The Bartle Test

Tycho said this recently:

In order to play a game with someone, more often than not, you usually have to be playing the same game. That probably seems like something you wouldn’t have to enunciate, but not everybody likes to do the same stuff, let alone has spent their lives building the skillset required to perform pixel-accurate jumps.

There are something like a trillion ways to define types of gamers. One that’s gotten some play in this blog is GNS theory, or the idea that gamers can be divided into gamists (who like to win), narrativists (who like to tell a story) and simulationists (who like to play in a world that makes sense). I’ve discussed it briefly a couple of times, but for a better analysis there are a bunch of places you can go.

I don’t buy the idea that there’s some trinary variable in a gamer that determines what they like. Instead imagine a triangle, with one style at each point, and I think a person falls somewhere inside that triangle’s area. Perhaps they like narrativism and simulation equally but don’t have much respect for gamism, or whatever. So that makes it a little difficult to sit down and say for a fact “system X is for people who like style Y”, because that could mean that system X is lousy for people who are only partially Y. I suppose that for the purposes of figuring out how gamers identify and forming a launching point for discussion, it’s perfectly fine.

But I didn’t start in role-playing games. I joined my first D&D campaign in 2003, relatively recently. I have, however, been playing video games since about 1989. A lot of my worldview is based around video gaming, whether I’m building worlds (Is this a place I want to interact with?), characters (Is this a character that will still be fun at the end of the game?), monsters (Does this guy have something that makes him memorable but not frustrating?), even plots (Is there a point here where a player would be fine putting the controller down and never coming back?). Heck, even my first D&D book was the Monster Manual, which I picked up because it had parallels to a game I was playing at the time. There’s an anecdotal blog post in that story.

The amount of overlap between gamers (pen and paper) and gamers (video) is a lot closer than one might expect, evidenced by the fact that these groups of people share a name. Before I had ever heard of GNS Theory, I was already looking at my characters, players, campaigns, et cetera through the lens of a different theory: the Bartle Test.

Very roughly, it comes from the idea that different players want different things from games, even from the same game, and it presents a way to determine roughly what a given person wants based on two axes: action vs. interaction, and game world x other players. It gives designers a simple, shared analysis to consider how players are playing the game, and it gives them some guidelines on how the different types interact and how to foster or deter each group. It’s designed for multiplayer games so it wasn’t hard to extend to tabletop gaming. There’s a more evolved three-axis version, but I initially internalized the two-axis version, so it makes sense to start there. It also comes with an easy-to-remember naming system based on playing card suits.

A spade wants to interact with the game world. Bartle calls them “explorers” because they’re interested in testing the limits of the game itself more than testing the people who populate it. In a MMORPG, these are the people who find the easter eggs, who notice when NPCs refer to each other or share similarities, who understand the lore behind the game. At the table, these are the players who want to know about the world and the campaign background, who like NPC conversations, and who actually think about things like “how does the ecosystem in this dungeon work?”.

A heart wants to interact with the other players. Bartle calls them “socialisers” because he’s British but also because these are the players who form social networks regardless of the game around them. In a MMORPG, these players join guilds and adventure with their friends. At the table, these are people who are there because it’s a fun sort of social outlet, and often the system or campaign isn’t as relevant as the people who attend.

A club wants to act on the other players. Bartle calls them “killers” because this usually manifests as some sort of PvP, though it can also manifest as currency manipulation, changing the world in ways that affect other players, and so on. At the table this can appear as the unlikeable backstabber, and as such there isn’t a lot of support for it in systems like D&D. But in systems where conflict between players is expected and encouraged (Paranoia, for example), clubs can thrive.

A diamond wants to act on the game world. Bartle calls them “achievers” because their goal is triumphing over the game in a way that produces rewards. In a MMORPG, these are people who work toward the best gear or get all of the achievements. At the table, they’re players who want to know the plot so that they can resolve it, because finding a problem and fixing it is most rewarding.

There’s a lot of room here for different player types to exhibit the same behaviors, which is fine. For example, a diamond might design an optimal combat build to defeat encounters more easier, a club to show up or kill other characters, and a spade to explore what the system can do. A heart might benefit from a game rule that rewards player interaction, but so might a diamond who uses it for a mechanical advantage. Everybody but a club might enjoy dungeon exploration and everybody but a diamond might enjoy quick one-shot sessions.

The thing I really like about it is that is provides some sort of semi-objective measurement (valuable for any DM who’s ever noticed that players have a harder time vocalizing what they want than vocalizing what they don’t want) but also puts preferences on a scale rather than a hard identification. Unlike GNS, there’s no “I’m a heart”, but rather “My primary focus is heart, but a close second is diamond and I really don’t like club”. It gets past the issues I’ve seen where a player might, for example, claim to be a narrativist but play like a gamist, because it acknowledges that a person can enjoy two things but allows them to express those likes as something understandable and comparable rather than vague guesswork.

It’s important to note that, like any such analysis, this is a tool rather than a hard definition because every player has each role in some measure. A heart can enjoy beating a dragon; it’s just more fun beating it in a party. A club can enjoy playing D&D with a group, but they’ll prefer exerting their influence on the NPCs and other players. Rather than telling you exactly what type of gamer you and your players are, it gives you a framework to understand the sorts of things you might enjoy and use that to figure out what sort of game you might want to play.

You can take the Bartle test here.

This entry was posted in Game Design. Bookmark the permalink.

5 Responses to The Bartle Test

  1. Yanni says:

    So I took it and my results were:
    E 93%
    S 73%
    A 40%
    K 0%
    Now, maybe I don’t know myself very well, but I kind of feel like my Achiever score should have been a lot higher.

    • Duncan says:

      I got:
      A 73
      E 53
      S 47
      K 47
      Which seems about right to me. I would have called myself social before explorer (but they panned out being pretty close). Maybe it’s just that ‘social’ and ‘MMORPG’ don’t dwell in the same place in my mind whereas ‘social’ and ‘pen and paper RPG’ do.

    • MssngrDeath says:

      Based on what I know about you, I’m a little surprised too. Maybe it’s a sign of the difference in how you play tabletop games versus other, competitive games? I know I like achievements, for example (check my Kongregate!), but not so much in D&D, and I was something like E86 / S53 / A47 / K14.

      • Yanni says:

        Dude, I am totes going to catch up to you on Kongregate someday! However you’re right in that I do tend to try and get different things out of Online vs. Tabletop games vs. Videogames … and online tests can only be so accurate anyways.

  2. Dave Fried says:

    S 67% E 60% A 46% K 27% – my payouts are varied, but I am a heart/spade first (and a murdering jerk last). And if you remove (S) – i.e. in a single-player game – the remaining percentages are also very accurate for my play style. For me, CRPGs are all about exploration first, then finding the best stuff, then (when I’m all done) killing whoever I have to kill to finish the game.

    As an aside, I should also mention that for RPGs, G/N/S is better at classifying games than gamers. People have complex payoffs, and theory has kind of moved away from trying to pigeonhole them (though IMO it’s still quite useful). A recent thing though – Blake and I were talking about how you could overlay GNS with Wizard’s M:TG “Timmy, Johnny, Spike” as a second set of axes to create a much more representative player profile for tabletop RPGs that takes why a player plays into account in addition to what they like in a game. (I’m mostly N-T.)

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.