« Pseudo-Live Blog From U Minnesota | Main | Why Are In-Game Economies So Hard To Get Right? »

Feb 24, 2004



How do we define group?

There seem to be formal groups – clans, and informal groups – people that one happens to play with semi regularly.

I feel that in general formal groups could be characterised as Achievers or Killers.

Whereas informal groups can be Achievers, Killers or Explorers.

Can a group be a Socializer, I wonder?



I see one axis: groups that meet for internal causes, and groups that mean for external causes.

Former type would spend time doing Four Types stuff within their own setting (probably); Latter type would be moving and shaking the world, however you define "world". Other players, other clans, setting up events, etc.



1. Guilds for long-term goals.
2. Ad hoc parties for short-term goals.

We can probably do a personality map by using the median of individual personality profiles.

This is my cynical list:

1. Monopolists
2. Mercantilists
3. Gangsters
4. Masochists
5. PlayerZ (with a capital Z)

By L50 there are no longer victims; they are all professional players in the game of virtual life. PlayerZ is the catch-all category for the non-classifiable.

In Asia, many virtual guilds are simulacrums of RL gangs. Cultures in Eastern Asia are communal and achievement oriented. In urban areas, population density is high. In an era of youthful individualism, competition in social and achievement areas is equally high. Stars will fly high while others are left to carve their niche of self-worth. The gap between the have and have-not is great. The impact of the genocide on middle-class professionals and technocrats over the last century is even greater. Gangs ands guilds are the groupings many find their support and belonging.

OK, what’s my point with this amateurish preamble?

In Asia, MMORPGs are cheap escape from the harsh realities of RL. Moreover, with the organization gangs and guilds can provide, you can make a better living and definable progression of achievement in VW than RL.

Therefore, by L50 all groupings are professional workers, either freelance or salaried.

Before L50? They're just working their way to L50 :)

Oh, groups can can be primarily socializers too, especially in SWG.



In Eve-Online, "corporations" are roughly equivalent to "guilds" in Everquest, say.

From the I-Eve website newbie guide on Corporations (Eve-Online) we see this proposed taxonomy of Corporation types:

1. Mega corporations :
2. Pirate Corporations :
3. Laid-Back corporations :
4. Manufacturing Corporations :
5. Bounty Hunters :
6. Capitalist:

(see http://www.eve-i.com/article.php?id=17 for full text)

These types of corporations "emerge" from the player community in the sense that there is nothing in Eve that says "Laid-back corporation" or "Bounty-hunter corporation", although the game design does channel behavior along certain activity lines: efficient money making, and flavors of PvP.

With a bit of squinting, I think these too can (sorta) map into Frank's list:

1. Monopolists = Mega Corporations
2. Mercantilists = Capitalist Corporations
3. Gangsters = Pirate Corporations
4. Masochists = Bounty Hunter Corporations
5. PlayerZ (with a capital Z) = Laid-back Corporations

Are commonalities (across games) in the ways players like to express themselves in groups.



Nathan, it is interesting to know, but not surprised, that my cynical list can be mapped into existing games. I wrote the list in half-jest. Maybe I'm showing too much angst over my love/hate of VWs :)

The dynamic I see in MMORPGs is that in a sci-fi setting, players are expecting that social and economic modes are advanced. Therefore, social interactions in groups reflect this advancement.

However, in MMORPGs that have a medieval setting players expect on one hand the modes befitting a medieval civilization, but the efficiency of a modern system.

In desigining VWs, how do we predict group dynamics for a particular design choice? Telepathing Nathan: are commonalites across games in the ways players like to express themselves in groups predictive of group dynamics in new VWs?

Any prior research on this?

Personally, the smaller the group I'm in, the more specific the dynamic of the group. The bigger the group I'm in, the more general/common the goals and dynamic of the group. The minority gets marginalized and drop out to form their own groups.

I think personality enneagrams can augment Richard's personality types. I think the buzz now in the biz sector is corporate enneagrams and personality test as ways to understand team dynamics and manage hiring.



Ren>How do we define group?

I was thinking mainly in terms of the smallest units of more than one person that get together on a regular basis, but if useful generalisations can be made about larger groups, that's good too.



> Can a group be a Socializer, I wonder?

I've seen examples of this in SWG. There are a few musical bands that play gigs at player owned cantinas. The goals of the groups I've met tend to be providing a backdrop/boost for cantina roleplay drama. There are a few traveling bands that set out 'on tour' to meet new people as well.


Richard wrote
>>Ren>How do we define group?
>I was thinking mainly in terms of the smallest units of more than
>one person that get together on a regular basis

Define regular :)
But seriously, I was reflecting on the way that I spend time in VWs, well in SWG mainly and I guess I spend time either with random hunt groups that are put together on the spot, or with a nebulas group of people where we may hunt or just hang out, so the full set seldom if ever gets together but I frequently group with subsets with a verity of purposes – we are loosely coupled as it were. Maybe this just all falls into Socilizer, but it is the irregular nature of the grouping that gives me pause before applying a type to them.



> In other words, are there "group types"
> like "player types"?

Something like:

A. Couples - Fairly exclusive, rarely group with others - though they do occassionally. When they do group with others, it's almost never with complete strangers. These seem to be more common in UO and SWG than in EQ, where it's a burden to remain exclusive beyond mid-ish levels. Though "romantic partners" fall into this category, it's just as likely these are merely partners.

B. Ad Hoc - Groups consisting of players that never seem to mind grouping with strangers. I wonder if they only join these sorts of pick-up groups until they find a partner (or larger trusted network of friends such as a guild, etc) that they want to play with all the time. The least-likely bunch of people to be allowed to group with couples (whom are the most exclusive of the groups).

C. Group is sort of like the Ad Hocs, in that they group with "whomever is available", but usually only from a trusted subset of the population (i.e. their guild). Sometimes they let a B join.

D. Basically the couple plus some number of trusted allies: Probably people from the same guild or town or chat-channel or some other meta association.


*A* logs in and looks for his partner. They group-up and run off to play with each other.

*B* logs in and stands around yelling "ne1 want to group!?"

*C* logs in and hits the guild-channel. "Hey guys, what's everybody doing tonight?" Will hook-up with a B if he has to.

*D* happens when either B or C is allowed to join A (and it's hella more likely it'll be a C allowed to join them than a B).


One type of group that I think got labeled above (but I'm not really sure) is the group that simply is out to help each other advance in the game via character development. Is this the 'achiever' group some people have identified? Back in my UO days, I would consider a variety of groups to fall under achievers:
- Achieve money/items
- Achieve 'rares' or items of scarcity worth a lot of money
- Achieve maximum character skill sets

Personally, I had 3 different groups for each of these in UO. I had a guild that I hunted with (both NPCs and players) for money and items. I had a small group of people (non of whom were in my guild) who I would go rare hunting/exploring with. Then I had another group of about 3-4 other people who I would macro with almost every night. There was some cross over between my guild and macro friends, but only a little. I think that macro group sometimes goes unoticed...although I'm not sure how current games handle macroing, or if macroing is as big of a problem in current games as it used to be in UO.


From a personal experience, I find that what Jeff described is often the most common occurance for me and my friends. I am most definitely in the "a" category that he mentions, often avoiding those that would be even in my own guild.

When reading through this and the comments I thought that it was perhaps being missed that a 'group type' might just be the sum of the 'player types'. While everyone has a scoring method and rarely forms up so extremely strong to any one type, what if the group adopted the dynamics of averaging all its sums?

A group that was made up primarily of socializers becomes the 'laid back' group, taking things slow and socializing more than worrying about the track to advancement. A group of explorers being more likely to move around with the combat as they become bored of a surrounding. Is there some possible way to track this idea based off of player types?

"Is there some required 'long time reader, first time poster' statement required like calling in a talk show here? =)"


Jeff Freeman>Couples - Fairly exclusive, rarely group with others - though they do occassionally.

Are there "atomic" groups of three or more players, or is the couple (which I take to mean 2 players) the only tightly-knit long-term group?



Richard> Are there "atomic" groups of three or more players

I don't think so.

Richard> or is the couple (which I take to mean 2 players) the only tightly-knit long-term group?

I do believe two is the limit, long-term.


Well, what my guild usually sees in games is the other groups are busy killing monsters, harvesting equipment and money, and not watching their backs.

We're usually out looking for the groups above in order to forcibly seize their assets :) Long live PvP :)

That would make us the "hunter / gatherer" group.



Felt like naming an unnamed type in Freeman's: C is "Club Members", where everyone's got a pass into a little club -- nothing big -- and they'll hang out with whomever's around. Occasionally, they'll bring a guest or a stranger will be around.

Is there any other kind of categorization of groups other than willingness to group with one another? (If that's a good description at all.., which I'm not sure of.)


One of the reasons players cite for playing a virtual world when they should be doing other things is so they don't get left behind by their friends. Similarly, one of the reasons they cite for buying characters from commodifiers is because they've not played in awhile and their friends have got ahead.

So who are these friends they're talking about? Using Jeff's categorisation, it seems that only players of type D would be affected.



to Richard and Jeff --

There are "atomic groupings" larger than two, at times. Weekly hunting groups that meet same time each week, for example. Then there are groups that fall between A and C; I'd call them cliques. Guild officers in my guild have tended to be this way; we ask what officers will join us before anyone else. Right now I have a group of four that have been playing together as often as possible for a few months now. In two weeks this group of four is getting together for a face to face visit.


My short answer is that I can't offer a typology because I haven't spent enough time studying various groups.

Somewhat tangentially, I think the typologies are interesting, but more as an instrument for thinking than an empirical tool, right?
Typologies remind me of, e.g., the 12 Signs of the Zodiac, the 4 humors, the 3 body types.
Which is not to say they aren't useful -- see, e.g., all of taxonomy, which is an attempt to break down biodiversity formally...


Greglas>Somewhat tangentially, I think the typologies are interesting, but more as an instrument for thinking than an empirical tool, right?

They can be used for both, so long as you bear in mind their origins (and therefore their limitations).

The reason I was asking about types for groups is because I wanted to explore a link between levels of immersion and levels of community. In the West, players tend to play as individuals, grouping together for convenience. Communities do exist at deeper levels, but seemingly not in any formal way.

In Asia, things are different. From what I can tell, players often play together as groups, and they often seem to regard the group as the important thing, not themselves as individuals. In this sense, the group is acting as a coherent individual with a number of effectors, each of which is under the control of a player. Such "atomic" groups seem to behave as if they were ersatz players, and their progression to deeper levels of immersion are felt by the players as deepening levels of community.

Most of these groups in Asia seem to be achiever type, from what I can gather. There are some explorer type groups, and some (but not as many as you might think) killer type groups. Socialiser groups, though, are very few and far between from what little I can divine from reading stuff on the net.

I was thus wondering whether groups of this kind were extant in the West, and, if so, whether any of these were of the "hub of 30 other groups" socialiser types or whether they were the same achiever types seen in Asia.

From what people have been saying, it seems that groups are as loosely-knit as ever in the West, so it's not really necessary to provide group-oriented content in new virtual worlds aimed at Western tastes. In Asia, though, group-oriented content is much more appropriate. The player type model, applied to groups, might help designers in considering what content to add; however, given the large number of achiever groups it would be hard not to design content that suited the majority of them anyway!



Dee> Guild officers in my guild have tended to be this way; we ask what officers will join us before anyone else.

I see that as a type-C. It's just that the subset is within a larger subset. I would be surprised if there isn't a partnership or two within that subset, too.

I have to admit, it's got to be *possible* for there to be a type-a group with more than two people - I can't think why there wouldn't be such a thing - It's just that I've never seen it. So my impression is that it must be a rare thing if it exists at all.

Richard>So who are these friends they're talking about? Using Jeff's categorisation, it seems that only players of type D would be affected.

Both type C and type D would be affected, which I think is where most players find themselves, most of the time. Even people whose preferred group-type is A will wind-up in a C or D from time to time.

But I also think individual players migrate from one group-type to another out of necessity / availability. e.g. if your partner isn't online, maybe you look to the guildchat for some group.

'D' might be a superfluous categorization, or perhaps is too narrowly defined. Or maybe it's not really a category on its own right, but is the result of any mix of A's, B's and C's. It's not anyone "preferred group type", being more the product of compromise.

'Haven't really put a lot of thought into this. Just thinking out loud, really, so to speak.


Richard, great thoughts and questions.

I would love to datamine the server logs to get accurate data. Everything so far seems to be based on different perception and perspective.

Play groups in Asia are generally social achievers. When competitive density is low, they can be explorers. When the competitive density is high, however, they turn into killers.

A more communal culture and actual RL interactions in game centers are two key factors that strengthen group cohesion in Asia. Also, to a lesser extent, the historical and culture ideals of “brotherhood” influences group types. Bond are strong and hard to break.

If western culture can be categorized as individualistic and perhaps self-centered, then the act of divorce or bond-breaking is not a major life-decision. So, perhaps the newest player-type could be Freelancer.


P.S. Sorry I am too brief on the 3rd and 4th paragraph. Happy to expand upon request.


magicback> I would love to datamine the server logs to get accurate data. Everything so far seems to be based on different perception and perspective.

I prefer the term "Observation". :P

magicback> So, perhaps the newest player-type could be Freelancer.

I keep having to fight the urge to include the solo-ists as a group-type. There are people whose preferred group is "none". So far I've managed to resist the temptation because "group of one" is an oxymoron.


Anecdote: There's a player I know who so strongly prefers the "couple" groups that he is half of about five couples at any given time. Ok, I'm exaggerating but not by much. He pairs off each of his characters with a character belonging to one of his friends and always plays it with that friend for weeks or months on end. Eventually most of these pairs break up and he either finds that character a new partner, or deletes it and makes a new one. He's deleted more Everquest characters in their 50s than most players ever level up in the first place.

I don't know how long term is "long term." You could view a group of three or four where, whenever any two or more of them are on, they always do something together, as an "atomic group." But you could view it as several pairs of hunting partners who also group with each others' hunting partners preferentially.

When you go above 2 people, it's hard to differentiate cleanly between atomic units (eg couples) and cliques (eg guilds).

Another anecdote: A group I've adventured with a few times contains members from several different guilds. They call themselves "team green." All of them dyed their armor or robes green. Their leader is a level 58 who has refused to go up any more levels. There are about five who go on at least 50% of his outings. Then there are more peripheral folks like me who go along on a few because they made friends with the leader or one of his core five.

So that is not "choose randomly among preferred group of people" (c) nor is it "only adventure when these specific people are around, and preferably no one else" (a) but somewhere in between.

My own grouping preference fails to fall helpfully into any category discussed, but I think it's a very common one.

I prefer to group with people I know well and know I enjoy playing with. I like it when we don't have to recruit any outsiders to play important roles. I also like going along with chance-made friends on exploratory adventures to areas I don't go with my regular crowd because we don't know our way around there. I also like seeing if I can make new friends by grouping with complete strangers.

My preference usually goes: person I live with > person I have played with and like how they play > person I have played with and like as a person but they play a way I'm not so fond of > stranger > guild mate I don't really get along with > non-guild-mate I've had bad experience with. Sometimes the stranger slips down a peg when I'm doing my duty to strengthening guild bonds (ugh).

Sometimes I feel antisocial and stranger > all because I don't have to be my typical self around them or have to deal with "what's wrong?" type of questions.

At this point I feel bad for tearing down an earlier suggestion without giving my own.

Perhaps we can categorize groups a different way. What about:

"adventure group" would be the most common in Everquest. It would have three flavors:

- "friend adventure group" group of those who mostly know each other and usually play together, formed just for a day's fun (minor achiever treadmill stuff + socializer, though depending on makeup of the individual personality, you might get features of exploration too or even griefing if it was a race-to-the-spawn type group going to kill the rare mob before anyone else does)

- "stranger adventure group" group of those who don't know each other and if they've played together before it's a pure coincidence, formed to make minor progress toward separate individual goals (minor achiever treadmill stuff, possibly some socializer if anyone joined it to make new friends)

- "blend adventure group" a couple or whatever picks up some additional people (cross between two above)

Then there's the "help someone do a quest" group. Usually this group contains one person who is doing a quest, and several others who are there just to help them out. Mostly it has to be friends or partners; sometimes, though, there are quests that give a chance of reward to the helpers, and sometimes you get strangers assisting on those. Or bored strangers who will help anyone for something to do ;)

(You can also have groups where the whole group is doing the quest. Trials groups fit this category, or a group all hunting drops for the same quest at the same time. Not sure if this is a separate type of group or just another "adventure group" with a more specific bit of the treadmill than just more experience. Groups camping drops are another subtype of that. I think the only one that is truly a separate *type* of group is where the whole group is there just for one person's quest or goal.)

In Everquest there would also be the "scout group," less common but seen often enough. Its most typical subtype would be:

- "guild scout group" a group who is setting out together on behalf of a larger, more official grouping to enable some goal of the latter. Examples might be checking for a spawn being up, or learning strategies and difficulty level for a new zone (exploration / competitive with other guilds)

In SWG or perhaps other games with PvP elements, you might also find:

- "mutual defense group" a group that gets together to defend one another from attack. This would be in a game with PvP. It might have flavors of friend/stranger/blend/guild especially in games where the PvP is between teams. In an everyone vs everyone PvP environment, it would probably be friends only -- who else would you trust to watch your back?


>Is your group doing basically the same kind of >thing as these other groups? Or will different >groups of friends focus on doing very different >things? Do these things change, depending on >the level at which the group operates?

In my experience the difference concering what type of activies groups do together is highly dependent on the guild/allegiance that these people are members of. (assuming that the majority of these players are part of the same guild, ie has done the *C* stage of collecting friends that Jeff talks about)

In Asheron's call 1 it was (in my experience) quite common that ppl basically interviewed other players who tried to recruit them to an allegiance about what type of gameplay they used to do. If they were adventuring - how many times together a month etc.
In acherons call 2 the gameplay itself is not so diversified, so there the questions are (in my experience) quite simplistic: Do we do PvP? Can I be of Dominion/Order and still join? The (comparatively) indiverified gameplay results in players doing similar things: doing quests and hunting.
But then again - SWG! This week i made a new char, an entertainer. As Bob says there are these traveling bands doing gigs, id love to see one! My experience was somewhat disappointing. The grouping with other entertainers results in more experience points, but in my case it was mostly a chat room. Beautiful mocap animations tho :). I read some dev notes where players had asked to be part of the story arc... and browsed player forums where my overall impression was that players are a bit disappointed in how the entertainer class is intergrated (ie not integrated) in the wholeness of the game.

However - i've been thinking of the composition of small groups. Probably most of us have made personality tests of the Meyer-Briggs type. When playing AC2 i have been asking players about their choice of character class, if and how the choices are affected of their conception of their 'self' in their everyday lives outside the game. This is of course totally informal, but there seem to be correlations. Furthermore, if a player has several player characters the playstyle and conception of the playing self changes depending on character. (Richard B's chapter on players is btw so amazingly useful, thanks :)) Anyway - I toss with the idea to make a system which would support grouping based on not only activities but on complementary personality types. These would be expressed by character properties... but i have no idea on how to actually support the "ideal grouping". Might be a risk of acting big brother. What do you think?

/Mirjam Eladhari


Mirjam Eladhari>I toss with the idea to make a system which would support grouping based on not only activities but on complementary personality types. These would be expressed by character properties... but i have no idea on how to actually support the "ideal grouping". Might be a risk of acting big brother. What do you think?

On the one hand, I can see how players would be very happy to be dropped into a group and find that they immediately got on with everybody else. On the other hand, it removes an opportunity for personal development: if everyone thinks the same way as you, you have no reason to change (or to try to change them).



What Mirjam’s post made me wonder was whether either consciously or un-consciously devs were using the kind of Meyer-Briggs type work on what makes an ideal group is used in the construction of character classes and more specifically the mix of classes that are needed for multi-person quests.



Ren>whether either consciously or un-consciously devs were using the kind of Meyer-Briggs type work

There may be some of this going on, but on the whole I think not. Much design seems to be incremental and evolutionary: take what works, and throw some new stuff into the mix that you think might be better. I guess there could be room there for "we need a class that does this", but I suspect that it's likely to be some way behind "wouldn't it be cool if we had this!".

Although I wouldn't like to see paint-by-numbers design, where people just apply a theory they don't necessarily understand rather than trying to advance it, that doesn't mean there isn't a significant role for such theories to play. It always helps to know why something works, rather than merely knowing that it does work.



Looks like there is a project in there for someone who understand the personality type combination and groups stuff to do a mapping that links from indivduals and how the fit these types, through character classes, the groups that people tend to form and the groups the the game requires them to form.
Comeon psycology depts, you know you want to.



>rihard: On the other hand, it removes an opportunity for personal development: if everyone thinks the same way as you, you have no reason to change (or to try to change them).

Usually groups with a positive dynamic consist with people who are different, and therefore complement each other (or not :). Groups consisting of very similar personality types tend to not work out so well in the long run - one is decieved in the beginning by "oh how very alike we are, we probably ´can work very well together" Groups with different, and complementary persons tend to be more dynamic, and indiviuals learn more about themselves. In short: more exiting, more difficult, higher challenge with the possibility of more interesting group dynamics. Not reaching equilibrium/static state with the effect of being bored and dissolving group... Or having so similar motives that a competition for certain outcomes (groupwise) becomes a problem.
So the experiment would be to have a simple profile of what role a person might aspire in a group, and fascilitate group formations with complementary roles. This would be easier in a VW... at least if you compare with everyday life where groups in for example work environments are fairly static. Oh i whish i had more time :)

The comments to this entry are closed.