A Deeper Look at Richard Bartle’s Player Types
Achievers are motivated to win. Explorers like to discover the intricacies and secrets of their world. Socializers enjoy human interaction, helping others, and building alliances, while killers like to dominate those around them.
If you’ve played any kind of multi-player game or been involved in a community organization (whether online or in the real world), you’ve run into all of these player types. This player typology was developed by Richard Bartle, a multi-user dungeon (MUD) creator and academic, during the 1980′s and formally published in 1996. Since then, Bartle’s player types have become one of the best-known design patterns in online gaming and in the burgeoning gamification field.
The appeal is clear. Player types provide application designers with a new way to look at psychographics and motivations and at the different ways we have fun. Once they understand Bartle’s typology, designers can easily enable specific social interactions targeted at each type. Amy Jo Kim provides an excellent an example of how to do this in her Gamification 101 workshop:
In this series of posts, we will be taking a deeper look at Bartle’s player types. We will look at his original 1996 treatise and subsequent writings and explore why Bartle’s typology has been more appealing and enduring that other possible models.
What Bartle Says
First off, how does Bartle define player types. Bartle’s basic model maps players on two axes that describe how they like to interact as seen in Amy Jo Kim’s social interaction matrix above. The two axes express a player’s degree of preference for acting on or interacting with the game world itself or its players. The four quadrants therefore map as follows to the four different player types.
- Achievers are interested in ACTING on the WORLD. They are typical gamers, playing to “win”. They give themselves game-related goals, and vigorously set out to achieve them.
- Explorers like INTERACTING with the WORLD. They delight in discovery. They try to find out as much about the environment’s topology and physics.
- Socializers are interested in INTERACTING with other PLAYERS. They spend a lot of time chatting, and empathize with other players.
- Killers like ACTING on other PLAYERS. They wish to dominate them, either through bullying or politicking. They use the tools of the game to cause distress to other players.
Beyond the simple classification scheme, Bartle offers several additional insights into player behavior and multi-user games.
1. Each player type has its own internal dynamic
Looking at each player type individually, Bartle made the following observations:
- Achievers do not need the presence of any other type to play. Therefore, a multi-user game can’t have too many achievers.
- Explorers are notoriously difficult to increase because not many people have the type of personality which is riveted by single-minded exploring. If you have explorers in a game, hold on to them!
- The more Socializers there are in a game, the more new ones will be attracted to it.
- Established Killers will only decide to stop playing if there is very drastic reduction in the number of players. However, Killers can be held in check if there are a sufficient number of Explorers who often have formidable fighting skills and who tend not be bothered by attacks. Thus, when Killers attack Explorers, they are denied actual or emotional victory neither of which is satisfying.
2. Multi-user games need a balance of all player types
The next insight is that a stable multi-user game “is one in which the four principle styles of player are in equilibrium.” This is not to say that the number of each type needs to be equal to have a stable system, but rather that the proportion of each type is relatively constant in such a system. Bartle goes on to say that it is the combination of all four player types that makes multi-user games unique.
This last point might seem to be counter-intuitive in light of the fact that, for example, a game designer might rationally decide that they want to eliminate or minimize opportunities for Killers to have their kind of fun. However, tilting away from Killers necessarily means tilting the game play away from the action oriented play that achievers enjoy and away from the player interaction that socializers crave. Thus by shying away from killers, a designer “would add depth and interest, but remove much of the activity. Spectacle would dominate over action, and again there would be no need for other players. The result of this is basically an online book”
In the end, tilting the balance of the game too much toward any individual player type would result in a single-user game, a book, a chat line, or an arcade game.
3. Player types interact with each other in complex ways
So, assuming that we’re not setting out to create any of these other types of entertainment, what happens when all of these player types mix together? I have summarized the relationships between the player types in the diagram below.
(See note  below on how to read influence flow diagram.)
One of the first things we should observe from the influence diagram is that the relationship between Killers and Achievers is relatively stable. For example, increases to the number of Achievers causes a proportional increase to the number of Killers. However, that increase in Killers results in a proportional decrease in the number of Achievers, which offsets the initial change.
Second, we see that there is a highly volatile relationship between Killers and Socializers. The number of Socializers is highly sensitive to the number of Killers and is self re-enforcing. Increases in Killers tend to create a large exodus of Socializers, and the remaining Socializers become more likely to leave as they see fewer and fewer of their fellow Socializers. Reductions in Killers can have a similarly opposite effect.
Next, we see that the relationships between Socializers and Achievers, and Explorers and Killers are fairly weak. Achievers are unaffected by changes to the number Socializers and Explorers, while Socializers are negatively affected only by large changes to the number of Achievers. Explorers on the other hand tend to be altogether isolated. Killers are only affected by large changes to the number of Explorers while conversely Explorers are only impacted by large increases to the number of Killers.
4. There are only a few stable player type configurations
Bartle determined that only the following three configurations of player types were stable.
- Action-oriented MUDs dominated by Killers and Achievers.
- Games dominated by Socializers
- Games with a balances of all four types
These configurations are necessarily balanced along the y-axis between action and social interaction. Stable configurations are unlikely to be found balanced along the x-axis between player-centric and world-centric games because of the volatile relationship between Killers and Socializers and the non-existent relationship between Achievers and Explorers
Bartle found that the first two configurations were the most commonly found and most easy to create. In the world of MUDs, there has been an ongoing division between game-like MUDs, which allow player killing, and social MUDs, which emphasize positive social interactions. The existence of this schism provides evidence of the stability of these configurations.
The third configuration, which balances all four types, is possible but, as Bartle states, is harder to create and maintain. To develop a balanced game, designers and administrators need a high level of skill to sustain an environment that attracts and retains that rarest of breeds, the Explorers. Bartle explains that by nurturing Explorers, their overall population will gradually rise and they will hold the Killer population in check. The Killers who remain exert enough of an influence on Socializers to stop them from going into fast-breeder mode, while not enough to cause an exodus. Achievers clash with Killers often enough to feel that their achievements have meaning.
Thus, as we set out to use Bartle’s player types in our gamified applications, we should note not only that different player types exist, but also that a balance of all types must exist in multi-user games. Furthermore, this balancing act must be performed consciously to prevent the system from tilting so far in one direction that it devolves into a single-player game, a hobby, a sport, or a social club for the essential nature of multi-user games to be all of these things.
In future posts, we will look at how Bartle used an expanded model to describe how players move between types over the course of a game and investigate the basis for applying Bartle’s player types beyond the confines of the multi-user games for which they were originally conceived.
 Bartle, Richard A. “Hearts, Clubs, Diamonds, Spades: Players Who Suit MUDs”. Journal of MUD Research. Vol. 1, Issue 1. June, 1996.
 Bartle, Richard A. “Virtual Worlds: Why People Play.” Massively Multiplayer Game Development: v.2, Ed. Thor Alexander. Charles River Media. 2005.
 Key to Player Type Influence Diagram: Green indicates increasing numbers and red decreasing numbers. Line thickness indicates the magnitude of the change. Thin lines indicate small changes and thick large changes. Arrowhead size indicates the magnitude of the influence of a change on the affected group. Large arrowheads mean that there is a great effect, magnifying the influence of the change. Curved lines indicate the existence of feedback loops, while straight lines denote a one-way interaction.