Achieve, Create, Socialise, Immerse
I've decided to take a break from designing games that primarily interest me (intricate, elegant, abstract, mechanics-driven games) and turn my attentions towards designing games that primarily interest other people (???). This is partly because getting out of my comfort zone is necessary if I want to expand my skill set, and partly because finding like-minded playtesters is proving a bit of a burden.
The analytical side of me should still be useful, though. Game design is still an optimisation problem, at it's heart – the objective you're trying to maximise ("funs-per-second") just depends on your target audience. It's easy to design for yourself, as you know exactly what you like, and can use this to guide your iteration. If you want to design for other people, you have to be aware of what they value, which is harder – and can differ wildly from person to person. What would make things easier is some kind of categorisation system for player motivations, so you can design for the categories rather than for individual people.
Thankfully, other people (designers and researchers) have already done the hard work and extracted this information for me. Here's what they found out!
Bartle's Taxonomy of Player Types
Richard Bartle, a veteran designer from the massively-multiplayer/online gaming world, theorised that players can be broadly sorted into the following four types:
Achievers This group is happiest when achieving the goals of the game. They like clearly defined challenges of an appropriate level, and enjoy overcoming these. This might be as part of a solo venture (e.g. levelling), through cooperation, or by defeating other players, but it's the joy of winning that appeals to them most.
Explorers This group wants to learn more about how the game works, and discovering all it has to offer. They like uncovering new information and being put into unfamiliar situations in order to figure out optimal strategies – possibly by trial and error. Winning is less important to them compared to experiencing something new.
Socialisers This group is happiest when building positive relationships with other players. Cooperative and team games appeal more to them over direct competition, and of the four types they're less interested in the actual game more than what goes on outside it. They value everyone having a good time, and will often go out of their way (in-game) to encourage this.
Killers This group is happiest when proving dominance over other players. If the game is purely competive, then this aligns with the Achiever type, but outside of this it differs because winning isn't the prime motivator – instead, it's about doing better than others. This isn't necessarily by defeating them – in purely cooperative games this can be expressed by contributing more than other members of the team. The name "Killer" is a bit problematic in this regard, because of the clearly negative connotations, but I'll stick with using Bartle's term for consistency.
Bartle seemed to suggest that these categories were mutually exclusive – that is, a player who scores highly as a Killer would score very low as an Explorer. However, subsequent research (#foreshadowing) suggests that this often isn't the case - in a lot of cases players will straddle several of these categories, or have neutral responses to many of them.
Rosewater's Taxonomies of Player Types
Mark Rosewater, head designer of the Magic: The Gathering TCG, writes a lot about design – primarily about his own game, but occasionally on topics with a broader scope. One of his most important contributions to the design of this gaming behemoth (in his opinion, as well as my own) is the realisation that its players can be divided up into three main categories, based on their different approaches to extracting pleasure from this incredibly complex game system. These are:
Timmy This type of player is all about maximising experience – performing the coolest actions, getting to the most ridiculous gamestates, and generating the largest reversals of fortune. They don't particularly care about winning, or being better than others, they just want to do awesome things (either by themselves or involving others).
Johnny This type of player is all about expressing themselves within the game system – trying out strategies that no-one else has attempted, or setting themselves arbitrary restrictions and trying to maximise winning despite these constraints.
Spike This type of player values winning above all else – the game is a mental competition, and to view it in any other way seems absurd. Winning usually means someone else has to lose, and the pleasure can come from either the victory itself or the domination over the opponent. They value optimal strategies and efficiency over all else.
These don't neatly align with Bartle's types – sure, Johnny is mostly an Explorer, but has elements of Achiever and Killer; similarly, Timmy is mostly a Socialiser, but has elements of Explorer and Achiever. In addition, it's quite common in the Magic community for people to identify as a hybrid of these types – for example, a Johnny/Spike. Taken together, this suggests that both Bartle and Rosewater's categories aren't rigidly defined buckets in which to throw players, but more like focal points in some "motivation space" – emphasising certain traits and deemphasising others.
More recently, Rosewater has promoted a parallel categorisation for a player's appreciation of a game's mechanical aesthetics:
Melvin This type of player appreciates mechanical design – elegance, intricacy and depth. They prefer game mechanics that are trying to explore cool mathematical concepts, independently of the in-game flavour (if it's present at all).
Vorthos This type of player appreciates flavourful design – visuals, setting, and narrative. They prefer game mechanics that evoke the flavourful concept they're trying to imitate, regardless of elegance or complexity.
These are more about appreciating a game from afar, rather than actually playing it, so are probably less useful for immediate design. However, they are presumably useful ways of thinking about how people will think and talk about your game when they're not playing it, which is necessary if you want people to recommend it to others or decide to try it in the first place (i.e. "this game is cool because (a) it has a cool trading mechanic! / (b) you get to be a dragon!").
Yee's Taxonomy of Motivations
The researcher Nick Yee noticed the categorisation problems with Bartle's model, and investigated them using data from a large sample of MMORPG players. He argues that player motivations can be broken down into ten main bimodal components (i.e. can broadly divide players into valuing these attributes as "important" or "unimportant") – grouping these under the three main categories of Achievement (A), Social (S), and Immersion (I), as there was little correlation between each of these.
The ten components were:
Advancement (A) This component emphasised rewards given to the player by the game for task completion. As this study was for MMOs, this emphasised levelling, quest completion and resource acquisition - though I'd assume that "winning" games that can be won would fit here as well.
Mechanics (A) This component emphasised the more rules-oriented side of exploration – both working out the underlying mechanics of the game and using this information to optimise personal progress.
Competition (A) This component emphasised dominance over other players. Like Bartle's Killer, this likely doesn't have to be through direct conflict, and should apply to any situation where players can be ranked.
Socialising (S) This component emphasised meeting and getting to know other players in general – being friendly to newcomers, and helping them out if need be within the confines of the game.
Relationships (S) This component emphasised the creation of meaningful, long lasting relationships with a select few other players – helping them out in more depth, including discussing situations outside of the game.
Teamwork (S) This component emphasised collaboration between players to achieve in-game goals, and a greater valuation of team rewards over individual rewards.
Discovery (I) This component emphasised exploration of the game world, instead of the game structure (as Mechanics did). Players who scored highly in this enjoyed acquiring new knowledge and experiences, whether they were regarding unfamiliar locations, tasks, or items.
Role-play (I) This component emphasised the game setting itself (lore and flavour), and submitting to the fantasy that your character exists within this world.
Customisation (I) This component emphasised tailoring the game experience to better express the player's desires. For MMOs, Yee interprets this purely as visual customisation (fancy hats etc.), but I feel this could be more broadly expanded to any mechanic where the player makes choices that represent themselves towards other people (e.g. deck construction in trading card games). As such I'm not too comfortable listing this as a pure "immersion" trait – it's possible that it contributes on several of the primary axes, if not an axis unto itself.
Escapism (I) This component emphasised the ability of the game to be disconnected from the real world, freeing players from external stress while participating in the system. This feels more of a circumstantial motivation rather than an intrinsic one – the amount of stress people are under can fluctuate, and affect how much they may want to relax. It's also not really something you can design for – it's just a property that games and most other media possess. (I guess making games dissimilar from real life would count?)
The biggest difference from Bartle's model is the fact that players can score high or low on each of these components, rather than being assigned to a single dominant one. While this probably doesn't mean there are 2^10 types of player existing with equal frequency, it does mean we have a 10-dimensional space we can use to identify clusters (in fact, the three supercomponents from Yee's data does precisely this for the MMO players, yielding at most 8 "player types"). Other differences are also present, most notably the split of the Explorer type between "Mechanics" and "Discovery", and the different types of Socialiser.
However, due to the type of game being examined, it's possible that Yee missed a few components. Marc LeBlanc has a similar list of "game pleasures", which overlaps a lot with the above (while condensing multiple components, such as [Socialising+Relationship+Teamwork] and [Mechanics+Discovery]). The most important additions seem to be:
Narrative (S? I?) This motivation emphasises drama and progression – this need not be in the sense of an actual fictional plot, but more the same patterns of interest peaks and lulls. Players who value this presumably get rewarded by "cool moments". MMOs tend to be more freeform experiences and I would assume largely miss out on this.
Sensation (I?) This motivation emphasises the aesthetics of the game – be it visual, audible or tactile (in the case of a physical product). There's a reason games aren't just a dry rulebook and generic playing pieces (except in the prototyping stage), and people will in general prefer something that looks good over something that plays well, because the former is assumed to correlate with the latter. The huge push by the AAA videogames industry to have superior visuals over innovative gameplay is proof of this.
Not wanting to be left out, Jesse Schell extends this list of "pleasures" further:
Gift Giving
Humour
Anticipation
Possibility (large action space, freedom)
Thrill (fear without danger)
Purification (tidying up)
Fiero (triumph over adversity)
Schadenfreude (pleasure in misfortune of others)
Naches (pride in accomplishments of others)
Wonder (awe at accomplishments of others)
It's true that these are all pleasurable activities (to different people), but they don't all feel like motivation axes in their own right – for example, "Thrill" seems to be a blend of Discovery and Role-play/fantasy.
The Problem with Johnny
My gut feeling, supported by Yee, is that the motivational component view is the right approach. Within this "motivation-vector-space", there would be a number of orthogonal axes – it seems likely that these will include Achievement, Socialising, and Immersion as three of the primary axes, but there's nothing to say others don't exist.
All the motivational components discussed above (Competition, Discovery, Narrative etc.) would then be directions in this space – a lot of them would align closely with a primary axis, but it's clearly possible for certain motivations to have components in multiple (e.g. Competition is primarily an Achievement-based motivation, but it has a small (anti) Social component as well. It's clearly orthogonal to Immersion, though).
Under this model, each player exists as a point in this space – they'll have different levels of response to each axis. Using Yee's three axes (A, S, I), we can attempt to fit Bartle and Rosewater's player types into the space as rough cluster centres:
B: Achiever (A+)
B: Explorer (A+, I+, ???)
B: Socialiser (S+)
B: Killer (A+, S-)
R: Timmy (S+, I+)
R: Johnny (???)
R: Spike (A+)
R: Melvin (I-, ???)
R: Vorthos (I+)
While most of these categorisations feel appropriate, this is somewhat awkward on the Explorer/Johnny types – it feels as if there's an axis missing, along the lines of exploration/creation/innovation. Yee subsumes "expression" into his Immersion component, and though he improves on Bartle by acknowledging that "exploration" has at least an Achievement component ("engineer") and an Immersion component ("traveler") the model still doesn't feel complete. Where's the scientist, or the builder?
Consider the game Minecraft. While it does have a typical RPG/levelling system (Achievement), and you can play it with friends (Socialising), neither of these things are core to the gameplay – and you can easily play it without either. What appeals to people most is the act of slowly constructing something that you can take proud ownership of, or, for more advanced players, constructing elaborate electronic machines within the simulation. This is achievement of a sort, but it's not as a result of beating the game or dominating other players – it's about trying things out and meeting goals you set yourself. With Magic, the fact that this creative motivation has warranted a distinct psychographic from the clearly achievement based Spike is yet another reason why this may deserve a fourth axis.
In the absence of further, formal research – which I lack all of the time, inclination and ability to perform – the existence of a Creativity axis will unfortunately have to remain highly speculative. But I'll be using it in my personal model regardless.
[[ It's also very possible that this low-dimension motivation space is an oversimplification, and while correlations exist between individual, specialised motivations it doesn't mean that they'll always hold true. I'm definitely someone who hates low level socialising, but places a much higher value on closer, more meaningful relationships – while this may score me as a mid-low-level Socialiser, I probably shouldn't classified in the same way as someone who prefers frequent superficial interactions and hates depth, as the experiences we value are likely to differ. ]]
My Definition
So, bearing all of this in mind, how does this help me?
I'm fairly sure my dominant axes are Achievement and Creativity. I like finding out how things work, solving problems and inventing stuff. If you've read any of my previous posts, this is probably not much of a surprise.
My weaknesses, therefore, are in the Social and Immersion side of things – I still like both aspects (in the same way that everyone likes achieving/learning/making things, to some degree), but to me they're less important. There's nothing wrong with this as a player – I get to choose what I play, after all – but it does mean that as a creator I'm probably going to be worse at making stuff for people that DO care about these, simply because I can't easily empathise with their desires and am less experienced with such products.
I'll need to correct this.
Conclusion
One of the things I love most about games is that they directly mirror social interaction in the real world. What motivates game players usually corresponds to what motivates them as people – so in a Shyamalanian twist all of the above is applicable to real life as well. Knowing your gamertype and the gamertype of others might be useful – well, as far as more mainstream personality tests such as Myers-Briggs are useful anyway.
So, I'm now going to try my hand at a more flavour-driven social game, guided by the submotivations listed above (role-play, teamwork, narrative etc.). This sounds a suspiciously good fit for the horror game idea I toyed with late last year, so that's probably where I'm going to start.