There are a whole bunch of people who reach for the Bartle’s Four when discussing MMO concepts. Frequently, they are talking about it as if it is gospel and all design should respond to it, whereas others believe that the whole thing is old and busted, and we should move on.

That being said, I see both advocates and detractors often use Richard’s paper in ways that conflicts with my own experience. Here are some of the ways I would suggest reconsidering how you use this important paper (and note, this is solely my point of view, and even Richard’s may vary).

1. Don’t assume that players are only one of the above. I see it constantly – the idea that explorers should never have to play the achiever game, or that socializers will never want to get in the arena. It’s just not accurate. While most players will tend to have a strong tendency towards one type of activity, most Bartle’s Test scores shows players are much more mixed and nuanced. Players who are 95% socializers, and below 20% on the rest, are aberrations.

2. Don’t assume player types are fixed. Player types mutate over time. Most players start as psychopathic achievers – psychopathic because they don’t know the social rules of a space, and achievers because they are looking for early validation and success. As they mature, they identify aspects of the game that speak more clearly to them, and wander towards those. Some are different kinds of players in different kinds of games — I was a pure socializer in SWG, but I’m a pretty heavy achiever-explorer in WoW. And don’t discount that sometimes people are just changed by life. When I was a CEO trying to start a startup, my killer rating taking the Bartle’s Test was 80% – now it’s 20% as I’ve become much more mellow in my old age.

3. Don’t take the Bartle’s Test as gospel. I’m sure Richard would want me to say that the Bartle’s Test was not developed by him, and that he has always had some problems with some of its methodology. It’s still fun to take.

4. Don’t assume that an equal division of players exists, or even is desirable. Richard’s thesis (which I agree with) suggests that there is a correct mix of players for a game, and that you need some of each to create the right community recipe. In the old days of UO, only 1-2% of the player base were hardcore killers. However, that 1-2% added one hell of a cajun kick to the game, creating a sense that they were much more pervasive than they actually were.

5. Don’t assume each merits an equal amount of development time. Really, that’s silly. Explorers will always need the most content and development effort, followed by achievers. Meanwhile, socializers need a handful of strong tools and a supporting atmosphere.

6. Don’t be too literalist with the title names. The title names are handy descriptors, but they often encompass larger motivations. As an example, many people seem to think that ‘explorers’ just want to explore places with a physical topography. In actuality, most explorers are just as eager to explore game mechanics, and can be satisfied with an excellent combinatorial advancement system. Killers aren’t just player killers – they are players that like competing in any arena. And socializers aren’t just limited to people who want to cyber in Silvermoon City – it includes any group of people who primarily play to play with other people. Raiding and dungeon running actually scratch the socializer itch.

7. Don’t think killers are griefers. An important subset of the previous point. I see this all the time, an assumption that killers are all griefers and should be removed from the game. Killers are people who are primarily motivated by schaudenfreude. Griefers are a subset of this, but it also includes people who like clean, healthy competition as well – as long as they win.

8. Don’t make features for just one quadrant. Features that are designed for just one of the player types tends to feel odd and tacked on to the other three player types (I’m looking at you, Blood Type Comparison in Star Wars Galaxies). Strong features appeal to multiple player types simultaneously. Consider, for example, Raiding in EQ and WoW (a feature oft-maligned by armchair designers):

  • Achievers enjoy the loot, which provides a sense of progress and increased power.
  • Explorers enjoy seeing new and interesting locales, and being forced to innovate in terms of solving the ‘puzzle’ that is beating a difficult boss encounter with their own party’s mix of capabilities.
  • Socializers enjoy having an activity that not only encourages, but mandates, extensive teamplay, coordination and communication (as well as an excuse to be in close proximity to their guildmates.
  • Killers have a metagame to appreciate. Yes, it’s not as visceral as laying the smackdown directly, but there was undoubtedly a bit of the competitive juices when my guild got the server first kill on Hydross a mere 2 minutes after our competitors wiped on him.

This is not to say that raiding appeals to everyone – it clearly doesn’t. However, for the people that do enjoy it, it is firing on multiple pistons at once.

9. Don’t assume that Bartle’s is ‘the answer’. It is not a blueprint on how to make a game, nor is it the definitive answer on how all players behave. It is, however, a very powerful and descriptive game model, and is a useful lens with which to examine your own game, to see if you’ve wandered off track. Bartle’s work should be used in conjunction with other gameplay models, such as that of Nicole Lazzaro, Scott Rigby, and Amy Jo Kim. Not to mention your gut.

Original comments thread is here.