Good Grief

There are no grief players, only players that grief – discuss.

This pithy conundrum came up during an after hours discussion following Richard Bartle’s presentation at ComWork last Friday (thx to TL and all at the IT University of Copenhagen’s Games Research bit for putting on the event).

Bartle’s presentation concerned the relationship between game design and player behaviour, eventually taking into account the interrelated nature of ‘real world’ culture, virtual world culture, players and designers – phew.

During the talk he made reference to a group of players that would descend on a new MMO Beta, play it dry (my words), then move to the next. That is, Bartle seemed to be making a very strong link between people and play styles.

Flipping to Chapter 3 of Designing Virtual Worlds one seems to detect a similar strong identification of individual with behaviour – there are Achievers, Killers, Socializes and Explorers.

Which got us thinking about the difference between types of person and types of behaviour – do we have to give up on the idea that players fall these archetypes and consider the alternative view that players exhibit certain of these behaviours in certain contexts (admitting that indeed some players do tend to exhibit particular behaviours a lot of the time, but that these may be on the further edges of the distribution).

Would such a move enrich all types of analysis of virtual worlds or is it in fact the very same view that Bartle is espousing and are we just counting angles on the end of a pin here.

To give a concrete example of why I think this might matter – in design terms it seems to boil down to whether one attempts to design grief out of a game (change people's behaviour) or to discourage griefers (change the people).

Comments

[Nathan Combs]

1.

in design terms it seems to boil down to whether one attempts to design grief out of a game (change people's behaviour) or to discourage griefers (change the people).

An excellent question - which easily expands to behaviors in general, beyond griefing. My take is that most players transition, more or less predictably, between a number of states they find pleasurable (dependent upon context). These players are, more or less, responsive to the shaping influences of game design. However, I also believe there are pathological groups - outliers who are unresponsive to game design reward systems.

[Staarkhand]

2.

Personally I think that there is a certain mentality of person who derives pleasure simply from tormenting others (i.e. there are grief players). It's not limited to games (evidenced by younger siblings and men wearing only a trenchcoat), but the relative anonymity can bring it out. It seems that as long as players interact some will find a way to annoy, pester, inhibit or otherwise degrade others' experience.

The best approach is to empower players to find, form and develop intentional communities of like-minded players, and to establish and enforce their agreed upon codes of conduct.

It's quite possible that there are some "oportunistic" grief players, who are driven towards grief play by poor game design - they're bored, or the game is so confusing that it's easy to swindle others. But I feel in the end game design can only limit the effects, not the existence, of grief play.

[Dave Rickey]

3.

I've sung this tune before, but not in a while and not here: In my opinion, Bartle's "Killers" are poorly labelled, and should be referred to as something like "Controllers", as their motivation is to excercise control over other players. Most of them do so in fairly inoffensive ways, intra-guild and inter-guild politics. A small percentage (1-3% of the total playerbase) are pathological in their methods, trying to excercise control by motiveless player-killing, harassment, or whatever methods are feasible in the game system.

To put it another way, griefers are to "Killers" as 80 hour a week power-gamers are to Achievers: extreme examples of a common type. Both are highly visible and to a degree undesirable denizens of the worlds, but most of the low-cost methods of discouraging them fail because they act against the same capabilities that the less extreme depend on for entertainment.

--Dave

[Peter Ludlow]

4.

Interestingly, I just posted an interview with one of TSO's more notorious griefers, Grannie Celestie (yes, of Daily Show fame), who has departed TSO of SL with a group of other griefers and scammers. Odd thing is that s/he isn't a griefer in TSO. None of them are. I really think this comes down to game design. Here is an interesting passage:

urizenus_sklar: Celestie, I hear you are leaving TSO, maybe, for a while, maybe. What's up?
granny_celestie: Well, these past few months have been dead. Why invest my time in something that is dying?
urizenus_sklar: Well, have you found something to do instead?
granny_celestie: I'm experimenting at the moment.
urizenus_sklar: experimenting with who or what?
granny_celestie: I'm experimenting with Second Life with Mimi E aka. Gladys, Nick W, Kimmers, and Bradley.
urizenus_sklar: the whole crew is going to SL?
urizenus_sklar: Are you going to be scammers there or do something else?
granny_celestie: We've cleaned up our act, no one knows who we are there.
urizenus_sklar: Does that mean you *don't* scam there?
granny_celestie: Correct.
urizenus_sklar: Well, how come you couldn't stop scamming here and don't feel like scamming there.
granny_celestie: It's so easy in TSO.
urizenus_sklar: (and you did try to stop doing it here no?)
granny_celestie: Yes, I tried to stop scamming.

[Mike Rozak]

5.

Sometimes it's easier to analyze human behavior by thinking about them as primates (such as chimps).

If griefers are put in terms of animal behavior, you could compare them to social animals that assert dominance within the group.

In primate communities, there are many ways to assert dominance. Sometimes it's violent, sometimes political/social.

Dominance usually has a social ladder, so stonger members will assert their dominance over the weaker ones. Members who are close on the ranking are more likely to get in a fight, while those of clearly-different ranks results in the weaker one retreating before anything violent happens. A much weaker animal may completely avoid the stronger one, preventing all conflicts.

Dominance/violence shows itself when:

- Two members are closely ranked, so fights must be used to disambiguate.

- A desired resource is scare (food, mating, territory, etc.); the dominant one gets it.

- Invasion of personal space.

- Boredom? Maybe this isn't the right term. If there is an ongoing dispute, it will be forgotten about by both parties when something interesting is happening (such as a nearby predator). However, when nothing is happening then working out social disputes rises to the animal's mind.

- Sub-adults (teenagers) will spend more time vying for dominance because their place on the social ladder isn't yet established. (And they're constantly getting larger/stronger.) They first start dominating those younger than themselves, and gradually work up to the adults.

- In males, dominance is linked to testosterone. I don't recall hearing what hormone is associated with female dominance. Human teenage males have higher levels of testosterone than adult males (as I recall).

Some ideas for controlling grief players may come out of these observations.

[Nick Yee]

6.

This is referred to as the trait-situation debate (reasonably good review here) in Personaity Psychology, which in turn falls under the more general Nature-Nurture debate. Most personality psychologists tend to be Interactionists these days - believing that traits interact with situation, and that salient situations (job interviews) influence behavior more than mundane situations (dinner with friends). But since traits are easier to study (stable, and accessible through questionnaires), the literature is still very trait-focused.

What is clear is that accepting the influence of situations doesn't mean we have to discard traits altogether. And vice versa.

In terms of the current post, I'd like to think of it using a threshold model. Every player has a certain propensity to "grief" (or "relate" or "achieve"), and the environment has a set of constraints that makes it easy or hard to grief. The two interact to create your current pool of overt griefers.

It's like at the supermarket. There are things I'll always buy unless the price skyrockets (milk). There are things I'll never buy no matter how cheap they are (liver). And then there's the set of things I just might buy if they go on sale.

So if I had to rephrase the "pithy conundrum", it would be - players become griefers depending on the constraints of the mechanics.

Of course, every user operates on several propensities, and they stay with the game as long as you don't make it hard to do everything they like.

[Jonas Heide Smith]

7.

REN SAID:
"in design terms it seems to boil down to whether one attempts to design grief out of a game (change people's behaviour) or to discourage griefers (change the people). "

One of the central questions to my research project, so thanks for bringing that up.
But I'm curious, Ren. It's obviously a theoretically important issues when wrestling with deviance/griefing. But from the designer's point of view how are the two design approaches different? Designers, after all, experience player behaviour and not "the people" in some internal/phenomenological sense.
Could you give a few examples of how one might direct one's design efforts at either of the two?

And thanks for great discussions at the Comwork seminar to all who came.

[John Prevost]

8.

Griefing is also *sometimes* in the eye of the beholder. Looking at games with various types of player vs. player competition, you have (loosely), people like the following:

1) Engages in "fair" player vs. player combat--that is to say, not 'ganking', and not taking the fight to people who are clearly not interested. An example of this from SWG might be a player who does not immediately engage a player of the opposing faction immediately upon sighting that player. In Lineage 2, this might be the kind of person who lets a non-aggressive "red" (player killer with bad karma) character go past, assuming that the player may have had a reason to PK. In any game, this type of person is likely to avoid fights with characters who are clearly outclassed.

2) Engages in "anything goes" player vs. player combat--if someone is flagged as an enemy, they're considered fair game by any available tactic. This type of player is likely to jump any available target without worrying about relative strengths. In general, this type of player will be seen as a griefer by players of type 1, due to use of "cheap" tactics.

3) Actively seeks out weaker opponents, or opponents that are clearly not interested in participating in player vs. player combat. These are the types that become very quickly labeled as "griefers", because they're so obvious. Typical behaviors include both preferring to attack characters that are much weaker than their own and actively attempting to bait such characters into being flagged as an enemy. In Lineage 2, these are the types of people who attack players who are just starting the game, or who get together gangs of friends to pound people who are unwary enough to be baited into "turning purple" (allowing other players to kill them without acquiring bad karma.)

I'd say that it's the type 3 folks who are generally considered to be a danger to virtual communities--the L2 example being quite a good one. By placing the weakest (and most likely newest) players in uncomfortable positions over an over again, this type of player will drive other players away from the game.

This same sort of scale (from a sort of latter-day chivalry to ruthless brutality) can be seen in other realms of competition, by the way. Economic competition (specifically in trading) is a good example:

Type 1 players will attempt to set a "fair" price based on their costs and needs. If a resource is very rare, they are not unlikely to use the going market value, but normally they will price based on their own need and the other party's perceived need--as long as they have enough to supply themselves. If they are offered an "unfair" price for something, they are likely to correct the other player.

Type 2 players will attempt to gain the highest profit possible in the current market situation--most likely by a combination of hoarding and aggressively selling at the highest price (and buying at the lowest) they can get in the current market. If offered a seemingly-unknowing trade which is slanted heavily in their advantage, this type of player would be likely to jump on it.

Type 3 players are somewhat more restricted than in a typical game's combat situation--most games do not give characters different levels of economic power. Having more resources does count (and in fact, in SWG this was a problem, when the economic powerhouses would buy up a weaponsmith's entire stock in a few minutes just to get a good "slice" on their weapon), but this kind of leverage is typically not used for the purpose of malice. What type 3 players *are* likely to do is make use of whatever "duping" exploits are available in the system, actively attempt to swindle other players, and use whatever other bugs are available to wreak havoc on the economy.

As a final thought: It's interesting to note that the "bar" for being counted as a griefer seems to be higher in the economic arena than in the combat arena. By that, I mean that players who are more ruthless at trading are unlikely to be scorned as much as those who are ruthless in battle. There are a number of reasons this could be the case.

First, the playing field for economics is generally much more level than it is for combat. Economic games are more about the abilities of the player than they are about the abilities of the character. As a result, a larger imbalance of ability is considered to be fair. (Although it might not be that much larger--the order of magnitude of differences between players' ability to manage economic transactions is tiny compared to the difference in combat power of a low-level and a high-level fighter.)

Second, very few people see the economic side of these games as being a real realm for competition. Acquiring wealth is certainly a goal, but few see it as being the primary goal, and fewer yet have the goal of depriving others of wealth.

Third, the economies of virtual worlds are usually designed (or mis-designed?) in such a way that currency is continuously entering the system. As a result, very few characters are actually needy to such a degree that they feel "grief" at their losses.

Fourth, this play-style is perhaps just not "hands-on" enough for the type of person who's going out of their way to cause people grief.

And finally, succeeding in the trade game generally requires an extremely large reputation component. If you build up a reputation of being a bad person to trade with, people will go to your competitors. As a result, most "scammers" (scams being, essentially the economic version of griefing) have to work by using secondary characters, or rotating characters when reputation is lost.

Some of the differences between these two varieties of grief-behavior might illuminate strategies to make combat (or further dimensions of) griefing less attractive.

Hope my ramble wasn't too long. :)

[Richard Bartle]

9.

My player types model isn't driven so much by "what do players do" as "why do they do it". An achiever can show all the symptoms of being any of the other types, but be behaving in that "non-achiever" way for reasons that are purely to do with being an achiever. If they want the Sword of Lewt so much that they're prepared to explore for it, OK, that's what they'll do. They can exhibit traits that change from situation to situation, and they can transition between types. Nevertheless, they will usually be doing whatever they're doing for some reason that keys into their current player type.

Ren>it seems to boil down to whether one attempts to design grief out of a game (change people's behaviour) or to discourage griefers (change the people).

How could you change the behaviour without changing the people?

Richard

[Jesper Juul]

10.

"There are no grief players, only players that grief – discuss."

"Would such a move enrich all types of analysis of virtual worlds or is it in fact the very same view that Bartle is espousing and are we just counting angels on the end of a pin here."

What you're suggesting is the kind of move that I believe tends to limit analysis because it strongly encourages people to consistently make only one specific point - in this case that there are no fixed personality traits, only behavior and so on.

I rather think that the term "grief players" is already sufficiently open to cover RL players that consistently grief play as well as the act of griefplaying itself.

[Jonas Heide Smith]

11.

Richard asks:
"How could you change the behaviour without changing the people?"

I suppose that depends on what you mean by people:
A) The Inner Core of Personality
B) The displayed behavour

If you evaluate/categorize people by their actions (B) then of course you can't change the behaviour without changing the people.

If - on the other hand - you were to consider people as essentially point-maximizers (i.e. good old Economic Man) then changing the game structure would change their behavior while at heart they remain the same rational point-maximizers.

I'm still curious - why would the game designer care whether he/she is changing the behaviour or the people?

[Ren]

12.

Jonas Heide Smith > I'm still curious - why would the game designer care whether he/she is changing the behaviour or the people?

Two reasons (that I can think of right now):
1) These might be different design decisions so one would need to know which goal one was aiming for.
2) If we assumed a fixed market size and we assume an MMO wants to maximise its revenues then people become a precious commodity, hence you want to keep the people an change what they do, thus increasing you subscriptions.

[Bart Simon]

13.

Just a general comment on ideal types. There are three schools of thought on these.

1. You treat the ideal types as archetypes - that is analytical categories given by the structure of the system (whether you go nature or nurture on this it doesn't matter).

2. You treat ideal types as an analytical heuristic... which is to say you don't presume that actual people fit these types but rather that the types provide a useful way of analytically organizing people into groups.

3. You treat ideal types as inter-subjective categories or actor's categories. That is the types are generated from actors (MMO players)in their own attempts to make sense of their experience.

IMO - problems arise when these different ways of using ideal types (whether Bartle's or others) get confused. Its especially fun when the levels of analysis mesh - as when a player says "i've read Bartle and you're a killer buddy"

Now it used to be that ideal typifications were all the rage in the social sciences (ala version #1). The rise of constructivism has made us back off to version #2... the difference is critical since in version #1 the type pre-exists player behavior (i.e. all behavior can be classed in accordance with n types) whereas in version #2 the type is a post-hoc tool for organizing observed behavior - strong constructivists would not attempt to predict behavior on the basis of their typologies.

'nuff said - version #3 is the most interesting since these types are deployed as part of the behavior of the players themselves... ethnography anyone?

hey - my first TN post... right on.

cheers,
Bart Simon

[Richard Bartle]

14.

Jonas Heide Smith>changing the game structure would change their behavior while at heart they remain the same rational point-maximizers.

In the same way that you can look at people as either sets of behavioural traits or as some central core of being that manifests itself through behavioural traits, you can regard "game structure" as either sets of reactions to actions or a central core of design that manifests itself through reactions to actions. [Hmm, this may be that "soul" I was talking about at the conference].

If you want to change someone's behaviour, one option is to force it ("yes, your honour, I did lock my daughter in the cellar - she kept dating boys!"). This doesn't change behaviour so much as make the behaviour impossible, however.

The ONLY other way to change someone's behaviour is to cause them to WANT to change their behaviour. To do this, you have to change THEM. Now you can do this by bribery, bullying, reasoned argument, appeals to emotion - there are many ways. However, all of them involve making a change to the person whose behaviour you want to change.

>why would the game designer care whether he/she is changing the behaviour or the people?

Because if people don't change they're not having fun.

Richard

[Jonas Heide Smith]

15.

Thanks for those comments, Richard.
BTW: It just occured to me that my questions above were sparked by my misinterpretation of Ren's original post.

Ren's by-now-famous words were:
"in design terms it seems to boil down to whether one attempts to design grief out of a game (change people's behaviour) or to discourage griefers (change the people). "

I took this to mean "change the thinking/values" of existing users. But what Ren meant (he confirms) was: Exchange current users for different users. So the answer to my "Who cares what you change?" is: You don't want to loose current players if you care about profit.

I agree, Richard, that how you change people's behaviour can range from the primitive (locking them in basements) to the more elegant (changing payoff structures etc.). But I don't know that I agree that The Basement Solution is necessarily so bad (might be because I have a daughter myself).

You say that "Because if people don't change they're not having fun."

But isn't The Basement Solution used to great effect in MMORPGs without PvsP? The designers may have percieved PKing to be a problem in other MMORPGs so they made it impossible in theirs. And lots of people like it that way...

- Jonas

[Ren]

16.

Jonas Heide Smith > But isn't The Basement Solution used to great effect in MMORPGs without PvsP?

This time it might be me miss-reading. To me, removing PvP only takes out one type of grief play.

But - what are the others?

Well for a very good definition of that you need to wait till next week when: Foo, C.Y. & Koivisto E. M. I, (2004 forthcoming), Defining Grief Play in MMORPGs: Player and Developer Perceptions – is published at ACE 2004. I’ll post a link just as soon as its up on the web,

[Chek Yang Foo]

17.

Thanks for the reference to this blog, Ren.:)

I wanted to add a comment to the notes here. Just a quick brief first; I'm looking at grief play for my Ph.D research. One of the respondents I'm speaking to in my (very) qualitative study is a correctional psychologist who counsels criminals - and he also just so happens to be a long time Ultima Online player, and a one-time guild leader whose guild was caught up with the rampant griefing that took place in pre-2000 UO.

His take of it, quote and unquote, is that there is little difference between griefers and criminals. They think the same way, they justify their behavior the same way, and they probably get the same thing out of it that antisocials get out of taking advantage of others - a feeling of power.

CY (the "grief play" guy)

[T.L.]

18.

Great to read all the interesting discussion here. Especially terrific to see a couple new names pop up - hey and welcome Bart and Chek! :) Well, as Ren tipped off I floated this idea the other night over dinner and I seem to have been tossing it around whenever I can for the last few weeks ;) I've been curious to get different takes on it. Basically I've been trying to think through what it means theoretically and methodologically to take as the frame action (grief play or cheating) vs identity (griefer and cheater). This isn't of course to say that there probably aren't some people for whom these definitive labels fit. I've been struck however within my own research by players who are probably typically thought of as "good" and "honest" nonetheless at times engaging in activities that shade into grief/cheating. To complicate things of course those very categories seem pretty up in the air at times to players themselves, or at the least locally negotiated and worked over constantly. I've done the identity trope in past work myself (ie powergamers) but I increasingly wonder if framing things in terms of action does a better job of providing space for the ways grief/cheating have deep social and contextual groundings. I really look forward to seeing Chek and Elina's work on this :)

In terms of the import for design - I think it's a great question Jonas and one I'm still trying to think through a bit more. I am very interested in hearing more from our practicing designers about this issue (or if it's simply not a very useful discussion for on-the-ground development).

[Jim Landes]

19.

Richard Bartle> ”The ONLY other way to change someone's behaviour is to cause them to WANT to change their behaviour. To do this, you have to change THEM. Now you can do this by bribery, bullying, reasoned argument, appeals to emotion - there are many ways. However, all of them involve making a change to the person whose behaviour you want to change.”

While this is true, it may not be applicable in this context. Most game designs limit the player range of responses from the stimulus within the game environment. Complex and satisfying interactions seem to be between the players themselves against the game environment back-drop. Thus, when applied to the intrinsic needs of power (control), achievement (recognition), and socialization (acceptance) of the virtual world, we have a lack of satisfying player responses. When the player needs are not being met, they appear to improvise and substitute goals and activities that satisfy their needs (a unique combination of control, recognition and acceptance). These needs sometimes take the form of what we are terming “Griefing” in our current line of conversation.

Creating dynamic game environments that broaden the reward structure and allow complex interactions and/or varying types of competition is the key to addressing this underlying problem. These do not necessarily change the player, but will deliver to many the intrinsic needs they are looking for in game play, which will in turn alter their behavior.

[Scott Moore]

20.

I think I am missing something in the either/or aspect of changing the behavior of a person or causing the person to "want" to change their own behavior. Are there no shades of gray between these two points? Where is the role of influence? Is that a mild form of force? And if physical coercion (the basement) changes one's behavior, but not themselves, how is bullying (an act of physical coercion) different in that is can cause a person to want to change their behavior?

Here's a concrete example of one attempt I took in changing behavior. In WorldsAway, avatars could change names at will (it was an experiment), but we wanted to limit it somewhat by increasing the price of the new name with each change. One of the reasons for the increasing charge was to limit the ease with which troublemakers could hide their identity from the community.

This could be interpreted as "bullying" in which case, according to Richard, I would be trying to change the person. I'm confused because we never had the illusion that we were changing the person. We expected that if the increasing name charge were dropped, we would have people using it to grief and then change their names.

Nick's assessment makes the most sense to me from a practical standpoint. There is a point of diminishing returns on griefing. The problem for designers is that each person brings their own threshold of that point. The additional problem that I think is being overlooked is that the group, as a whole, will likely have a different (I guess lower) threshold.

Chek (and Ren), I am really looking forward to that paper. Having worked in non-game virtual environments (graphical and text-based), I can confirm that there are ways to grief without PvP. In fact, I cynically maintain that any bit of code that allows one person to act on another without the prior's consent can be used for griefing. Heck, I could take it further. Allow someone to influence the play of another and it can be griefed. I'd be very curious if there is anyone involved with ToonTown reading who might like to point out he creative ways people grief there.

[Nathan Combs]

21.

Are there no shades of gray between these two points? ...There is a point of diminishing returns on griefing. The problem for designers is that each person brings their own threshold of that point.

I think the bulk of folk in the middle are most "norm-abiding" - I've known folk who one might consider as "occasional" griefers - heck, most PvP players "go nuts" once in a while.

It is at the extremes (the pathological nooks) that lay the problems. So, yes, sure, they too can be described as "having a higher threshold" than the rest. But unlike those huddled in and around the norm, I suspect their "threshold" harder to reach via modest efforts.

btw, I also wonder how group dynamics play in all this. I suspect a lot. Put it another way, sure someone may be just barely "supra threshold", but because of group reinforcement and feedback, it could seem a mile depending upon context...

[Mike Rozak]

22.

From Jonas Heide Smith
> I'm still curious - why would the game
> designer care whether he/she is changing the
> behaviour or the people?

Literature.

From my POV, contemporary virtual worlds (and computer games in general) are empty calories. They emphasize primitive emotions (kill, fear). They contain no ideas or concepts. Except for a few instances, I don't walk away from games feeling changed/enlightened, like I do when I watch a classic movie like "2001" or read a classic book like "Watership Down".

If I do ever design a game, the reason I will design the game is to change people. Not through coercive means (locking them in a basement), but by giving them some new ideas/concets to ponder (seeing the world from a different POV, etc.).

[magicback]

23.

Nathan Combs> btw, I also wonder how group dynamics play in all this. I suspect a lot. Put it another way, sure someone may be just barely "supra threshold", but because of group reinforcement and feedback, it could seem a mile depending upon context...

I'm in the "players that grief" and "players don't grief alone" camp.

I think group dynamics plays a larger role than personal dynamics and the design lever should be directed at the group level: changing groups of people rather than changing people (or using group dynamics to affect behavor).

If you review the rise of civil affairs among different cultures, you can see how city-states form to protect the "norm" within city walls and outcast the "outliers" into the wilderness. Connecting the city-states are armed caravans that established a mobile zone of civility. Once in a while the “bandit king” becomes the “king and protector” of city-states.

Take a look at the desert Bedouin society. Bedouin groups are the uber-guilds of the desert who grief each other via raids and camp key desert resources like wells, grazing pastures, etc. They look ruthless on the outside, but when you enter the camp as a guest, you find that they are men and women of civility and you are treated like a family member.

Furthermore, if you take an expanded look at primate societies, you’ll understand how primal griefing is a force of nature that keeps the pack evolving (too much Discovery Channel shows for me.)

Frank

[Barry Kearns]

24.

Jonas wrote:

> I took this to mean "change the
> thinking/values" of existing users. But what
> Ren meant (he confirms) was: Exchange current
> users for different users. So the answer to
> my "Who cares what you change?" is: You don't
> want to loose current players if you care about
> profit.

But the whole point of a rational set of game policies with respect to griefing is that (hopefully) the developers/GMs are attempting to maximize the total aggregate number of player-months spent in the game by subscribers (and thus maximize profits).

It's perfectly rational to want to lose CERTAIN current players if, by doing so, you either ensure that you're not going to lose a larger fraction of OTHER players due to them quitting out of frustration... or better yet, you show that the "police force" of the game will act to protect the interests of those who are often unable to "protect themselves" due to game mechanics... and thereby encourage the sense of a safer, more enjoyable community.

One of Frank's observations is dead-on with respect to this point: 'If you review the rise of civil affairs among different cultures, you can see how city-states form to protect the "norm" within city walls and outcast the "outliers" into the wilderness.'

In the vast majority of cases where the game mechanics allow griefing of some form (no "basement" provisions), players typically lack the mechanics to enforce proper 'civil' behaviors on others (they can't effectively cast the outliers into the wilderness THEMSELVES), and thus cannot protect their own ability to enjoy their gaming experience.

The developers/GMs act as enforcers of a societal norm... and the extent that they FAIL to enforce proper punishments for those who set out to ruin the experience of others, they create a sense of lawlessness.

What option is then left to players unable to insulate themselves from such anti-social behavior? Why, they vote with their dollars, and leave.

Establishing (and much more importantly, fairly and rigorously enforcing) community standards serves to enhance the long-term profitability of the enterprise as a whole... you "sacrifice" those customers who would cost you FAR MORE if you allow them to drive away OTHER customers.

I see it as little different than a restaurant throwing out (and refusing future service to) an unruly and disturbing patron fond of throwing temper tantrums.

Do restaurants want to lose customers? No... and often the best way to minimize their overall losses is to quite visibly and permanently lose THAT particular kind of customer.

[Chek Yang Foo]

25.

The discussion here is really meaningful. One of the objectives of my 4 year study is to investigate how grief is being managed right now, and if it can be better managed at all - e.g. through player justice.

After reading through the posts here, I decided to go back to some of the data I've accumulated in the study. The below are selected notes from a first series of semi-structured interviews I did with a group of several dozen respondents (comprising players, griefers and developers) over Sep 2003 to Jan 2004, with the theme on "why players grief". Data from discussion room postings, and my field observation are not included below as they're (at the moment) in a separate part of my study. Here's a (disorganised!) sample of what the data says:

- griefers grief because they have no power in real life (e.g. kids vs. parents). In the game, they have power, so they impose that power on others.

- In the real world, they lack some physical characteristic – e.g. looks, money, girl friend – so they grief in the virtual world as a harmless way of replacing what they lack.

- "Idle hands, idle minds" - they grief because they have nothing else better to do, as another player puts it. This sentiment is disputed by a developer respondent though, who said players will grief only if they're bored AND it's easy to grief. If they're bored but it's hard to grief, they'd typically just move to some other game.

- Reputation, as one griefer put it, just so that they can brag about it to their peers, e.g. how many times they got banned or suspended.

- Another explained this incident in which some reluctant fellows joined in by way of getting accepted into a group and 'brotherhood'.

- Frontier justice: one reason cited by a griefer was that he wanted to put down "the high and mighty" players, and specifically the role-playing ones who say that "they are more mature players, and the pvpers are immature little brats."

- Another griefer said he griefs Trammel players in UO because he hated the cowardness of many players, and the feeling of favouritism given to them by OSI that led to their smugness.

- Making a statement to game management: a bug or class misbalance that a player feels has been ignored by developers for a long time. So he decided to take it out on the player base to provoke a response.

- Anger - someone griefed because his (real world) dog took a crap on the rug, and he got mad.

CY

[Barry Kearns]

26.

One of the mechanics that I have been rolling around for some time deals with a game-internal "player rating system".

I've had a game in development for a long (long, long) time now, and one of the concepts that I wanted to specifically incorporate was a method by which GMs could get a determination for who the "problem children" were in a given population.

I incorporate concepts along the lines of a Google "page ranking", where players are able to rate others, and their ratings carry proportionally more or less weight depending on how well they are (themselves) rated... the theory being that it's much harder to use this system ITSELF as a form of griefing.

One of the significant aspects is that no player would be able to see their own "rating", nor the aggregate ratings of others... this would simply be a mechanic that would allow the developers/GMs to track the in-game "opinion" of other players. Players, of course, could use it as a handy form of "sticky notepad" for interacting with people whose name you might remember, but you can't remember quite WHY at the moment... =)

I was visualizing a -10 to +10 scale, or something like that. Something quick and easy to set on-the-fly either from the UI or the "command line".

The comedian Gallagher inspired the idea... he has a bit about giving everyone suction-cup dart guns for mounting on their cars, and darts with a "Stupid" label on them. When someone does something moronic, you shoot their car, and when the cops see someone with a huge collection of darts, they pull them over and give them a ticket for being an idiot.

Many games have an in-game mechanic for reporting other players, but they are generally cumbersome, and lack a context for establishing the credibility of reporters and reportees in a "my word versus yours" scenario, which tends to lead to a sense of uselessness in the system. Of course, such a system gives the rule-enforcers statistical insight not only into the reportee in an incident, but also the reporter.

Such a system could be used for anything from a simple "credibility measure" for GMs, all the way up to a full "player justice" mechanism (though the latter holds the possibility of significant grief-like behavior unless proper safeguards are built-in.)

I could see this leading to some interesting dynamic tension between the player base's image of "good behavior" versus that of the game designers / rule enforcers.

Should the king define the bounds of proper behavior, or should the people?

[Richard Bartle]

27.

Jonas Heide Smith>But what Ren meant (he confirms) was: Exchange current users for different users.

It did occur to me that this might be what Ren meant, but that doesn't mean we can't continue a conversation about what you thought he meant.

>I don't know that I agree that The Basement Solution is necessarily so bad (might be because I have a daughter myself).

I have two daughters, but no basement (sigh).

>But isn't The Basement Solution used to great effect in MMORPGs without PvsP?

It's used to effect; whether it's used to great effect depends on your definition of "great". It might be short-term great but long-term ungreat.

>The designers may have percieved PKing to be a problem in other MMORPGs so they made it impossible in theirs. And lots of people like it that way...

But not having it brings in other, longer-term effects that people don't like yet which they don't associate with not having PvP.

Physical prevention of actions occurs all the time in virtual worlds. If there's no system for teleportation, you don't get to teleport; if there's no system for setting fire to rocks, you don't get to set fire to rocks. These are hard-wired rules of play about which the players can do little (except complain). Changing the rules can change what players do (by adding or removing choices), but does it change why they do it? I guess it's possible over time, but not something you can rely on. More likely, what would happen is what Ren was asking originally: if you change the rules, do you replace the people whom the change affected negatively with new people whom the change affected positively? If the change were radical enough, well yes, you would. You'd need to look at the long-term effects to determine whether or not overall the change was a good idea, though.

Richard

[Richard Bartle]

28.

Jim Landes>Most game designs limit the player range of responses from the stimulus within the game environment.

Yes, they'd have to. Otherwise, everything would be just so much white noise.

>When the player needs are not being met, they appear to improvise and substitute goals and activities that satisfy their needs (a unique combination of control, recognition and acceptance). These needs sometimes take the form of what we are terming “Griefing” in our current line of conversation.

I agree: the less there is for people to do, the more they have to make their own entertainment. This involves determining what building-block actions are available, with a view to putting them together in interesting ways. Unfortunately, those "interesting ways" could lead to a meta-game rather than something in the spirit of the original design.

>Creating dynamic game environments that broaden the reward structure and allow complex interactions and/or varying types of competition is the key to addressing this underlying problem.

I think there's a limit to the broadness, but yes, on the whole I'm with you on this.

>These do not necessarily change the player, but will deliver to many the intrinsic needs they are looking for in game play, which will in turn alter their behavior.

This is where we diverge. I see those "intrinsic" player needs as only temporary steps of a much longer journey. When players are having fun, they are advancing (slowly, but nevertheless surely) along a path of personal development. As they develop, their needs change. As their needs change, their behaviour changes: this is because they themselves have changed.

It could be we're in agreement, just looking at different timescales.

Richard

[Richard Bartle]

29.

Barry Kearns>It's perfectly rational to want to lose CERTAIN current players if, by doing so, you either ensure that you're not going to lose a larger fraction of OTHER players due to them quitting out of frustration...

This is correct, but there are underlying complexities to it. Firstly, it's rarely one change that causes people to leave, just the accumulation of changes - "the last straw". Even then, people usually have other reasons for wanting to leave, which they often have difficulty articulating; a change can act as a trigger, but they were probably unhappy before then. Ironically, the reason they were unhappy could well be related to the knock-on effects of earlier design decisions of which they approve. It could be that without a certain level of griefing, the virtual world loses features that they find attractive.

By way of analogy, suppose a prosperous small town had a big factory that produced a lot of pollution. No-one likes pollution, people complain, they don't want to leave but they will if the factory doesn't clean up its act. OK, so the factory cuts back on production which reduces pollution, but also puts a lot of people out of work. Now there are groups of unemployed people wandering the streets. The people who complained about the pollution now perceive the town to be dead-end so they move to somewhere more upmarket.

>In the vast majority of cases where the game mechanics allow griefing of some form (no "basement" provisions), players typically lack the mechanics to enforce proper 'civil' behaviors on others (they can't effectively cast the outliers into the wilderness THEMSELVES), and thus cannot protect their own ability to enjoy their gaming experience.

Game mechanics always allow griefing to some degree: even free-form communication allows for griefing. Although I agree that allowing players to enforce local norms on one another is a reasonable way to control griefing, I don't think it's the only way - or even necessarily the best way, at least at an individual level. I'd much more prefer it if the griefers were themselves to realise that if they stopped griefing they would have a better experience.

Some people just get stuck in a grief rut, though. Here's an exchange I heard on the radio a month or so ago:
Interviewer: So the reason you steal is because you need the money?
Criminal: Yeah, that's right. I have a wife and kids to support, I need the money for their sakes.
Interviewer: Is there nothing you wouldn't do for money? You'd beat people up, steal their wallets? Kill them?
Criminal (proudly): There's NOTHING I wouldn't do for money!
Interviewer: How about working?
Criminal: Ah. Except working.

Richard

[Scott Moore]

30.

Nathan, thanks and sorry to make you repeat what you had already said.

Richard, thank you also, I understand your intent to change griefers. Picking up your small town factory analogy, it seems as though we are looking for ways to reduce the population without necesarily reducing production. Say, a starting up a business (or civic project) devoted to filtering or disposing of the pollution of the factory.

What forms of player enforcement have we seen work? Given that any game mechanic itself can be griefed, it seems there would be some form of checks and balances to the group defining the norm.

[Bart Simon]

31.

my - these threads do get large don't they.

On Jonas'issue of designing out griefing/griefers. Certainly 'A Clockwork Orange' comes to mind (are we not talking about social engineering and behavior modification in effect)but might I also suggest (and maybe this has been said... its getting difficult for me to follow)that there is a fundamental design error in this desire (i.e. to make it so the only rational choice is to play nice).

Arguably it is the tension between in-groups and out-groups, deviance and normalcy, right and wrong, good and evil that gives MMO games their fascinating quality... a player derives some satisfaction from chastizing or complaining about the griefer as much as the griefer enjoys annoying others.

The result (as seen long ago in MUDs and long before in all manner of utopian social experiments) are exercises in constituting innovative forms of social order.

The desire to eliminate the tension over the boundary deviant/normal through social software engineering (via rational choice theory or otherwise) might actually be counter-productive if stable social orders (happy, cooperating, communing people) are the goal (er.. plus I don't think it'll ever work anyway).

From a design standpoint perhaps the issue is how to make player solutions to the player defined problems of griefing evident to the player communities themselves. That is, how to give players more tools for innovating their own solutions. I might think of things like voting systems, scandal sheets, most wanted posters, public 'executions', rehab and community service, surveillance systems, etc... but my idea would not be to place these in a game as modules but rather create mechanisms for players to construct their own systems/modules (which they already do in a sense) so that instead of the virtual equivalent of our criminal justice system some innovative hybridization might develop.

cheers,
bart

[magicback]

32.

Bart> From a design standpoint perhaps the issue is how to make player solutions to the player defined problems of griefing evident to the player communities themselves. That is, how to give players more tools for innovating their own solutions.

This design standpoint also applies to the discussion on guild support functions.

I agree that there must be sufficient room for both normalcy and deviancy, civil and uncivil (constructive tension). I agree also that it is good to make possible solutions more evident with flexible tools to allow innovate.

A Tale in the Desert has its petition system and Eve Online has its different security zones. Short of very structured games like City of Heroes or Toontown, dynamic VWs should accept a certain level of grief-play as a part of the dynamics and empower the players to manage satisfactory this dynamic factor.

Frank

[Jeff Lotton]

33.

It's always intersting when, within a MMOG community, the players try and define the difference between cheating and griefing.

The difference is often about bannable offenses and usually start when someone griefs. Most TOS say things like : " cheating will be dealt with...blah blah blah" but griefing hasn't really been so clearly defined as an offense.

JHL

[Alan Stern]

34.

Bart Simon> ...but my idea would not be to place these in a game as modules but rather create mechanisms for players to construct their own systems/modules (which they already do in a sense) so that instead of the virtual equivalent of our criminal justice system some innovative hybridization might develop.

I really like the idea of trying to find and/or let players develop new, innovative solutions or hybridizations to address the issue. My question to you is how would one enable such a process/freedom? You don't want the 'answers' provided as in-game modules so how do you give the players enough freedom to accomplish self-regulation while still maintaining not only the underlying bases and rules of the game but also while not opening up these very solutions/freedoms to griefers to be exploited against the other players?

[Hellinar]

35.

Ren wrote> The ONLY other way to change someone's behaviour is to cause them to WANT to change their behaviour. To do this, you have to change THEM.

That seems to assume that peoples wants are binary, they want this, but not its opposite. Seems to me more typical that people want something, and its opposite simultaneously. Context influences which want wins out. So changing the context will change their behaviour without making a fundamental change in their makeup. So behaviour change isn’t generally as tough as the forcing new desires onto people.

For example, someone might want to the glory of being the leader. At the same time, they want the lack of hassle of being a follower. The mechanics and demands of leadership in a world will push them into one behaviour or another. Though they will still "want" both states. Some outliers of the population will be very hard to push into one role or the other, but I think the norm is flexible.

The Basement Solution is being actively tried in A Tale in the Desert. In that world, Egypt, laws when passed are coded into the server, so the forbidden action becomes impossible. In general, non codeable laws are vetoed by the developers. There are now a number of grief actions that are not longer doable in the world, forbidden by the Players through the law system. The devs have been surprised how few laws have been passed. Generally, the Players are wary of such a draconian and inflexible system.

As I see it, the ultimate sanction against griefers is let the Players choose with whom they share the world. And more strongly, with whom they share their world history. That would require a more radical design of a VW than has been attempted to date. I don't know how many people would want to play in such a world, but it would be more resistant to greifing than the current models.

[ren]

36.

Hellinar wrote:
>Ren wrote> The ONLY other way to change someone's

Actually that was Richard, but do feel free to attribute more of his comments to me :)

[Jonas Heide Smith]

37.

Bart said:
"On Jonas'issue of designing out griefing/griefers. Certainly 'A Clockwork Orange' comes to mind (are we not talking about social engineering and behavior modification in effect)"

I don't see why A Clockwork Orange springs to mind as opposed to, say, Plato, Rawls or indeed any political philosopher. Clearly its behaviour modification, all institutions are.

Bart:
"Arguably it is the tension between in-groups and out-groups, deviance and normalcy, right and wrong, good and evil that gives MMO games their fascinating quality... a player derives some satisfaction from chastizing or complaining about the griefer as much as the griefer enjoys annoying others."

Well, to some degree. But would you then deny that there have been clear cases of behavior being so destructive that the community has been shattered or seriously damaged? (one example would be far too rampant PKing in early OU).

Bart again:
"The desire to eliminate the tension over the boundary deviant/normal through social software engineering (via rational choice theory or otherwise) might actually be counter-productive if stable social orders (happy, cooperating, communing people) are the goal (er.. plus I don't think it'll ever work anyway)."

Again, I mostly agree. But we don't have to be talking about friction-free utopia, we might just want to reward certain behaviour types and punish others. All succesful communitities do this - in social software, however, the techniques that these communities use in RL are sometimes not available.

I'm thrilled by the thought of bottom-up innovative player solutions. I'm just not all that convinced that we shouldn't just adopt the best systems that have sprung up through the history of physical world communities (not convinced as in: not YET convinced).

- Jonas

[Richard Bartle]

38.

Hellinar>Seems to me more typical that people want something, and its opposite simultaneously.

They can want all sorts of things, including the impossible. I don't think that's incompatible with what I said, though.

If you want them to do something then basically you either have to force them to do (the Basement Solution) it or to cause them to want to do it. If you cause them to want to do it, then you're changing them.

Changing the context so they'll do one thing rather than another is, I submit, a soft version of the Basement Solution. A goat wants to go into one of two fields to eat grass, it doesn't care which field, but when you put a wolf in one field then it will choose the other. You haven't changed the goat, but you've changed the world to make one course of action (as far as the goat is concerned) undesirable.

Interestingly, if the goat went away and came back with 50 other goats and together they beat up the wolf, then you'd have changed the goat. The same action that can direct mere selection from an underconstrained system can also lead to change in the self.

Richard

[Bart Simon]

39.

From Jonas:"I don't see why A Clockwork Orange springs to mind as opposed to, say, Plato, Rawls or indeed any political philosopher. Clearly its behaviour modification, all institutions are"

Ah but the potentials of "virtual worlds" are so much greater - Plato and Rawls would be in awe. How would you like to tweak the code so that after being convicted of a griefing offense (via the automated griefing detection system) the player's avatar breaks into an incapacitating spasm attack upon the next grief defined behavior. This form of behavior modification works on the "coded" body as a form of discipline - better ala Burgess actually than Rawls.

The ontology of Burgess and Rawls are different (and so are yours and mine Jonas... which is fine by me; makes for nice raunchy discussion - since (for me and Burgess) 'man' is no rational animal it will be action on the body rather than the mind that do the trick (or so we are left to ponder)

But if its political philosophers we need - I'd grab Bentham because he does well to articulate the link between body and mind (or soul) through something like the panopticon. If I may skip a beat ('cause is on my mind)

The 'virtual' body (or one's instanciation in the virtual world) is the only link to the source of intentionality/motivation (as a quick aside this is usually located in the player... but this is like sooo problematic in role-play situations - "hey i'm a chaotic evil anti-paladin... i'm supposed to be anti-social and break all the rules"... also interesting is the possibility of automating motivations in game worlds - like auction sub-routines that get the avatar to sell to the highest bidder... why not sell to the lowest, friendliest, prettiest bidder?).

Anyway, back to Bentham - is not the virtual world almost perfectly surveillable... an uber-panopticon? There's your solution to griefing (or any behavior anyone with coding power in the game may not like... an old lession from Julian Dibble eh?)... if the players know for sure they are being watched and they know the consequences of an action deemed deviant then it will be in their best interest to conform.

The panopticon was conceptualized as a prison work-house but a fantasy virtual world would suit Bentham just as well -- might virtual worlds be used as a form of rehab for "criminals"?

Perhaps these thoughts are a little extreme but the idea is simply to raise the design issue of where to draw the line - in virtual worlds you can structure the environment, the bodies, the conditions of action and arguably even the minds.. this is dizzying power Plato could only dream of (and of course some say that is the point - its only supposed to be a dream).

cheers,
Bart

[Ren]

40.

Bart – thanks for brining Bentham to the party, v interesting point…

I’m not sure if we can apply the panopticon model sufficiently far for it to have the effects that Bentham envisaged. First off, of course, we have to assume that a panopticon does have desired effect – anyone one have references of empirical studies?

I take the ‘effect’ to be the idea that people will conform to a norm through the internalisation of the normative force of the gaze. The ‘force’, I take it, comes from a set of power asymmetries established by the structure operation of the panopticon.

Certainly we can construct a VW where there is tracking of behaviour and text input, and this could be mixed with some kind of sanctioning e.g. giving players positive / negative ratings though to kicking them out.

As with a lot of issues of establishing norms in VWs one has the difficulty that people are there by choice, hence there is a tension between the value of the world to the individual and the type of sanction that one can use, this really limits the power of sanctions. And is a major difference with the panopticon where people are there by force and the sanctions that may be applied are non trivial.

There is also a kind of structural difference. In the panopticon the potential of the gaze is constantly perceived (I say potential as the guard is both seen – when an inspection takes place, and semi hidden i.e. the shadow of the guard combined with the ability to come and go as they please is important). This constant potential is what forces the internalization of the gaze. To have this in a VW one would have to engineer a constant representation of the presence of the potential of normative power.

Also, I’m not sure about the way that Bentham was making the connection between mind and body, and how applicable this is here. While the panopticon is very much about seeing / being seen and the behavioural impact of this, there are also the physical facts of incineration etc that have to be taken into account or at the very least the involuntary nature of the observation.

Lastly there is the contrast between the panopticon as a system of knowledge (where the observed are watched and recorded – setting up yet another power asymmetry) and the ways in which our relationship with a VW spills out all of the place i.e. there is much about us as people that the VW does not know.

I’ve tried to pick apart some of the differences as future VWs may close several of these gaps the thus VWs may become more normative, things that would seem to have a direct effect are: less anonymity, better AI / Data mining systems so that behaviour can be tracked in near real time etc.

But all this might be a very bad thing. Griefing in VWs comes form the fact that VWs are less panopticon like than real life. At least some people do fee more free (or at least perceive a different set of norms). And to take Richard’s line – the journey that people go through (which I take to be enabled due to the freedom) is valuable, or to take Foucault (again as that’s where the power analysis of Bentham comes from) – knowledge or self is a moral good, and issue is that griefing seems to be a necessary evil, but is a too big a price to pay.

[Chris Mancil]

41.

"A small percentage (1-3% of the total playerbase) are pathological in their methods, trying to excercise control by motiveless player-killing, harassment, or whatever methods are feasible in the game system.

To put it another way, griefers are to "Killers" as 80 hour a week power-gamers are to Achievers: extreme examples of a common type."

Nicely said Dave.

[Richard Bartle]

42.

Ren>First off, of course, we have to assume that a panopticon does have desired effect – anyone one have references of empirical studies?

I only have anecdotes, but I can tell you that if you log absolutely everything in a textual world (ie. you record exactly what is transmitted to the players) and if you record all gameplay decisions the server makes, aberrant behaviour will drop practically overnight to only a small fraction of what it was before. In some cases, it's eliminated.

There are two problems with this approach, though. Firstly, you need someone to look at the logs when there's a dispute; this isn't hard for textual worlds because they have manageable numbers of players.

Secondly, you need to be able to record absolutely everything (and do so at the server side). This is easy with textual worlds, but involves saving huge amounts of data for graphical ones. Furthermore, in a textual world you can search through the text for specific events fairly easily; in a graphical world, it's much, much harder to find out whether something might have happened earlier.

If there were better tools for finding where an incident was (cross-indexing with easily-searchable the gameplay logs, for example) and if the logs were kept long enough (at least a week, and possibly indefinitely for people flagged as ones to watch - that includes all the CS staff) then I'm sure we'd see a reduction in griefing as a result. However, the hardware, software and CS support for it wouldn't be trivial.

Richard

[Jessica Mulligan]

43.

They may be small as group, but that diminuative, pathological group of griefers suck man-hours out of your dev team, customer service, community relations, et al, far out of proportion to their numbers. As Richard noted above, you can ease the investigation, but it still requires man-hours to resolve and the cost is not trivial.

At OSI, for example, we had a harrassment report button for the UO players; click it and the last few minutes of your character log, including chat, went out as an email to us for later investigation. This did weed out the stupid ones, but caused something of a Darwinian effect by making the less-than-stupid ones more clever in their efforts and thus more difficult to pin down. In effect, we traded the least time-intensive griefers for the most time-intensive ones, who also turned out to be those most willing to accept enforcement efforts as a challenge and increase the frequency of the behavior.

About the only way I see to come close to curtailing this behavior (and you'll never get rid of it entirely) is to have the right CS tools at launch (which means extensive data-mining and logging capability and players tools to report behavior), reduce account and character anonymity to some level you can live with and still look at yourself in the mirror in the morning, have a large staff of people who can respond quickly to such events as they happen to create a feeling among players that you take it seriously (especially at launch, where it matters the most), and be completely, utterly merciless about banning griefer accounts.

And be prepared to spend more on customer relations and service than you ever dreamed would be necessary.

[Xango]

44.

earth Distributor

Xango

Xango Juice

Mangosteen

xengo juice

Distributor

Natural

http://xengo.net/xango/9636.cfm?

[vimax]

45.

How does one determine whether a law is just or unjust? A just law is a man-made code that squares with the moral law or the law of God.