Christopher Harz looks at the fundamental change that may be occurring in online game development, with implications for game producers, players and toolset developers.
One of the common themes in recent Serious Games conferences is that many game designers are out of touch with their audience, the players or end users of their games. Surprisingly, this theme has also been echoed at entertainment game expos, as major gaming gurus spoke of designers thinking that they "know too much" -- that they design games for themselves rather than for the users. No less a person than Phil Harrison, president of Sony Computer Ent., called for a fundamental shift in online gaming, to games that were oriented to -- and heavily influenced by -- the user. "It's about community. It's about collaboration. It's about customization. It's about emergent entertainment powered by the audience, with the audience at the center of this universe," he noted at the recent GDC.
Designing a game (in either serious or entertainment genres) with heavy user involvement runs counter to the conventional wisdom, which calls for essentially finishing a game design and then -- and only then -- releasing a "beta" version for actual users to try out; the earlier "alpha" version is usually only tested with an in-house audience. By the time the game is at the beta stage it is mere months before "shipping gold" (sending the master to the DVD factory), far too late to incorporate any significant user input.
Designing a game of the type that Harrison and others are calling for involves a radical shift in thinking for much of the game community, a change that can be painful but could also engender a lot of success and new opportunities. Rather than use the cliché term of "shifting" paradigms to describe such a change, let's use the gaming term of "power-up" -- a gain in capability that can change everything that was once familiar. We need a paradigm power-up.
So what does a Serious Game that was evolved with user interaction look like? How do you actually get to the stated goal of "community, collaboration, customization?" To get some answers, let's look at three games that are successful (that is, that are actually being used by user communities to learn and train with), to see what we can learn from them. They are: America's Army,
Sim Ops Studios has now renamed Hazmat: Hotzone and its new moniker is Code3D, which is being released for nationwide sale and use on May 15, 2007. It is similar to Hazmat: Hotzone in that it is oriented to firefighters, but covers more types of training, and offers an even stronger toolset to instructors or facilitators. Instead of using Unreal, this game uses the Panda 3D engine, which allows the facilitator more graphic manipulation of game elements as he/she sets up each scenario.
Code3D was also developed in close cooperation with firefighters, this time not just with one group, but with 20 Fire Department and Emergency Response sites nationwide. "The unprecedented aspect of this game is the ease with which instructors can adapt it for their specific training requirements," Tellerman noted. "Instead of paying someone $50,000 or $100,000 to adapt it for them, they can do it themselves. They can build the basic 3D environment, select different types of hazards, insert people, and insert vehicles such as tanker trucks that can move around and even crash and burst into flame. They can speed up or slow down the time line, and go back to a point in time to proceed differently. They can pause the game if they want to make a point, and show the scene from different camera angles."
The instructors do not have to generate their own scenarios, as the game comes with a library of pre-built scenarios. However, it is expected that most instructors will want to "tweak" the basic scenarios, and will use these set-ups more as templates. Enabling non-animators to in essence build their own game levels took an extraordinary amount of GUI development. "Whereas Hazmat: Hotzone's GUI is simple to use, it is still primarily text based," noted Tellerman. Code3D is really graphic, around 75% of the elements on the control screen are actual graphic elements in the 3D world, and the user has the ability to manipulate an element such as a fire truck, turn it around, change it by putting different placards or signs on the truck, insert it into the environment, and then preview what the final setup would look like."
Getting feedback from users with succeeding prototypes elucidated what elements were considered essential to authenticity for this community. "Anything that related to the community's traditions or artifacts was highly important," Tellerman said. "We had to render six different helmet types accurately, with proper coloring, as well as three different uniform types, with proper badges and ranks." The technique also helped fine tune some expectations, and avoid unpleasant surprises. "We had believed that people in a scenario could be moved from one point in time to another, from one physical point to another, with no "tweening" (intermediate movements between the two physical points) necessary, but that proved to be wrong -- the users insisted on it.
On the other hand, they did not care about a 100% perfect walk cycle -- where the person's feet perfectly meet the floor as in real life, so we were able to compromise on that. What the users prioritized was often a surprise. But we were able to draw this out early by getting the response from the prototypes. This also helped us greatly with the GUI, in identifying parts that were difficult to use." Tellerman definitely recommends the prototype-with-users approach. "As painful as it feels -- get your prototype into users' hands as early as you can, even if there are still rough parts," she noted.
User interaction will continue even after Code3D is released. The company will continue to ask for and examine user feedback and suggestions, and expects the user community to react enthusiastically, especially since the users are "stakeholders" in the game -- they have a personal stake in its continued relevance and utility. Whereas Sim Ops will offer future upgrades for the game, Tellerman also expects that a cottage industry may develop for local animation houses -- to develop photorealistic buildings, landmarks and other features that are specific to a particular fire department's locale and that can be inserted into the game.
America's Army is a multiplayer online game that is in wide usage, both as a PR tool for the Army (in its popular version, which is available as a free download, and has registered over eight million users) and as an actual training tool for soldiers (where it can be played up to the Secret level). The civilian version of the game can be downloaded for free from the America's Army site and it is also available for the Sony PlayStation 2 and the Xbox.
America's Army was also involved with extensive interaction with the user communities, especially Army Infantry and Special Forces. According to Dr. Michael Capps, exec producer, the design and production team spent a lot of time at Ft. Benning, the Army Infantry School, as well as at other bases, firing Army weapons, operating Army vehicles, and interacting with real Army personnel. The initial production team, housed at the MOVES Institute at the Naval Postgraduate School in Monterey, California, also worked tightly with military training experts with dual knowledge of both games and Army requirements, who served as intermediate agents between the design and user communities, according to Dr. Michael Zyda, who was director of the Institute.
Again, tight interaction with users enabled the design team to determine what game features were necessary to appear authentic to Army users, and what features could be compromised in order to reduce rendering times and speed up gameplay. Generating prototypes and getting user feedback played a major role in game development, as did thorough up-front planning, according to Zyda.
The project's initial goal was to give inclined young Americans an idea of what life in the military is really like, in the hope of arousing their interest and making them more comfortable with the idea of enlisting, and to give the American public a better image of the Army in general. The initial America's Army offered two different modes of play. In the Soldiers mode, the player guides his character through an Army career by setting personal values, resources, and goals, beginning in basic training, then moving on to through Airborne and Sniper School on the way to becoming an Army Ranger.
The more action oriented (and more popular) Operations mode allowed players to take the first-person roles of skilled soldiers in the field to accomplish a variety of missions. Each player must work as part of a team -- teams that do not collaborate usually fail. The player character's particular role in his unit is determined by such factors as his leadership ability and MOS (Military Occupational Specialty).
The game offers extensive support to newbies (unlike many other games, which throw them right into the fray). New players get extensive help in getting introduced into the skills and traditions of the group. Pre-mission briefs give a clear explanation of what is expected for each mission. The professional version of the game, used by the troops, also has post-mission briefs with experts that explain what happened, and what went right and what needs improvement.
Having started as a PR vehicle for the Army, the game evidenced such a level of realism that a form of it is now being used for actual training for the U.S. troops. The game's success has also led to the genesis of a new group, the America's Army Government Applications office, which opened in Cary, North Carolina, where it has easy access to a number of Special Operations bases in the area, as well as to the home of Epic, which makes the Unreal engine that powers the game, and which continues to offer major support in continued improvement in the game software. The Government Applications Group is evolving new forms of the game for training other parts of the government, such as the Secret Service and the IC (Intelligence Community).
Capps, now the ceo of Epic Games, producer of the award-winning Gears of War and Unreal Tournament games, continues to be a strong believer in a tight partnership between the designer and user communities, which should continue even after the game has shipped. He noted, "We've always benefited from user-created content; our map/mod community keeps our game alive on shelves long after we've finished. We also release multiplayer betas for user feedback. MMOs like America's Army certainly benefit from user feedback long before release, with their extended beta periods. Ongoing feedback is the reason we, Blizzard, Bioware, and many others spend so much time and effort on community websites and forums."
Zyda is now the director of the USC GamePipe Lab in the Los Angeles area, where he teaches new generations of game producers in successful game technology and design techniques.
Winning Game Design Elements
It appears that there is a lot to learn from these successful games, which really stand out in a field where failure (non-adoption by the user community) appears to be the rule rather than the exception. Much of the design philosophy in these three games appears to be remarkably strong and consistent, and runs counter to the design approaches of many Serious Game production groups -- and, as Michael Capps noted, many of these design principles should be used by multiplayer entertainment games, as well.
The most obvious common factor is that all three games considered the game creation process to be a partnership between the design and user communities, with the designers going far out of their way to live with, eat with (and drink with!) and learn with the users in their native setting. Designers also paid a great deal of attention to authenticity, as perceived by the users. They found that authenticity does not equate with reality, and that it can be really surprising to learn what elements users consider to be necessary for an environment or story element to appear "real" to them -- a real danger for designers, who had to overcome their own prepossessions as to what constituted such a perceived reality.
The designers of these games also had a great deal of respect for the users as a community, and assured that they could learn together as a group, not just as a collection of individuals. Support to the community included enabling authentic ways to communicate among its members, high-resolution rendering of important community traditions and artifacts such as uniforms and badges of rank, and the collection of common knowledge as it was created. One means of identifying and validating such knowledge was by post-mission debriefing sessions, called "After Action Reviews" by the military, where players review and reflect on what they have learned, usually in the presence of an expert who serves as a facilitator for the group. All three games acknowledged the role of such an expert, whether live (in an actual classroom) or simulated in-game (by AI, in Tactical Iraqi). The games also made use of one or more intermediary agents, who could understand both "userspeak" and "gamespeak" and were thus able to function in both worlds, thereby avoiding many translation mistakes.
Interestingly, there is an academic theory of learning that systematically describes the above aspects as being necessary for learning by a group of professionals -- not just the firefighter and military communities targeted here, but by any group of professionals that functions and learns together. This form of social learning theory is termed "Communities of Practice," or CoPs, and is described in Wenger's Communities of Practice: Learning, Meaning and Identity (Cambridge Press, 1998), and in many other books and research reports, which describe the social, psychological and anthropological underpinnings of what is necessary to make a community "tick" either in real life or in a gaming environment.
One of the hallmarks of a CoP is that it can be very specific, that is, that its traditions can be very local, down at the level of a specific firehouse or military unit, making it all the more important to enlist such users in the design process, because no game production team could customize a game to that level by itself -- as all three teams confirmed, such customization would bankrupt any game budget.
In order to achieve the above objectives, these design teams used the strategy of rapid interactive prototyping, rather than the straight linear development process that is far more popular. This approach can be "painful," as one designer mentioned, and can lead to many radical changes in assumptions and design at an early stage. But all three teams voiced a unanimous opinion of, "Better early than later." Feedback from users at the very early stages avoided a lot of unpleasant surprises at the beta stage, when it was usually too late to make any major changes to the game.
All three teams chose to personally witness and document the release of each prototype as end users were trying it out, to mark how well it did and get user appraisals. "Documentation" in this case did not mean the very extensive (and time consuming) formal write-ups that government bureaucracy usually demands in a training development process, but consisted of much more informal note taking and videotaping of group sessions, with the use of screen capture tools such as FRAPS or 3D in-game recording (when possible) to record gameplay from the user's point of view.
It is also interesting to note that all three teams did not consider this partnership to be over upon game completion, but planned for ongoing feedback from the users, especially users that were involved in the original development, with provision for incorporating such feedback into future game upgrades.
Userware and New Toolsets for Game Development
Tools to enable the modification or even the creation of game levels are nothing new. Will Wright pioneered much of this capability with his Sims family of games, each of which is basically a game about creating a game. Modding a game (changing the game's characters and environment) has been around for quite a while -- many expert players are so used to modding games, especially First Person Shooters, that they refer to un-modded games as "Vanilla" versions, or "V" for short, such as VQ3 (Vanilla Quake 3). Persistent worlds such as