design: limiting the real money trading cheat

Real money trading (RMT) is a result of users being able to organize themselves on the web and selling their in-world assets for real money. This detracts from the virtual world as a coherent self-sustaining entity and real-world inequalities spill over into the game. So what can you do about it?

There have been many discussions in places like Terranova, Raph's blog and Tobold's blog, but most designers seem to agree that RMT is an evil that cannot be eliminated without harming the game. I don't disgree with this assertion, although I believe it can be controlled.

So what are the most obvious options for a designer wanting to control RMT? The simple answer: To design for it or remove what enables it.

The Trading Card Game

Embrace and extend what is going on in existing games on a relatively limited level by making it the core focus of the game. Design the game from the ground up as a trading card game. By selling random undisclosed assets you get to sell the same useless equipment over and over, forcing players wanting the best stuff to keep spinning the lottery wheel, paying thousands of dollars. As there are no opportunities to gain assets cheaply the RMT market is curbed. The RMT market is further curbed by making cards either consumables or timelimited and having them stick to the character on first use.

The Fading Template Game

This is basically a class-based design in which you don't pay for an account, but purchase the right to play a specific class of characters. You can then provide opportunities for upgrades and multiclassing. Let characters' abilities rise and fade over time, following a predetermined curve, basically forcing the player to purchasing upgrades from the game company at regular intervals. This also curbs the RMT market as there is no way to gain assets cheaply and assets loose value over time. The net advantage is that you can provide the service at multiple pricepoints and that players can switch back and forth between cheap and expensive options depending on their activity level.

The Inflation Game

Forget everything you know about creating a balanced design. Try to get the inflation rate as high as possible. Deflate the value of all assets continuously. Investing in assets become a lot less attractive. The key example of this type of game is a reset based game.

The No Feedback Game

Make sure that there is no way for the player to objectively tell what the capabilities of a character is and no way to tell what the the capabilities or remaining duration of assets are. This curbs the RMT market as there is no way to know whether you get what you pay for. The net advantage is that it doesn't curb gifting.

The Neutral Game

All assets are average, exchangable and easy and fun to obtain. Nothing is special, players have to depend on social relations, knowledge and real skills.

The Social Game

Focus heavily on prestige related to group membership. Each guild is a hierarchy to climb with their own rules about how to climb the hierarchy and how to go about it, and it is supervised by guild representatives. You don’t eliminate RMT, as you can still have a RMT guild, but they would have less prestige. So it matters less.

The bullet list.

Limiting RMT means attacking the foundations which makes RMT attractive.

• Focus on how, not what. Assets are easier to trade than situastions.
• Focus on collective achievements. Collective achievements encourage group identity, group prestige and moral and makes hiding RMT-bought assets from your peers more difficult.
• Focus on lots of tiny assistance (sp?), discourage players from giving big one-shot favours. That makes selling a service tedious and you thus have to rely on real friends to support you.
• No sex before marriage: Require players to spend time with people they receive major favours from. This doesn't prevent RMT, but ensures that you are tied to the person, thus you have to think twice about who you receive major favours from. It better be someone that you want close ties with.
• Focus on diversity. Selling assets which the purchaser has no knowledge of is difficult. If all items are different and personalized then the market for each item will be difficult to find.
• Focus on customization and use-once assets. It doesn't prevent RMT, but limits trades to whole charcters.
• Focus on transparency. Let players prove that they have gained their trophees themselves, you cannot sell honour.
• Focus on independent groups. RMT-based groups should not be allowed to rule non-RMT groups.

Label the cheaters and attach fingerprints to creations

The most potent tool for reducing the negative effects of RMT is entirely psychological. Strive for a cohesive culture in which RMT is viewed as non-threatening. RMT is most damaging to the magic circle if it is viewed as the main route to success. If those who object to RMT can choose to view RMT as a failure or non-important then it will trouble them less.

Remember that do-it-yourselfers, roleplayers and artists don’t suffer from RMT. Encourage those activities which involves personal expression. For an artist the template is a failure, breaking away from the template is (partial) success. Everbody in a group can be “an artist” within that group’s culture (a great moderator, a great leader, a great joker etc). Associate prestige with that which have your own personal stamp on it. You cannot purchase your own fingerprints. Having other people appreciate that which is essentially you is an invaluable aspect of virtual worlds.

In essence, make sure that what the users value the most cannot be bought for a high price. Signs of honour have to be authentic.

design: basic constraints

When designing a virtual world you need to consider how the online reality differs from physical reality...

This is a list of some of the major constraints that applies to most virtual worlds, their games and economies. Particular designs have to deal with even more constraints, of course. This list is temporary.

Strong constraints

• New users arrive all the time: a key design challenge is to allow new players to join at any time while keeping the game fair and interesting for them.

• A single user might have multiple accounts: Designs which gives a new character exchangable resources is prone to abuse, as a user may create multiple accounts and transfer all the resources to one account.

• A user might never return: so you have to make sure that user's who never return don't create bottle-necks or dead-lock like system behaviour.

• Users have access to alternative communication channels

• Users might be able to use the network protocol directly

Weak constraints

• A user may have to log off instantanously

• Fraudulent behaviour: users will try to cheat on other players if the system allows for it, by faking system messages or other means.

• Infrastructure attacks: whatever is exposed of the network infrastructure opens it for attacks from vile users.

• A user may have technical issues: network latency, lack of audio etc.

• Some users cannot deal with complicated interfaces

• Some users suffer from disorders: epilepsy, motion sickness etc

• Minors may obtain access to the system

Desirable properties

• Encourage friendships and altruistic actions: create meetingplaces, and make it easy and cheap to do favours for other users.

• Allow users to spend time together: avoid forcing single-user gameplay, allow other users to tag along and help out.

• Allow identity formation: allow enough varieties of creative outlets for the user to establish an identity and form which he can identify with.

philosophy: cheating is impossible

Designers and users are easily annoyed by cheating in their game worlds. This mentality is destructive for the world...

There are two main types of exploits:

Cheating

In-world cheating takes place in the world and involves users breaking the norms that is imposed on them by themselves or others. The term is often used to refer to users who gain advantages by innovative exploitation of the environment.

Infrastructure abuse

Out-of-world infrastructure abuse involves using the interface, computer-system, network or other out-of-world means in unexpected ways to achieve effects which affects the world and the users of it in a negative fashion.

Let me start by a contradiction. Cheating doesn't happen within the world, as that is clearly not possible. If you think somebody else is cheating it just means that their local norms are different from yours. Cheating is all about breaking norms, but breaking norms should be allowed in a virtual world. That is, if you want it to be a world and not merely a game.

Games versus world

Games are all about global norms (rules) and climbing some power-ladder while staying within the limits of the rule set (either by lowering the powers of others or increasing one's own power). Worlds are all about not having to know the rules on which the game world physics relies.

In a game you are meant to stay within the confines of the intentions behind the rules. If you are not, the referee (game master) will make sure to interpret the rules for you and enforce it. In a world you shouldn't have to learn the rules, but do what you feel like and learn the consequences by success and failure. Whether you should be happy with the consequences of these successes and failures is really up to you and not up to global norms.

A game is different in the sense that it assumes that you want a dominating position. If other users optimize for loosing rather than winning, the game breaks. Not so with worlds. If it is a world, you simply live in a world full of loosers...

Design flaws

What you do have in virtual worlds are bugs and flawed designs. Whenever a user has to appeal to admin power the world has failed. The stronger presence admin power has in the world, the less it is a world, simply because admins are out-of-worldish and the ultimate failure you risk by challenging their rules is also out-of-worldish (i.e. a ban).

One could also argue that admin presence makes a shift from world to game, as the sole purpose of having admin moderation is to enforce global rules that are not embedded within the world. Such rules will be highly gameable as their enforcement is less strict, more subjective and easier to influence than hardcoded rules.

Immersion breaking exploits

When the world operator finds that users are using the world in ways which he doesn't like, he will think of them as cheaters and demand that they stop exploiting his system.

This is where the water gets really muddy. Its muddieness is best shown by the fact that some operators not only forbid users to use exploits, they also forbid users to communicate the exploits to others. You are basically assumed to know or accept that some consequences are intended while others are not. This is where the world dies and the game wins. It institutionalizes the idea that you should know the rules before you interact with the world and that you are not allowed to take the world for what it is.

This contradicts the main pre-requisite for deep immersion. In order to achieve deep immersion users will have to give up their norms and let their character's personality interact directly with the environment as experienced. Deep immersion doesn't allow for two parallel consciousnesses, the user must be allowed to yield to the character-in-the-world.

Role-acting

So what about role-acting where you rely on norms about separating what is done in-character (IC) and what is done out-of-character (OOC)?

Clearly this is a global norm generally enforced by the users themselves. The need for the norm arise because users always have an out-of-worldish component to them. In the case of role-acting other users constitute a vital part of the virtual world which you are exploring, and the isolation of the out-of-worldish aspect becomes important.

I will however still argue that the argument I have made about cheating also holds for the IC/OOC case. If a user starts to act OOC then role-actors will either ignore him, reply in IC or switch to OOC. The user is allowed to learn by success or failure. In the odd case that you have an operators that is enforcing IC/OOC separation by moderation then the very definition of OOC becomes gameable. If users are able to make the definition of what is IC a powerbase then the moderated role-acting system takes a swing from world towards game.

Conclusion

Cheating isn't possible in the world, only in the world's games. All attempts to get rid of cheating are detrimental to the key characteristic of virtual worlds.

The rules of the games in a virtual world work against the world when enforced. Such rules should instead be embedded in the world physics and deduced from the use context by the users. If they are enforced and defined out-of-the-world then you no longer have a contained world. If you forbid cheating you get a game, but you don't get a world.

Together we fall... Designers and users.

Related

This article is based on a mud-dev discussion from 2004. It was triggered by the constant disturbing murmur of game-players who keep ranting about players who cheat. And, I might add, a slight annoyance with Espen Aarseth's claim that cheating is a play-style on the same level as Bartle's four play-styles.

pattern: inflation based expansion

Virtual worlds usually have static and dynamic aspects. For pragmatic reasons most online games provide static environments where the gameplay is fairly predictable. Common wisdom says that metamorphosis is more problematic than expansion...

Unfortunately, continuous expansion without any clean-up can turn the world into an incoherent inflation mess.

Advantages

Inflation solves the long term consequences of hoarding and content starvation without taking anything away from the players by force. If the expansion treats all players fairly this is perceived as being more acceptable than a nerf, even though the overall effect isn't all that different.

Introducing new more valuable content can shorten the gap between older players and newer players by gradually deprecating the old world in favour of a newer world. This assumes that the new content doesn't favour older players over newer players. Unfortunately, reality is often different.

You can keep the world fresh and competitive, which in turn gives you free press.

Disadvantages

The real danger of focusing on growth by expansion is that you create social gaps between groups of players. If the land areas expand too much player density may be too low for good sociability.

You also risk getting a widening gap between hardcore oldbies and casual newbies, creating fractures in the social fabric of established social groups as well as in the overall social sphere of the world.

The addition of new items tend to make older items useless, making the world confusing for both newbies and revisiting players. Revisiting players may feel like newbies and choose not to come back after having a brief look. In addition old play-guides and other non-controlled content become misleading which may cause additional confusion.

And finally, continuous development is expensive. Especially if the foundational architecture isn't solid.

Solutions

Some of the disadvantages can be addressed by recycling and deprecation.

Problem: When adding content it is tempting to focus on high level content. The effect is that casual players feel left out. This increases the gap between harcore and casual, oldbies and newbies.

Solution: Rescaling of the achievement-ladder. Some games do this by extending the number of levels and increasing XP gain at lower levels. Basically, as real time progress, the lower level players get their efficiency improved. This we might call level-inflation. Developers should not forget the lower level players when they design expansions. The synergetic effect is that more low-level content increases replayability.

Problem: Expansions are easier to make than changes, both technologically and socially. Unfortunately, the population growth might not match your content growth. Too much expansion may lead to a less socializable desert, too many trash items and overall lower usability for newbies.

Solution: deprecate content. Maintain several sets of content on all levels, of which one set is meant to be less attractive and phased out. When the popularity of this set is low, remove it or refurbish it.

Problem: Expansions are expensive.

Solution: Plan and design for recycling and refurbishment. High level content that has been deprecated can be introduced as mid-level content, thus retaining those players who never will make it to the highest levels. Removed monsters can be refurbished and tweakd and play secondary roles in new content.

Problem: Expansions might feel like a nerf for revisiting players.

Solution: Design for replayability and let players gain some in-game advantages based on how long they have been subscribers. Even XP or other types of capital.

Checklist

This list is temporary.

• What effects and how varied are the effects of introducing new content?Have you considered all groups, including newbies and revisiting players?

• Have you planned right from the start how to deal with rescaling of character-levels? How far can you expand your level range?

• What are the weak spots in your architecture? Where does it handle growth, deprecation and refurbing? Where does it not?

Related

This article is based on some posts I made to mud-dev in 2004. A level based design is assumed to simplify the discussion, but most of the pattern might apply to level-less designs as well.

devil: worlds as web

What can you learn about virtual world design from the web-sphere? Game design guru Raph Koster sees a lot of connections. Of which some hit closer to home than others. I look at the others... and take the liberty of playing the Devil's Advocate. Feel free to put me in place. Just remember that this Devil has a pitchfork. You have been warned.

From the slides:

The value of the software is proportional to the scale and dynamism of the data it helps to manage.

Hardly! This is 100% contextual and depends on the particularities of the application. Does this even translate to a game or virtual worlds? Are bigger more convoluted games, better? I am sure players and researchers of GO! have a different opinion on the matter. Are bigger and more complex worlds better? Not for a roleplayer it isn't. What happend to the aesthetics of minimalism? When did maximalism become a goal?

The fail fast, fail often method: Users must be treated as co-developers.

Hardly! By exposing the evolution of the system to the playerbase a small segment get the upper hand. From an immersiveness perspective you want the underlying engine to be completely invisible. Bugs and evolution expose the details. You want players to live in a world, not in an engine. Unfortunately, current worlds are too shallow to benefit from this, but that is no excuse! We want the perfect world, after all.

Release early, release often

No, no, no, no, nooo! Release often if you want the user-base to be forum-whiners craving changes to the world rather than experience the world as is. Virtual worlds aren't TV, they are self-contained universes to be explored as is, not a series of soap episodes.

Small pieces, loosely joined: lightweight programming [...] Loose couplings (e.g. XML and HTTP, not SOAP, not custom protocols) [...]

Oh dear! Let's create a massive bandwidth and space problem, right from the start! Anything that is stitched together with a mess of verbose protocols is going to be costly to develop and maintain. XML and HTTP? Hello, welcome to the stone-age! We want compressed binary dedicated protocols designed with a particular world model in mind. Anything less is throwing bandwidth out of the the window in favor of sloppy performance.

Above the level of a single device. Meaning, make no platform assumptions.

Rrrriiight! Let's create the entire game with SMS in mind and then scale it up to a full-blown Playstation 3 world. You cannot design without making assumptions. The more assumptions you can make, the more freedom you get to make something great.

The service automatically gets better the more people use it.

I absolutely loved to see my favourite community being ruined by a swarm of snot-nosed eros-ridden teengers. Not not not! Few people, tends to mean higher identification with the system and better socialization. Call me a snob, if you wish. What's wrong with designing for snobs?

I guess that's enough whining for today. Go read all the slides, not just this tiny selection, it's worth it. Learn from the web if you must, just don't translate it into virtual worlds. Please?

(Disclaimer: I don't have audio hooked up so the above is entirely based on the slides. Which probably sound different than they read. Feel free to flame me over it. I am the Devils Advocate, after all.)