Thursday, October 8, 2009

Bartle Player Types Revisited

Welcome to my blog. I will be discussing “A” list Massives aimed at the American and European market, such as EverQuest®, Dark Age of Camelot®, World of Warcraft®, Lord of the Rings Online®, Eve Online®, Warhammer Online®, etc. Today I will be revisiting the Bartle Player Types. I believe there is a need for a Player Type test that is more predictive than Bartle.

The Bartle Test of Gamer Psychology is often used by game designers to ensure a game will appeal to the player type(s) they are targeting. Wikipedia: The test is based on a 1996 paper by Richard Bartle and was created in 1999-2000 by Erwin Andreasen and Brandon Downey. The test comprises 30 questions, each question asking the respondent “do you prefer (a) or (b).”

From Bartle’s paper: “So, labelling the four player types abstracted, we get: achievers, explorers, socialisers and killers. An easy way to remember these is to consider suits in a conventional pack of cards: achievers are Diamonds (they're always seeking treasure); explorers are Spades (they dig around for information); socialisers are Hearts (they empathise with other players); killers are Clubs (they hit people with them).”

“Naturally, these areas cross over, and players will often drift between all four, depending on their mood or current playing style. However, my experience having observed players in the light of this research suggests that many (if not most) players do have a primary style, and will only switch to other styles as a (deliberate or subconscious) means to advance their main interest.”

Bartle was and is a good place to start a discussion about player types. However, Bartle’s paper is now over ten years old. Massives have undergone many changes, and perhaps more importantly, the Massive player base has expanded from under 100,000 in 1997 to 18,000,000+ in 2008 (mmogchart.com). In 1997 we can categorize the player base as primarily early adaptors and the games themselves as “sand-box games”. Currently the player base is much more diverse and new types of Massives have emerged, such as the “theme park” game.


Bartle Player Type is Not Predictive

My Bartle Test defined me as a “Killer – Achiever,” which on the surface I would agree with. What the Bartle Test does not disclose is that I enjoy large-scale player vs. player (pvp) primarily with pick-up-groups (pugs). The Bartle Player type would not predict which game with player vs. player (pvp) I would enjoy: World of Warcraft, Guild Wars, Warhammer Online, Aion™ Online, etc. Additionally, since the Bartle Player Type is so broad, it does not give any indication of what percent of total possible players my player type represents and what could be done to enhance my player type’s gaming experience.

Bartle says “Killers get their kicks from imposing themselves on others.” A killer in this context means a player who primarily enjoys player vs. player (pvp) content. Player vs. player is a zero-sum game, or to put it another way, for me to win, you must lose. Examples of zero-sum games include: a marathon, a friendly basketball game at the gym, poker, NFL Football, a battle of the bands, even a game of monopoly® with your family. A marathon and playing poker are basically both solo pvp zero-sum games. A friendly basketball game is an example of a pickup group pvp zero-sum game. NFL Football and a battle of the bands are examples of team pvp zero-sum games. A game of Monopoly with the family is an example of a social pvp zero-sum game.

Even if we agree with Bartle that “Killers get their kicks from imposing themselves on others,” the experience of being a “Killer” can be radically different from player to player. A solo killer may enjoy dueling above all else. A pickup group killer may enjoy pvp instances, such as WoW’s Battlegrounds. A team killer may enjoy ladder-type organized pvp competition like WoW’s Arenas. A social killer may enjoy pvp paced to allow socializing, such as Warhammer’s realm vs. realm combat. These four player types: solo, pickup group, team and social are all subsets of Bartle’s Killer player type. I believe by more narrowly defining all Player Types in a similar fashion we can devise a predictive model of player types.


Bartle Question 4

Below is Question 4 from the original Bartle Test:
Would you rather:
· Know where to find things
· Know how to get things
I personally found this question very hard to answer.

Nicholas Yee writes: “The problem of employing a just-so model is that it becomes self-fulfilling. If a questionnaire is constructed such that a respondent has to choose between being an Achiever or an Explorer, then the end result will be a dichotomy where none may exist to begin with. It would be like asking - Do you prefer pizza or ice-cream?”



The Duel-Guy

I’m going to use the Duel-Guy in the next section, so let’s talk about him for a minute. In WoW and EQ a player can challenge another player to one-on-one combat called a duel. In games like Dark Age of Camelot and Warhammer Online, there are not game mechanics for dueling. In those games players use out-of-game communications to set up times and places for dueling. Some players duel to hone their skills, some for the challenge, some to kill time. I personally do not particularly care for duels, but I have witnessed dueling in all the games I have played. I know that some Duel-Guys will keep various sets of gear and such so they can maximize their chances against various classes of opponents. Anecdotally I know that there are some players that enjoy dueling more than any other activity in a Massive.


The Need for Predictive Player Types

There are four broad reasons that we all could use predictive player types
· Help a player choose a character at the character selection screen
· Marketing
· Define viable niche markets
· Fine-tune game play

Character Selection Screen. Let’s say we developed a new test to determine Player Types and offered it to players before they hit the character selection screen. We could then make suggestions on what characters they might enjoy and/or comment on selections they are looking at. For example, Bob is primarily a Duel-guy. Bob uses Professions to help get him just the right gear for Dueling. Now Bob may well enjoy other aspects of a game, but these are key for him. Knowing this about Bob, we might recommend in WoW that he choose a Druid or Paladin. In War, if Bob clicked on a Marauder, we might flash a comment that Marauders are really good at dueling but subpar at realm vs. realm combat.

Marketing. Let’s say our new Player Type test was easily available and used by prospective players. We could then aim our marketing at particular Player Types. Imagine Bob coming to your web-site and sharing his Player Type. You could then give him specific examples of why your game is superior for his particular player type.

Niche Markets. I’m going to blue sky a bit about niche markets, don’t treat these numbers as real. Let’s say our new Player Type test was industry wide and we had good data on a large number of gamers. Let’s say that the potential gamer pool is 40 million. Now if hardcore Duel-Guys were one percent of the potential gamer pool that would be 400,000. That would tell me, that a Duel Wars game would be viable. (See my discussions on niche markets for more information.) I believe that the future of AAA Massives is in creating viable niche games, and predictive player type tests are necessary to do so.

Fine-Tune Game Play. This is pretty self-explanatory. If a gameplay change would enhance Duel-Guys gameplay (at one percent of total) but would negatively affect Story-Guys gameplay (my estimate fifteen percent of total) whether to put it in or not is pretty easy. In other words it would give us a tool to help us allocate resources for maximum benefit.


Player Types

I’m going to throw out some unique Player Types I have encountered while playing. This is by no means a complete list, rather just the start of a discussion.

Raid-Guy. He likes the camaraderie of cooperative play. In another time and place he might have been a League Bowler, on a League Softball team, or perhaps a member of a community orchestra. If you think of new raids as playoff games my simile is complete. Raid-guy is happy to schedule 2-3 raids a week in advance, just as he might on a League Softball team.

Profession-Guy. He likes to be the go-to guy on crafting whether for his guild or for the accumulation of in-game currency. Although he doesn’t really enjoy other aspects of the game, he will do whatever it takes to get rare recipes and components. In WoW, professions are limited to two per character. Often WoW Profession-Guy will have a number of characters in an effort to be self-sufficient.

(My gut tells me there are enough Raid-Guys to support an “EverRaid” or “World of RaidCraft” game. Most Raid-guys like to raid 2-3 times a week, that leaves some downtime to fill. So if we added content to enhance raids available through farming materials and making them useful through Professions we might have a viable game.)

Story-Guy. He enjoys the story being told most of all. Will probably love the upcoming Star Wars: The Old Republic. Can be found arguing esoteric points of back-story on the forums. Key complaint is that he is blocked from seeing all content/story because he is not hard-core.

Power-Guy (aka Mini-Max-Guy). He wants to have the perfect set of gear for whatever his secondary interest is, such as dueling, twinking, raiding, etc. He is a guy that could not answer the Bartle Test Question 4 either, as to get the perfect set of gear he has to have almost complete knowledge of the game. He can be thought of as a data-miner, or a game-deconstructor. A quick trip to the Elitist Jerks website will give you more insight.

Challenge-Guy. To paraphrase Repo Man: Challenge-Guy spends his life getting into tense situations ordinary people run from. Challenge-Guy is leaning forward in his chair, music off, general chat off, refreshing beverage at hand, totally concentrating on the game. It is possible this complete immersion in the game is what he is looking for. Challenge-Guy is not just a raid guy. He is always looking for content to challenge himself with. More about him on my blog.

Alt-Guy. He just loves leveling. He may have dozens of different characters on numerous servers. He is all about choices in character development, and new areas to explore.

Sight-See’er-Guy. He wants a guided tour through the content. He avoids challenge. More about him on my blog.

Leader-board-Guy. Whether its leader-boards in Battlegrounds (BGs) or Damage-Per-Second (DPS) meters in raids, this guy is all about his big numbers. (insert inappropriate joke here).

Auction-House-Guy. He plays the Auction House (AH) like some people play the stock market. His motivation seems to be the accumulation of in-house currency more of a way to determine who won as a desire to actually use the currency.

Griefer-Guy, Leader-Guy, Arena-Guy, Role-Play-Guy, Collector-Guy, Open-pvp-Guy, Anti-PK-Guy, etc., etc. etc.

Take the above as just a pencil sketch of some Player Types. I’m sure you can identify more. Each unique player type deserves more than just the few sentences I’ve given them here. Also note that there are few people who are just one pure Player Type, most people are a combination of types as Bartle taught.


What Next

To study player types and produce a predictive player type test will require a team of people composed of game designers, players and a Sociologist to guide us. Additionally we will need easy access to players, so will need the support of at least one mainstream Massive company. I believe this study will help the industry as a whole, anyone else up for the challenge?