’s life cycle, the market for a particular game becomes saturated. When everyone has the game that might buy it, all the additional content in the world does little to effect sales. You end up with a lot of happy players who can now play their favorite game for years. This is great for the players, but it doesn’t put any more money in the pockets of the game developers.

The exception to this rule are titles like the Sims. They have a broad enough demographic that it is unlikely that they will ever saturate the market.

Second, in the retail world word of mouth is a smaller sales driver than things like brand, reviews and advertisements. When the official buzz starts to die down, the community can only maintain a trickle of additional sales.

The exception to this rule are titles like Counter Strike. The multiplayer aspects of the title end up creating a magnified word of mouth effect that are quite useful in driving additional sales.

Making money off user content online
Online games are once again a different beast. The goal of online titles is to keep your current customers for as long as possible. Selling boxes is nice, but the real money comes from keeping customers. The lifetime value of a customer in a service-based business is dramatically higher than the value of a customer who only purchases once.

Let’s look at a game like World of Warcraft and compare it to Doom 3. Suppose that a customer buys Doom 3 for $50. The developer might get $10 of revenue from this sale if they are very lucky. With WoW, the customer pays $14 a month. Let’s say Blizzard gets to keep $10 of this with the rest going to various publishing obligations. If the average life of a customer is 2 years, they’ll spend $220 on WoW.

In this scenario, keeping a customer for an extra month in an online game is worth just as much as selling one copy of a purely retail game like Doom 3. In addition, since you already have an existing relationship with an online customer, your cost of keeping them happy should be less than your cost of acquiring a new customer.

This changes the dynamics of the business model quite dramatically. In retail, you only care about making existing customers happy so that they will act as a marketing effort to pull in new users. In online games, you care about keeping customers happy so they will play your game for all eternity.

An infinite, evergreen source of high quality user content becomes a major strategic tool in keeping customers. Here are some ideas:
  • Create a source of free content that is distributed to long time users.
  • Create a system where players can purchase content created by other users. The game developer gets a cut of the proceeds
  • Create contests that set forth a theme and prizes. The community judges the content and the winning package is then collected into themed expansion packs
The most appropriate system will be dependent on the online game’s revenue model. For example, the subscription-based services may be happy with providing an archive of free content. A service that makes money by selling avatars may be best served by reselling user content. The sky is really the limit here. Over the next decade, I predict that we will see some innovative and highly successful business models built around user content.

Conclusion
Once again, I’ve meandered to the end of another essay. We’ve covered some rather important topics:
  • The basic benefits of data-driven development and how it leads naturally to mod friendly games
  • The potential market benefits of encouraging user content.
  • The various generations of user content system and the challenges with creation, distribution, filtering
  • The advent of 3rd generation user content systems in which the act of playing the game results in content generation.
  • A conceptual design tool known as “creation mechanics” for seamlessly integrating a user content system into the game design.
  • The importance of seeing a game as both a development platform and a community service, not just a fixed bundle of developer created content.
  • The importance of cultivating a game’s creative class
  • Business models for both retail and online games that take advantage of user content.
In the end, I’m most interested in production efficiency and making money. User content is a natural answer to both of these topics. Game developers can leverage their investment in data-driven development systems to empower their game player’s creative class. The end result is games with a much longer shelf life and only a marginally more expensive development cost.

Ultimately, this process will yield new genres of games that look distinctly different than those you find today. A game like Spore may superficially resemble Diablo or a RTS. However the compromises necessary to embed a robust user content system into the title produces game play that feels radically different than previous games based off the same core game mechanics.

Properly done, user content is a win for the developer and a win for the game players. (Arrr…and the pirates are happy too)

Take care
Danc.