February 28, 2011

Open PvP via randomized Points of Interest

None of the games I know of has gotten open world PvP "right". Large scale PvP usually leads to zerging of enemy players through organized, prioritized, and focussed damage on targets. Clear objective-driven open world PvP areas like WoW's Lake Wintergrasp or Grim Batol result in larger scale battles in certain time intervals and a temporary zone lockdown. Anything else is pretty much "gank & spank". So, how to get it "right"?

Well, juggling around with the new secondary profession in WoW Cataclysm, Archaeology, brought me to a pretty interesting idea that could be worthwhile investigating development-wise. Here's a brief rundown on how the, for a potential open world PvP solution, relevant part of WoW's archaeology works:
  • On a per user basis, users are exclusively assigned certain spots (small areas close to ruins, skeletons, excavations, etc.) within the game world where they can dig for artifact fragments.
  • After digging up fragments three times, the spot disappears and a new "random" exclusive digsite is made available to the user somewhere else within the game world.
  • At all times, there are 3 digsites per continent available for each user.
Here's my rough concept for a potentially intriguing open world PvP system, assuming a 2-faction setup like "Orcs vs. Humans" or "Guardians vs. Defilers":
  • 1 up to 5 players per faction can group up for an artifact (PvP) quest.
  • Two equal-size and equal level (range) groups, one per faction, are assigned one random digsite within the game world.
  • The first of both teams that has acquired, e.g., 5 artifact fragments can claim the artifact, and thus a reward (loot, gold, resources, whatever is valuable within the game world).
  • Artifact fragments can be acquired by digging them up or by looting them from killed enemy players.
  • Only members of both groups can loot artifact fragments from enemy players or by digging them up, however there are no more than 5 fragments per digsite at a time.
  • Basically, as soon as a team's player has at least acquired one artifact fragment, the other team will need to engage in open PvP with the other team in order to be able to claim the reward.
  • As soon as either of both teams has claimed a digsite reward, a new "randomly" selected spot is made available as a new digsite for both teams unless one team's group is disbanded (in which case the remaining team can queue up for yet another artifact quest).
To prevent domination via bringing "friends" for fast artifact acquisition, the system could only allow users of both groups to damage each other and make them immune to enemy player damage as soon as they enter the zone the artifact site is located in.

This could result in small-scale team vs. team battles at randomized locations throughout the whole game world, and actually provide a potentially endless meta game with a constant flow of ongoing and recurrent small-scale battles within the game. It would constantly re-use existing content while providing participating users clear objectives and rewards. It might also enhance community building and social interaction, as coordinated teams will likely be more effective than random pick-up groups.

December 17, 2009

Resolving Goldfarming through Concealed Validated Auctions

Introduction to the problem


The exchange of goods is an integral part of any social system. People buy goods to satisfy their personal needs, to honor their traditions, or to strengthen their social bonds. People sell produced goods to make a living, to pay their monthly bills, or to make profit. Community members can obtain wealth through adherence to the overarching system's ruleset, through contribution to the ongoing persistence of the system, and through the accumulation of commonly agreed-to status indicators, e.g. the amount of children, land ownership, crop seeds, or money. The same applies for online world societies.


Online world users, however, can as well take a shortcut. They can have others obtain those status indicators for them in exchange for surplus acquired through their regular day-to-day job. If used in ways that are not intended by the game developers and not under the control of the game operators, this so called real money trade (RMT) threatens the integrity of the affected online world systems. RMT contractors accumulate money used in the real world by exploiting resources provided within online worlds. That way, the system-internal online world economies are flooded with easily obtainable goods and commodities. The values of rather common goods are highly deflated. Uncommon commodities become highly inflated as users seek to reduce their losses through obtaining rare goods.

This so-called MUDflation causes frustration for both the users and the operators of online worlds. The efforts and commitments of the users to the integrity and well-being of the online world system are devalued, which reduces the overall systemic quality and social stability of the online worlds. The operators of online worlds have to invest substantial funds into combating real money trade while simultaneously facing the challenge of users leaving their online worlds due to RMT-based frustration. Their costs increase while their revenues decrease. Under careful calculations based on previous research by Prof. Edward Castranova, the annual losses caused by RMT add up to at least $300 mio., but very likely significantly more. Obviously, a way to resolve that issue has to be found.



Approaching the solution


One way to approach the problem is via the analysis of the process performed throughout an RMT transaction. Each real money trade transaction consists of two transactions, one per system involved. First, the buyer transmits the agreed-to amount of money to the seller, e.g. via credit card payment, and provides info on his virtual identity, e.g. the name of the buyer's avatar. Second, the seller transfers the acquired virtual commodities to the buyer via means provided by and within the online world, e.g. via a trade window or a PO box. Game operators may try to resolve the issue of RMT by not allowing any trading at all between avatars within their online world. As stated above, however, the exchange of goods is a vital subsystem of any society. Taking it out of the online world would greatly reduce the amount of social interactions performed.

The problematic interaction is the verifiable exchange of goods between two specific online world users. If there is a way to verify a transaction within an online world, both buyer and seller of an RMT service can track the completion of an ex ante agreed-to and paid-for exchange of goods. Thus, there must not be any exception to the following described mechanism. It has to apply to all transactions. Else, the exception, e.g. an allowed direct exchange of "cute glade flowers", will become the norm, glade flowers will become the new inofficial currency, and users will start complaining about RMT glade flower farmers.


Several identifiers can be used to verify a transaction: location, price, name, and time. Detaching transactions from this trade information could provide means to enable trading within online worlds, but disable trading between two parties involved in RMT. This full concealment of all transaction identifiers could be achieved via the following mechanism.



The proposed solution: secret auctioning


All avatar-to-avatar transactions are performed and mediated via a trade platform that establishes the exchange of goods between designated trade partners. To conceal the time identifier for transactions, all trade is performed after the expiry of a trade cycle, e.g. with a duration of 24 hours per cycle. At the beginning of each trade cycle, all goods put up for sale before the initiation of the trade cycle are listed on the trade platform. Additional amounts of commodities that were already listed at cycle initiation may be added at any time while the trade cycle lasts. New commodities will be listed at the initiation of the next cycle.

The actual amount of offers placed is kept undisclosed, as well as any information regarding the demanded price and the identity of the seller. Users can determine which commodities are listed for sale at trade cycle initiation, but cannot identify prices, seller identity, or available amount. Suppliers may modify their price at any time while the trade cycle lasts. They may not, however, offer the same commodity for different prices, e.g. one shoulder armor for 10 gold, one for 7.5 gold, and one for 5 gold pieces. This similarly applies to buyers: if they are interested in buying a commodity available on the trade platform, they can place one secret bid per commodity, and modify it anytime while the trade cycle lasts. That way, both suppliers and bidders reveal their actual individual willingness-to-pay or willingness-to-be-paid-for. All information that might lead to the identification of the buyer or seller, e.g. price, name, or time of bid / offer placed, is kept secret.

During the trade cycle, involved users reveal their individual willingness-to-pay or willingness-to-be-paid-for to the trade platform and deposit the according items or funds. The price information can be used by the platform at the end of a trade cycle to aggregate supply and demand and to determine the market equilibrium for each commodity, based on full market information, which is only available to the automated transaction mediator.

So far, however, this would not prevent certain RMT-related transactions where a user would offer a commodity of very low value which the RMT seller would then buy for a very large amount of online world money, which would effectively establish a real money trade transaction. This issue can be addressed via a transaction validation procedure.



Increasing economic stability through a transaction validation procedure


Computer-controlled shopkeepers, so-called NPC vendors, provide according means to validate transactions. Acting as buyer machines, they usually pay rather unsatisfying prices to the users, based on a formula that determines the base value of each commodity. This formula greatly varies from online world to online world, and can be based on various factors like i.e. rarity, commodity properties, or avatar requirements. Via that formula, computer-controlled vendors effectively establish a minimum price for each commodity. There is no need to keep this minimum price information undisclosed to the participants of the platform-mediated trade. It provides an anchor, a minimum relative orientation for the initial evaluation of each offered commodity.

Based on the trade data input by the users, the supply and demand curve and the market equilibrium for every single commodity can be determined. In addition, NPC shopkeepers establish a minimum price for each commodity via the general base value formula. Based on that information, a range of valid transactions can be identified.



The intersection of the minimum commodity price and the supply curve (pmin = S, xmin) is declared the minimum valid price. The intersection of the demand curve and the minimum amount of performed transactions (pmax = xmin, D) is declared the maximum valid price. That way, all offers and bids between the minimum valid price / maximum valid bid and the market equilibrium are declared valid. All remaining offers and bids are declared invalid. The according suppliers and bidders are refunded their items and funds. Valid bids are sorted in descending, and valid offers sorted in ascending order.


Suppliers with low validated prices have, in relative comparison to the other trade participants for the according commodity, proven their commitment to the integrity and well-being of the online world's economy. This similarly applies to bidders with a relatively high willingness-to-pay. They did more than they were required to and are thus rewarded accordingly: the lowest validated supplier is served first to the highest valid bid and so on, until a transaction cannot be established any longer due to missing commodities or bids.

A trade cycle ends. All validated transactions are performed, all invalidated offers and bids are refunded to the according users. No-one knows who bought what from whom but also, no-one is worse off than before participating in trade. All bidders and suppliers are either refunded their deposit, get what they wanted for the price they were willing to pay (bidders), or get even more than they asked for (suppliers). The Pareto-criterion is met. Due to first degree price discrimination, the welfare within the range of validated transactions is maximized (economists like this). A new trade cycle begins. All market power and information obtained during the last trade cycle is reset, while all market experience is kept. Effectively, a non-discriminatory black market is installed.



Conclusion and criticism


By applying the above described mechanism, online world users can still exchange goods between each other, maintain and nurture an integral part of their day-to-day interactions, and work to achieve goals related to this subsystem of the online world community. As a trade-off, users would have to wait until the end of a trade cycle before they may receive their goods. Immediate needs could no longer be met. In addition, this mechanism would prevent any form of materialistic gifting. According social interactions would need to be replaced, e.g. by shared collaborative activities, which might actually prove beneficial to the online world's society.


The goal of the mechanism is achieved. Illegitimate or unintended RMT transactions are effectively eliminated: when this mechanism is used, there is no way for suppliers and customers of RMT services to verify the delivery of goods within the online world agreed-to and paid-for outside it. The economy of the online world is kept unbiased from interests external to the online world. The integrity of the online world's economic system is preserved. Meanwhile, trading becomes a game, based on secret bids and educated guesses.



author's bio: Jan Pontzen, Dipl.-Kfm. (German equivalent to MBA), graduated in Media Economics at Ilmenau University of Technology, Germany, in March 2009. The full master thesis that this article is based on can be downloaded here. Jan is currently working as an Assistant Product Manager at Acony, a Germany-based games developer that is readying up Parabellum, a free-to-play tactical MMOFPS, for open beta.

June 11, 2009

Capture the Egg v1.0 release

Times of peace are hard times for true heroes and adventurers. No greater cause to join, no tales of their heroic deeds, no princess to save and rescue. In times of peace, true heroes become mercenaries for any cause. They help farmers getting rid of the wolves that feast on their cattle. They brawl with other adventurers to get a stake of the money that was bet on them. They fulfill any task, any opportunity, that helps to pay the tavern bills and their maidens.

The greater the risk, the better the wage. One of those high risk, high wage opportunities for true heroes lies beyond the cold and harsh valleys of Icewind Gorge. There, the frostwyrms reside to hatch their eggs within their temples. Powder scratched off the frostwyrm eggshells is a top quality ingredient that skilled craftsmen like Gizz the Alchemist and the dwarven engineers Ulf & Grim pay a high price for.

You, adventurer, want to earn a wage that keeps your oven warm and your maiden fed throughout the winter? Then come to Icewind Gorge and earn your reward! Join Gizz's or Ulf & Grim's band of mercenaries. Gather eggshell powder from the frostwyrm eggs. Of those that try to stand in your way, let none survive. A freshly brewn keg of beer and a beautiful maiden are waiting for you.

We are very happy to announce the release of Capture the Egg, a fast action team vs. team RPG map for Warcraft III: The Frozen Throne. This map features intense team vs. team PvP action with some of most unique and compelling classes you'll have ever played on the Battle.net and far beyond. Each class has individual class mechanics with abilities specifically tailored to each heroe's playstyle.

Capture the Egg features

  • intense team vs. team PvP action
  • king of the hill ruleset
  • unique, high-quality, and compelling classes
  • individual and playstyle-tailored class mechanics
  • detailed player statistics scoreboard

Download links for "Capture the Egg" v1.0:

May 19, 2009

CtE release "soon"

After a lengthier job-related break on the "Capture the Egg" Warcraft III mod project, Tzirrit and I got back to this uncut diamond a few weeks ago. I am happy to announce that the diamond is now approaching its final polish and that we will release version 1.0 of "Capture the Egg" "soon", meaning when we feel it is ready to be released to the public (which actually really is "soon").

"Capture the Egg" will feature intense King of the Hill ruleset Team PvP action with some of the best RPG classes you'll have ever played on the Battle.net and far beyond. Feel free to check out the class sneak peak for the Protector, the Thunderer, the Crusader, and the Berserker on the project website. Additional unique, compelling, and exceptional quality classes are planned for post-release patches.

I'm officially smart

Just graduated, well actually about a month ago. I am now a "Diplom-Kaufmann", the German equivalent to the Master of Business Administration.

The topic of my thesis: "The Development of a Price Control Mechanism for MMORPGs", which you can download at db-thueringen (click on "[Dateien anzeigen]" to view the .pdf file).

June 09, 2008

Capture the Egg

The Icewind Gorge is a menacing place. A place from which the living do not return. A place where the icy wind chills the breath of the incautious adventurer. Atop the peak of which the Icewind River emerges, Gmaug the Broodwatcher, keeper of the frostwyrm hatchlings, resides. Only during a short period in the winter, when the cold chill even partially freezes a dragon's wings, he returns to Icewind Peak from his flying patrols over the slumbering broodmothers, resting uneasily for a change towards a less chilling howl of the winds.

The moment scouts of the Goblin Henchmen and the Dwarven Merceneries have been waiting for months has come. A short timeframe has opened up to steal frostwyrm eggs from the slumbering Broodmothers, eggs of unmeasurable value to alchemists and nobles. Dispending messengers to the taverns in the south, they place large bountys on every frostwyrm egg being brought to them.

Despite the danger of angering the Broodwatcher, driven by the opportunity of great wealth and celebrity within their country, groups of mighty warriors gather at the Icewing Gorge to fight for the frostwyrm eggs. To bring their hero's journey to completion. To seek a fortunate ending for their adventurous lifes.

Prepare for battle, adventurers! The coldest time of winter in Icewind Gorge will soon have come. Tales will be told about your theft of the frostwyrm eggs. Tales of brave men that escaped the anger of Gmaug the Broodwatcher. Tales of wrathful screams of the Broodmothers, mourning the loss of their hatchlings. Tales of heroes. Heroes that captured an egg.

Capture the Egg - a Warcraft3 map project by Tzirrit and Tjuhl. Release 2008.

April 24, 2008

crowd control

Crowd control (CC) is always a highly discussed topic in MMORPGs. Crowd control means that certain classes or roles have the opportunity to hinder enemies in their actions to get a strategic advantage over them for a limited amount of time. It is usually annoying for the affected, but simultaneously an opportunity to change the outcome of a fight. It thus increases the fun for all involved players, because it makes overall combat more spicy for all involved players... if done right.

I'd suggest the following approach: give all classes CC according to their role to allow every player to perform according to / consistent with their role.

Thus, we 1st have to determine what exactly the different roles' job is. In MMORPGs, we usually differentiate between Tanks, Melee DPS, Ranged DPS, and Healers. The later can be distinguished in ranged healers and melee healers which should greatly affect their overall playstyle.

Tanks are heavily armored melee fighters. Their main purpose is to protect allies by soaking up and/or reducing overall damage. This can be accomplished by two means: single target protection and general protection. Single target protection is achieved by selecting an ally that is in need of protection (e.g. she is taking a heavy blow from enemies) and then performing actions that reduce the damage dealt to her by transfering it to the tank. General protection means giving enemies incentives to focus their attacks on the tank (of which the tank then mitigates a considerable amount through high armor / avoidance). Thus, concerning crowd control, a tank should be able to generally annoy enemies to turn their hatred towards him. How do you generally annoy enemies? 1St, they have to be annoyed, not taken out of combat. They have to actually be able to turn their hatred towards the tank. Consequently, a tank should slow down enemies, reduce damage afflicted to allies, block ranged attacks, and reduce overall enemy stats. This would lead to the following crowd control abilities:
  • PBAE and/or cone snare („Come to me!“)
  • PBAE and/or cone damage reduction to other targets than the tank („Hit me!“)
  • interception of ranged attacks („I'm your target!“)
  • interception of melee attackers („Stay with me!“)
  • damage transfer from a single ally to the tank („I'll take that punch!“)
Melee DPS classes are medium-armored melee fighters. Their main purpose is to inflict high damage to single targets in a short amount of time, close combat. They have to be very agile and able to close distances to enemy targets. When they are close to their target, they also have to be able to stay close to that target until their task is accomplished. For crowd control, that would lead to the following abilities:
  • anti-CC abilities to stay agile (desnare, break root) („You'll pay for that!“)
  • sprint to close distance gaps („Here I come!“)
  • short-time knockdowns, stuns („Hold still while I punch you!“)
  • interrupts („Don't do that, my turn!“)
Ranged DPS classes are fragile ranged damage dealers. Their role is to inflict high damage to multiple targets from a distance. While performing attacks, they usually have to stay in their current position. Due to their very low armor and fragility, they should try to stay away from enemies and continuously readjust their range. To stay away from threats, they have to be able to readjust the correct distance to them. If they fail to do that, they will die. Thus, the crowd control abilities should be as follows:
  • short-time single target root („Stay away from me!“)
  • short-time single target snare („Slow down there, let me finish that!“)
Healers are different. Their job is to keep allies alive by all means. Because of that ability, they are often the primary target for all enemies. If they are given direct crowd control abilities that put enemies out of combat, they would be too strong since they would have two means to greatly reduce damage: through healing allies and through taking enemies out of combat action. Thus, their crowd control abilities should be focussed on self-protection and recovery. To keep that consistent with their role, we should differentiate between melee healers and ranged healers here.

For ranged healers, the general class mechanics for ranged DPS apply: they need to stay away from enemies to fulfill their job. Concerning crowd control, they would then need the following:
  • pushbacks („Back to where you came from!“)
  • quick recovery from knockdowns, stuns, silence, and interrupts („Let me do my job!“)
  • emergency self-heals / self-protection („I'm not done yet!“)
Melee healers are designed for healing in the thick of things. They should not have to flee to stay effective. In general, they need a base mitigation that is similar to the base (!) mitigation of tanks. Similar to them, they need to be able to take a punch or two. This is because they are subject to several kinds of attacks at the same time: ranged and melee DPS focus damage, enemy tank annoyance, general ranged DPS AE damage. In comparison to melee DPS classes, they cannot be as mobile as them since they need to stay in range to many allies to get their heals through. Thus, they need a wide range of self-protection (because, again, direct crowd control would be too overpowered):
  • high mitigation / armor („I can take a few!“)
  • short-term damage immunity („Not now! Go to that other guy (tank)!“)
  • quick recovery from knockdowns, stuns, silence, and interrupts („Let me do my job!“)
  • high interruption resistance („That doesn't disturb me.“)
  • emergency self-heals / self-protection („I'm not done yet!“)

March 27, 2008

Why collision detection is NOT enough...

... and why you, imho, ALSO need (rudimentary) line of sight.

let's talk about the tactical purpose of melee characters. They basically build a front line that
  • constructs a mobile wall for enemy players
  • lures the enemys attention to them
  • enables wounded players to back up a little bit to (get) heal(ed)
the later two are only possible when, in addition to collision detection, a line of sight mechanism is implemented.

by collision, you only hinder enemy movement and thus melee characters. to provide opportunities for retreat, you also need to encumber ranged characters. that's what line of sight would be for then: if there's an obstacle b/w the ranged damage dealer / source and the target, the effectiveness of ranged damage sources should be reduced according to the degree of encumbrance. e.g. if a melee character / obstacle is very close to (quite far away from) a ranged damage dealer, his/her damage and/or chance to hit is significantly (slightly) decreased. to simplify, this may only be applied to player-controlled enemy (non-friendly) players.

implementation of line of sight AND collision detection would, imho, open up lots of interesting and highly enjoyable opportunities concerning tactics and overall gameplay:
  • searching for weak spots in the enemy lines
  • setting up traps or "feints"
  • flanked attacks and other maneuvers (to break through to the injured / ranged etc.)
  • give incentives for melee players to move the front line closer to the enemy artillery :)
  • provide a more realistic feel of combat
  • etc.
line of sight would only apply to ranged attacks, since melee are already hindered in several ways:
  • they have to be within melee / weapon range
  • they have to overcome additional defense (parry, block) at frontal attacks
  • they need to find a way to attack from the side / from behind to unleash their fury... ;)
  • ... while, simultaneously, trying to avoid turning their back / side to enemy players
also posted on the WarhammerAlliance forums.