An Introduction to distribution systems in MMO’s
1 The Need for Loot-Distribution Systems
In Massively Multiplayer Online Role-Playing games (MMO’s) the toughest challenges that offer the greatest rewards require the combined efforts of many players. In World of Warcraft (WoW), these challenges take the form of ‘raid instances’. Raids consist of pre-determined challenges, or ‘bosses’, that when defeated leave behind items, or ‘loot’, that the players can use to make their characters more powerful. How to distribute this loot between the participating members, or ‘raiders’, has been a long-standing and problematic concern, and one that has yielded a variety of attempted solutions.
The problem originates in the disparity between the number of raiders needed to defeat a boss and the amount of loot obtained. In their original form in WoW, raids consisted of 40 man teams and each boss would drop only 3 or 4 items when defeated. The difficulty of these raids is such that it would take a great number of unsuccessful attempts before a boss could be defeated for the first time. To tackle them, players form themselves into communities, or ‘guilds’, which aim to maintain a core team of participating raiders so that the guild could raid over several nights a week and progress towards defeating these bosses. The more powerful a guilds individual raiders, the better equipped the guild is as a whole to tackle more difficult challenges. Thus loot distribution is important in two respects: Firstly, loot is the primary source of character development for individual players, making loot acquisition desirable from an individual perspective. Secondly, by gearing up its members a guild can tackle new challenges and so loot distribution becomes important from a collective perspective, to find the optimal method to facilitate this aim.
The game itself provides a simplistic random number generator (RNG) based distribution system, termed ‘Need before Greed’. Under this system when an item becomes available players can select whether they ‘need’ the item or just desire it (‘greed’) and then roll a dice with the highest roller winning the item, with any need roll taking priority over any greed roll. In groups of unassociated players where individual gain is the priority, defection in the system is frequent. The term ‘ninja looter’ was coined to refer to people who would at the last minute need roll on an item when all had agreed to greed roll, thereby winning the item by default. Within guilds, such defection would inevitably lead to a player being expelled from the community, but more general concerns about the roll based system became manifest. Was the system fair to those players who put in more effort than others? Did the system reward loot to those who deserved it most? Did the system facilitate the overall aims of the guild? Such concerns led guilds to seek out alternative systems, an overview of which is provided in the next section.
2 DKP Systems
The first documented implementation of a player organized loot distribution system within an MMO is the “Dragon Kill Points” (DKP) system created by Thott of the guild ‘Afterlife’ for the game Everquest in 1999. Since this time ‘DKP system’ has become an umbrella term for a multitude of variations to Thott’s original system, which share the common theme of using DKP as form of currency but differ in whether these points are earned and spent in a varied or static manner.
2.1 Original DKP
Thott’s original DKP system is arguably one of the most simple and uses a combination of static earning and static spending (SE/SS). Players participating in raids are awarded a fixed amount of DKP per boss kill, which can then be spent on the boss loot for a fixed amount of DKP per item. When there is competition between players for a particular item, it is awarded to the player with the greatest DKP.
This method of DKP acquisition quickly became problematic for several reasons. As DKP is only earned per boss kill, it fails to incentivize and reward participation in the progress period of a guilds activity. At the start of a new raid tier, guilds need to learn how to defeat the new boss encounters and large amounts of time are invested by players without reward. Under this system there is no way to differentiate between the hard-workers and the free-riders. Those who invest the time to allow the guild to succeed are no better off than those who join in only when the boss can be defeated.
A related issue is caused by the weekly reset nature of raids in WoW which causes a large imbalance in time invested per DKP earned under this system. With raid instances typically containing multiple bosses and the difficulty of the bosses increasing throughout the raid, it is almost always the case that guilds progressing on an instance will be able to regularly defeat a few bosses whilst still progressing on a more difficult one. Consequently, the bosses the guild has already killed are quickly re-killed the following week and the rest of the week’s raid time is spent on progression. This means that the first raid of the week is far more profitable than subsequent ones and players who can attend the first raid are able to earn significantly more DKP than those who can only attend later in the week. Similarly, the system encourages a high attendance for the week’s first raid but fails to encourage further participation, which often stagnates a guilds progress.
A solution to these problems is to award DKP for time invested rather than per boss kill. This way, players have an incentive for participating in all raids and players with a greater attendance are better off, as with more DKP they have greater chance of acquiring loot when available. Indeed, many guilds employ a bonus structure, whereby guild members will receive bonus DKP for factors such as punctuality and good performance, while taking away DKP for lack of preparation and poor performance. Today, almost all DKP systems with static earning award DKP primarily for time invested rather than boss kills, although other systems will be examined which still employ boss kills as a central factor.
However, even with this modification SE/SS systems face problems with inflation and the related problem of the ‘DKP Gap’. Inflation occurs when the amount of DKP entering the system exceeds DKP expenditure. For static systems that award DKP for raid attendance and have fixed prices for items, inflation can quickly become a large problem. The relative value of attending each raid is quickly diminished as players accumulate large amounts of DKP, which in turn leads to little to no fluctuation in the priority list for loot. Players with high attendance early on can easily monopolize the system.
DKP taxation and price increase proportional to inflation are two potential solutions to this problem. However the added complication of these considerations typically means they are an undesirable option. Further, the probabilistic nature of which loot becomes available calls into question the fairness of taxation. A particular boss will have a list of items that may become available when it is killed, and a small number of these are randomly selected. It is not necessarily true that a particular player with a large DKP lead is guilty of DKP hoarding to affect distribution priority. It is perfectly possible that such a player is simply unlucky, and the items that they desire have thus far never been available. In such a case, any taxation policy will unfairly disadvantage veteran players.