The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.
While not a necessity, the author highly recommends that readers understand the concepts of Virtual Equity and Supremacy Goods before attempting to apply the information contained in this paper.
Persistent Gaming Collectives
When a player elects to participate in a multiplayer game, one of the primary motivations is to interact with other players. If this was not a strong motivator, the player likely would have a better experience playing a dedicated single player game. When a game design permits it, or better yet rewards it, gamers in multiplayer environments tend to rapidly and organically form groups. Because of the strong motivations and communication opportunities found in such games, groups of gamers quickly form a gaming collective if the interaction is allowed to be persistent.
Examples of games where Persistent Gaming Collectives (PGC) have formed would include World of Warcraft and EVE Online. Games like League of Legends and World of Tanks, where matchmaking is largely anonymous and the social interactions last only for the duration of a battle, do not foster the creation of PGCs. The key difference is whether friendships and repeating peer interactions are fostered, as these types of relationships create peer pressure and reinforcement that can greatly increase retention and willingness to spend. In environments where persistent groups are allowed to form, and interact with other groups (cooperatively or competitively) the PGC is functionally the size of the entire server. Note that both LoL and WoT have some clan/company/team modes that allow persistent interaction, but these are utilized by a relatively small percent of the total population, and these also allow tournament competitions.
Due to the competitive nature of a PGC, any information that becomes available to one part of the PGC tends to propagate rapidly across the entire PGC. Thus a PGC acts very much like a “hive mind”. If the average budget of a participant in a massively single player game like Star Wars: The Old Republic is “x”, then you would expect the aggregate budget of 1000 players to be “1000x”. In the presence of a PGC, the budget is not 1000x, it is “Q”, the group budget. Q is always initially greater than the predicted sum of the budgets of all participants, and continues to rise over time if not inhibited. Q is rapidly affected by the group perception of the game environment, and is capable of rising or falling in real time.
Game environment characteristics that can raise Q include cooperative play, fair competitive play, systems that favor new players, prestige-based reward systems, and a virtual economy that maintains equity. Good customer service and the perception that game makers are receptive to the whims of the PGC also boost Q. If any of these characteristics act to reinforce other positive characteristics, then a positive feedback loop can form. The result is that the consumer will have a very difficult time leaving the game environment willingly. This state is not coerced, it is consensual.
Game environment characteristics that can lower Q include unfair game mechanics of any sort, unfair monetization models (“pay to win”), griefers, “gold farmers”, spam, game design elements that favor veterans at the expense of new players, systems that promote or reward aggressive narcissism, equity losses, the presence of bugs and exploits, service interruptions, and any form of Supremacy Good, including systems that give unfair advantages to players who can play for extreme durations. The presence of multiple systems that negatively affect Q that also reinforce each other can cause a negative feedback loop, also known as a “death spiral”, that can cause Q to collapse rapidly to nearly zero.
I am currently performing Q-optimization on an experimental high budget AAA game where it is uncertain how organically the PGC will form due to the novel way players interact and restrictions employed to make the game child-safe. A good example of a currently deployed game with a strong PGC that is reasonably monetized is Glyph Worlds' Ways of History game, though the limited graphics and glacial pace of the game make the target audience quite limited, and studying the design a bit time consuming (matches take about a year on average).
Group Budget Capture
It is not enough to build a game with a high Q. It is just as important to be able to capture as much of Q as possible. Capturing Q without lowering Q is exceedingly difficult. To approach solutions to this problem, the concept of Group Budget Capture Percent (QC%) must be introduced. Having a high QC% means that you are getting your players to give you as much money as they are willing to spend on your game. If you have a low QC%, then that means most of the money that your players are willing to spend is still left on the table, uncollected.
Having a low QC% not only means you are losing a tremendous amount of revenue, but it also opens the door for 3rd party agents to collect that untapped consumer willingness to spend. We usually call these “gold farmers”, due to the way these agents ravaged World of Warcraft's primary currency. The presence of these agents is a symptom of a game with a high Q and a low QC%. What makes this situation dangerous is that the presence of these agents also rapidly lowers Q, thereby making your low revenue capture even lower. Examples of games with high Q values and low QC% values would be World of Warcraft and EVE Online. The designers of both games have taken extraordinary measures to combat the symptoms of a high Q / Low QC% design without doing much to remedy the root cause.
The Role of Analytics in QC%
The role of analytics use in game monetization is to raise QC%. The problem is that when this is done without an understanding of Q dynamics, the end result is that Q is lowered. Raising QC% at the expense of Q results in an initial increase in monetization rates followed by a progressive loss in revenue per user over time. The overall effect is a drop in total revenue over time, unless this technique is used to “squeeze” the QC% up for the purpose of displacing 3rd party agents. This is a poison pill approach and should only be used as a last resort.
It is human nature to take responsibility for our successes and to blame others for our failures. Numerous research studies show this. When an analytics team comes in and demonstrates a 300% increase in revenue generation by implementing a more aggressive method of QC%, and then sees a steady drop in revenue after that, it is natural for the analytics team to credit themselves with the revenue spike and to blame the game design for the later revenue degradation. Without an awareness of Q dynamics, this type of logic is likely to be convincing.
Please note that I am not suggesting that companies stop using analytics. Analytics can be very powerful. What I am suggesting is that if you take action based on the data you are collecting without proper knowledge of Q dynamics, you can do more harm than good. This is because conventional forms of retail commercial optimization are designed to optimize individual spending in environments where a PGC does not exist, like in a super market. I may not like white bread, but I'm not going to stop shopping at your store just because you sell it. If you sold guns in your store, and encouraged your customers to shoot each other with them, then I suspect your revenues would drop pretty fast.
As the “hard scientist” on the UCLA coaching staff 25 years ago, it was a revelation for me when my mentors taught me that in a sport so obviously dependent on physiological performance, “soft sciences” (like psychology and sociology) were just as important as “hard sciences” (like physiology, physics, and neuroscience). This is why the US Olympic Training Center in Colorado does not produce winners in track and field. I would assert that this also is why China, with their emphasis on hard sciences, can't seem to win an Olympic soccer game. Our industry is currently, thanks to the (in my opinion false) assertions of a company called Zynga, moving radically toward using hard science to drive game design. There are very real scientific reasons why the focus on hard science over soft science in the production of creative products will not have the intended results. The ideal balance actually does come from a balanced approach between creative elements and soft and hard sciences.
How to Raise QC% Safely
Some of the ways I go about doing this in my proprietary designs are by removing or reducing the rewards for individual narcissistic behavior and boosting the prestige rewards for group behavior. I also economically favor new players over veterans in my designs, which is upside down from the way 100% of the MMORPG's I've studied over the last 14 years have been built. This encourages new players to participate long enough for peer reinforcement to build, making it very difficult for them to exit the game environment.
If you can design a system where the players themselves are asking each other to spend more, then you are in the sweet spot. Now you can't be the “bad guy” for charging a lot, and marketing costs are reduced. Of course this requires that what the individual spends helps the group at least as much as it helps the individual.
Part of the inspiration for this game design philosophy comes from my time as a coach for the UCLA women's track team. We were ranked #2 in the nation and produced two world record holders the year I was on the coaching staff. Our secret was simple: We did not recruit narcissists. This may sound strange in a sport where almost everyone has some narcissistic qualities, but we made sure all of our runners put the team first and we were very careful in vetting them before they got an invite to UCLA. When extremely talented narcissists applied, we were careful to steer them to our direct competitors. You can't imagine how effective this was.
Actually, if you are an experienced studio head, you probably can. I've met a number of major studio heads in the last year, and almost universally the one question they have about me is “are you a narcissistic jerk?” (paraphrased for suitability) Given my zeal in writing, this is an understandable concern. What excites me about getting this question is that it shows me that the leader of this studio understands the danger of putting a “narcissism bomb” on their team, just like we did at UCLA.
So my question to the Masters of Interactive Media is this: If you know how dangerous this narcissism is on your teams, why are you rewarding and selling it in your games? That same narcissism bomb is going to ultimately explode in your game worlds, taking out your revenues.
If you are intent on making single player games, or multiplayer games with short (under a month) lifespans, then the above information will not benefit you. PGCs take time to grow. In that case, my only question would be “Why not?”, given how much more money is to be had tapping the unmet consumer demand for social interaction.