Ducheneaut, N., Yee, N., Nickell, E., Moore, R. (2006). Building an MMO with Mass Appeal: A Look at Gameplay in World of Warcraft. Games and Culture, 1, 281-317. PDF
There is some good work coming out of the Palo Alto Research Center lately about the design of MMOs. This data-driven look at World of Warcraft had some interesting insights about the ways people play the game and what motivates them.
The game is designed such that the players experience rapid progress and frequent rewards during their first play sessions—an important design strategy to encourage continued play. The rate of progress is then fairly stable up to Level 50, at which points it drops precipitously. As we discuss later, by the time players reach this “endgame,” interest in WoW is not driven by achievement through leveling anymore: Group activities (e.g., high-end raids in instances to collect powerful items) replace basic quests, and gamers turn to reputation as a marker of achievement (Taylor, 2003). (10)
Hooking players in the beginning is essential, because that's when players are at the bottom of the social ladder in terms of achievements and reputation. These beginning players may not have a social network yet that can acknowledge them for event-based achievements. By the time players have reached the highest level, at least in World of Warcraft, continued play typically depends on collaboration with guild mates. Therefore, there are witnesses to achievements.
The players' apparent reluctance to play “ugly” and “bad” characters could indicate that despite the anonymity of virtual worlds and their potential for experimenting with different identities (Turkle, 1997), social and cultural norms still shape an individual's choices in virtual worlds powerfully. In particular, the more “politically correct” races and classes are apparently more popular. (14)
New and less experimented players tend to pick “friendly” races initially: The forces of “good” portrayed in popular movies like Lord of the Rings , for instance. The Alliance races “look good” and can be more attractive at first. Conversely, achievement-oriented MMO players tend to pick the “bad guys,” either because they already tried the other side and want to experience something new or quite often, explicitly to avoid playing with noobs 3 (a derogatory term for inexperienced players). Our numbers clearly show this Horde bias toward achievement. (22-23)
I wonder if the main factor here is that players strive to see themselves as pretty/handsome or if pretty avatars are simply more pleasing to look at during play.
Support roles are less in favor than others. Being less familiar with the notion of a healer, these newcomers might pick the more offensive classes by default. (14-15)
Or these newcomers may simply want to be the ones doing the damage and making the obvious impact. There is a different kind of gratification that comes from pummeling the enemy then comes from supporting one's team. Experienced players may choose support roles simply because they've played and grown tired of the traditional damaging roles.
We wonder however if Blizzard intended such a clean cut between the two factions. Indeed, it can have some negative consequences in terms of game balance. On many servers, for instance, the Horde/Alliance imbalance seriously limits the number and quality of matches in the Battlegrounds (arenas where groups of players from each faction collectively battle against each other, leading to valuable rewards; battlegrounds open up only when a large enough and balanced number of players from each side sign up for them). (24)
It is difficult to fully predict who will choose to play which side and why. It's risky to design a competition system based on this balance. Is there a compelling reason to have battlegrounds strictly between factions?
It is quite clear that at Level 60, WoW becomes a much more intensely social game—an interesting contrast to the earlier stages of the game where a large majority of time is spent alone. These high-end raids are significant undertakings requiring a lot of planning and in-game communication. (28)
The game is clearly designed to steer players toward more and more hardcore activities requiring a lot of time and effort. In this, WoW is not different from any of its predecessors in the genre. (30)
Our social network metrics clearly show how this need for increased coordination translates into more frequent social contacts between the players. Although this may sound desirable at first, it is important to keep in mind that most WoW players progress through the game alone (Ducheneaut et al., 2006), and the sudden switch to large-scale groups could be jarring. (34)
The fact that there are really two different games here is disturbing. If a player is primarily interested in the game until level 60 -- the more independent aspects -- he may not want to give up that game upon reaching level 60. Beginning a new character is an option, but requires giving up achievements from the old character. Those achievements and the corresponding sense of competency and accomplishment are big parts of what keep players in the game, so asking these players to start from scratch is inviting them to quit. Perhaps these players should be allowed to continue the current play style, accumulating visible accomplishments other than leveling up. Similarly, if a player is primarily interested in the game after level 60 -- the more dependent aspects -- it's asking a lot to make a players slog through the first 59 levels before the real game begins. For these players, perhaps the endgame play style should be made available much earlier.
Characters on the PvP server were observed to be in a group more often than characters not on PvP servers. (31)
PvP combat apparently encourages players to group more and play more, suggesting that shared adversity might be a positive force in the social life of a MMO if it is correctly implemented. (33)
If game designers knew this already, they didn't communicate it to players. When picking a server to play on, it wasn't obvious to me that player versus player servers would involve more grouping. It may make sense in hindsight, but it didn't occur to me at the moment I was choosing a server. If the designer's job is to help me have a good time, he needs to educate me about the consequences of available choices.
