Guide no image

Published on May 18th, 2006 | by Matt Haberfeld


Finding a DKP system that works for your guild

This article is dedicated to explaining DKP systems, why they exist, the problems that they solve, and listing their pros and cons. After reading this article, you should have some idea of what will work for your guild and how to implement it.

Loot Systems before DKP

Loot systems are a very new concept in MMOs. The fact is that for a long time there was no need to have a formal loot distribution system. Everquest was the first game to introduce this idea, and the concept flourished in World of Warcraft. The reason for a loot system is that in Everquest, the really tough monsters of the world required a large amount of teamwork to bring down. Lady Vox and Lord Nagafen required dozens of players to take down the first time, and yet they only dropped a few items each. How do you fairly distribute 5 items between 40 players?

When I played Everquest, even the big raids used random rolls to give out equipment. All players that could use the item rolled between 1 and 100, and the highest random roll received the item. Sometimes we capped the number of items a single person could receive per run, sometimes not. This was acceptable for two reasons:

1) The same people raided weekly and understood that they would eventually get their items.
2) No one had developed a better system.

Random rolling was a fair loot distribution system in the extremely short term. When 40 people go into an instance, it’s because 1 person can’t do the instance alone. Nor could 5, 8, or 10. You may not need all 40 people, but you invited 40 and they are helping to clear the dungeon just as much as you are. So they should get an equal chance to roll on items that drop. However, as weeks turned to months, it became more apparent that a better system would be required. Random rolling can produce an extremely uneven distribution of loot. The same player could conceivably win an item every week, while another player won nothing for two months. Only as the number of items approached infinity would the distribution begin to reach uniformity.

DKP is a loot system that attempts to distribute loot fairly over short and the long term. DKP originally stood for “Dragon Kill Points” in Everquest, and stands for “Dungeon Kill Points” in World of Warcraft. The DKP system is just as “fair” as random rolling, but the definition of fair is completely different. DKP systems attempt to account for time spent in an instance, and previous loot won from the instance.

However, DKP has come to mean much more than that in World of Warcraft. World of Warcraft is a very loot-centric game. Once you hit level 60, the only way to advance your character in any meaningful way is to get better equipment. This is the facet of the game that creates more greed than any previous game could possibly hope to achieve. Therefore, DKP systems are much more than simple rule systems to govern the distribution of loot. A good DKP system ensures the survival of the entire guild by preventing the greediest players from taking everything, by encouraging raid attendance, and providing a clear measurement of where each player stands in relation to his peers.

How does DKP work?

The way DKP works is that every item on the monster’s drop table is assigned a value. Players receive points for time spent in a dungeon, and can spend those points on items when they drop. The player with the most points gets priority on the item, because that player has spent the most time in the dungeon without receiving an item. When the player takes an item, the points are deducted from his total and he has to wait until he is in the lead again before he can take another item. This helps to ensure that the items are distributed fairly evenly among the raid, and no single player gets all the equipment. It helps attendance because players cannot earn points unless they help the raid make progress in the dungeon. It also helps the guild as a whole because everyone knows where they stand in relation to their peers.

So here is an example of a DKP system:

Rule 1: Each player earns 1 point for 1 hour spent in the dungeon.
Rule 2: The player with the most points has priority over items when they drop.
Rule 3: In the event of a tie, players roll and the highest roll receives the item.

So three brave Warriors are competing with each other for items in Molten Core:
Warrior A: 3 points
Warrior B: 4 points
Warrior C: 5 points

Lucifron drops the Gauntlets of Might, which cost 3 points. Warrior C has the most points and takes them.
Warrior A: 3 points
Warrior B: 4 points
Warrior C: 2 points

Magmadar drops Legplates of Might, which cost 4 points. Warrior B has the most points and takes them.
Warrior A: 3 points
Warrior B: 0 points
Warrior C: 2 points

The next Warrior drop will go to Warrior A, and then the next drop to Warrior C, etc. Players are earning points as they clear the dungeon, whether Warrior loot drops or not.

Balancing DKP systems can be tricky. First of all, you have to know the values of all the items that can possibly drop from the monster before you kill him. Second, you have to match the item values to the amount of time it takes to kill the monster to get them to drop. If the item values are less than the time it takes to kill the monster, then everyone’s points begin to inflate, which is demoralizing to new players starting at 0 points. If the item values are more than the time it takes to kill the monster to get them to drop then lots of players end up negative and brand new players get priority on items. So far there have been two modifications to the DKP idea to fix this problem:

Bidding DKP

With bidding DKP, there are no item values, nor a need to know what any particular monster drops. When an item drops, each player bids the number of points that they wish to spend on an item and highest bid wins. In the case of a tie, both players randomly roll and the high roll gets the item. This is a much easier system to deal with, but the one big drawback is collusion. Collusion is the fancypants term for exploiting the system for personal gain. With a bidding system, all of the members of a single class can agree to bid 1, or not bid at all, and save their points for weapons used by multiple classes, etc. Believe me when I say that players will exploit the system. The only way to counteract that is to institute a minimum bid rule for each item, but then you once again need to know the values of all the items that can possibly drop. So basically the only advantage of bidding DKP is that it ignores inflation, and you introduce the huge disadvantage that the system is exploitable.

Zero Sum DKP

With a zero sum DKP system, no one ever earns points. Instead, when an item drops, the winner gives his points to all the other players in the raid. It’s a bit much to wrap your mind around the first time, so here’s an example:

Fourty players in a raid, and an item drops. The item is worth 40 points. Warrior A wins it, and now he has -40 points. He gives each person in the raid 1 point, so everyone has 1 and he has -39. When you add up the points of all the players in the raid, they equal zero. That’s why it’s called zero sum.

This system also removes the problem of inflation. However, it has its own problem. The system does not reward attendance unless there is the guarantee of loot dropping. “Wipe nights” do not reward the most dedicated raiders, and this can really hurt a guild’s progression through an instance.

Developing a DKP system that fits your guild.

World of Warcraft is unique in the fact that the game is easy enough and addictive enough to draw in a large demographic. You will get boys and girls, men and women, level headed adults, college students, teenagers and 8 year olds, Brazilians, Chinese, Russians, and Americans. Players will vary wildly in ability to speak English, understanding of game mechanics, and maturity. There is no single loot distribution system that works for everyone. As a guild leader or officer, you need to try to understand what your guild needs and what system is best for them. Try answering these questions:

Attendance – How can I use the point system to keep people attending the raids?
Loot Distribution – What is the best way to distribute the loot evenly, yet still allow for raid progression?
Reward – Are players guaranteed a reward for their efforts? Is that reward worth the effort?
Security – Can this system be exploited?
Fairness – Is the system fair for me?

Fairness is the one thing that ultimately every system lacks. That is because there is no fair way to distribute loot. Every item someone gets is an item that everyone else was not allowed to have. This will make someone unhappy. You have to choose the system that is acceptable for the greatest number of people, or the system that works the best for your guild (even if it’s not “fair”).

My guild’s DKP system

I felt that it would be more useful to explain my guild’s DKP system so you can see the sheer amount of ridiculous crap that you have to go through to make it work for everyone with a minimal amount of complaining. As stated previously, World of Warcraft is primarily a greed driven game, so you have to be very careful about the system you implement to try to please the greatest number of people and keep the system from being exploited.

Our first question as a guild was which DKP system to use. Previously, we had been using a standard DKP system where you got 1 point for every 2 hours, but we ended up with a lot of inflation. New players had no interest in raiding when the veterans had 70 points to spend. So we decided to switch to zero sum, because we had a few unscrupulous raiders and bidding DKP would never work for us.

So after deciding zero sum, another question of monumental importance arose. Were we going to use a single zero sum system for every dungeon in the game, or was each dungeon going to be its own system? A combined DKP system really rewards hardcore players for being hardcore. They can attend raids in which they don’t need anything just to earn points to spend in the higher instances. They can also hoard points for instances that don’t even exist yet, so that they can get the newest and the best items.

If each dungeon is its own DKP system, then everyone starts the dungeon on equal footing. The hardcore players will still rise to the top because they attend the most raids, but it’s much less pronounced, and they may lose a few high prestige items to more casual players.

Our guild varies wildly in character progression, so we opted for each dungeon being its own DKP system. Sometimes this does hurt us, but for the majority of the guild it is a good thing. However, this alone was not enough. Players in greens want to sign up for Molten Core, and Blackwing Lair, and double gear their characters. It’s not really a problem for experienced players, but the only types of players that would do this are usually the types of players that only want items and don’t really know how to raid. So we had to post and enforce minimum gear requirements before players were allowed to sign up for a raid. This prevents double gearing and gives everyone a clear indicator of where they stand in relation to other players.

We quickly discovered another problem with a zero sum system. Players only earn points if they are in the instance when an item drops. Our guild has more than 40 players online ready to raid at any given time, and those players were not getting points, even though we regularly need as many as 55 raiders just in cause people cannot attend.

If we did not allow players to earn points outside of the instance, we would lose those players to other raiding guilds. Therefore we decided to allow players beyond 40 to sign up and be “on call” during the raid. If someone disconnects or has to leave, we have a replacement waiting. Players on backup earn points as if they were in the instance, so it’s a pretty sweet deal. In fact, it was too sweet a deal and many people complained that players could earn points on backup without paying the repair bills. So we added the caveat that players on backup must pay the raid bank 30g worth of raid materials. This can be potions, crafting materials, repair bots, etc. Already the system is terribly complex.

Next the question of alts vs. mains arose. Should the point system treat each alt as a brand new character, or should the human playing the character get the points to distribute as they see fit? Again, here is a chance to reward hardcore players for being hardcore if it’s best for the guild. Allowing the human playing the character to earn the points means that the hardcore players will usually have a lot of extra points when they get everything they need on their main. This allows them to get a few items on an alt if they want. Some people may consider this loot “wasted” but players get burned out on playing the same character over and over and a change of pace can keep a person playing the game when they would have otherwise quit. What good are epics on a character if the human playing quits the game?

If each alt is treated as a brand new entry into the DKP system, then you are starting from scratch all over. This has its benefits as well, as any alt that takes gear has been there long enough on that character to earn it.

Although we decided not to reward hardcore players on the combined DKP for all dungeons, here we decided that points should belong to the human playing the character. This has resulted in drama once every few months, but for the most part it’s a better system and as long as the players respect each other people don’t massively gear up their alts over mains. Since alts keep the hardcore players playing and not quitting, it was the best decision for our guild even if a few items are not put to best use.

The final question was whether we award points on “wipe nights” where we are learning an encounter but no loot drops. Even the most hardcore raiders can get frustrated by not visibly earning a reward for their efforts, so as a guild we decided that DKP would be awarded on wipe nights. It’s usually equal to the DKP as if we had killed the boss we are attempting. Now this destroys the zero sum part of our DKP system, but so did awarding points to players on backup, so although it behaves like a zero sum system, the points do not actually add up to 0. It’s really not that big a deal.

Further Customizing the DKP System

We now had a system that worked for our guild. So we raided and raided and watched how things worked out. It became apparent almost immediately that certain players were hoarding their points for high prestige items off the last boss. Now, players are allowed to do this according to the rules of the DKP system, but it’s not a good thing. Unspent points don’t clear the dungeon any faster, in fact they don’t help your character in any way. Items that could be used by players were rotting because too many people were saving for the rare treasure at the end. This had to stop or we’d never make any progress in the dungeon.

Our solution to this problem may not be the best, but it works for our guild. We created 3 DKP charts, one for class armor sets, one for resist gear, and one for multi-class gear, which is basically everything else. Since resist gear only helps the raid clear the instance faster, and has no use outside of the dungeon, we allowed people to take it and those points spent would not affect their standing on weapons and armor. We also separated class sets so that people would not forsake important armor upgrades to save for a weapon. This DKP distinction has solved some problems and caused some new ones. Although items no longer rot, players are allowed to triple gear their characters from 3 separate DKP charts. Although it seems like a bad thing, it’s only a problem in the extreme short term and we deal with it on a case by case basis. Also, if players still refuse to gear up on the lower prestige items and insist on competing with the hardcore players for the extremely rare drops, the officers can elect to put him on backup until his attitude changes or they feel he deserves the next item.

Choosing DKP Values for Items

This is a really daunting task for anyone. Not only do you have to assign values for items that you want for your class, but you also need to understand what items are the best for classes that you’ve never even played. I will briefly explain the system my guild uses and hopefully this will give you a solid starting point to assign values to items for your guild.

The first thing I do is get a complete list of all the items that drop in the instance. Thotbott and Allakhazam are very good for this, and for the very newest items, sometimes you need to go to the public test realm forums or an impromptu website that lists everything. If you can’t get a complete list, just get as many items as you can, so you have some idea of if these items are big upgrades from the last instance, or just sidegrades.

Next, you work with the easiest stuff first. Any items that are only useful for elemental resistance or PvE progression have low point values. Since my guild uses zero sum, the lowest point value possible for us is 40 (we’re fraction bigots). So items like Flamewaker Legplates, Black Ash Robe, Necklace of Purity are all 40 points. Also included in this list are items that are exceedingly terrible and make the itemization team look retarded. Gloves of Rapid Evolution, Shadowstrike, etc. These items should be disenchanted for Nexus Shards, but if someone wants to pay 40 DKP for the novelty of them that is ok.

Now that you have established the low end of the spectrum, the next group of items to work with are the class sets. You should assign points to the class sets in the order they drop in the dungeon, so that the last piece is the most expensive. You need to decide whether you want to assign a value to each individual piece of class gear, or if you just want all tier 1 bracers to be the same value. The second method is much easier, and in general the class sets are not so unbalanced that this will cause much drama. So using Blackwing Lair as an example, our guild has the values set as follows:

Bracers – 200 (Razorgore)
Belt – 240 (Vaelastrasz the Corrupt)
Boots – 280 (Broodlord Lashlayer)
Gloves – 320 (Firemaw, Ebonroc, Flamegor)
Shoulders – 360 (Chrommagus)
Chest – 400 (Nefarion)

So why are the bracers 200 and not 40? This is a simple matter of ratios. If the bracers were 40 and the belt was 80, that would make the belt twice as expensive as the bracers. The chest would be 6 times more expensive, meaning that for the price of a single chest piece, you could get the bracers, belt and boots. If you start at 200, then the belt is still more expensive than the bracers, but not by such a large margin. There’s no wrong way to value these items, and in the end it’s a decision you’ll have to make for your guild.

Finally, you have a huge mess of random epics, some really good some mediocre. It’s your job to try and figure out which of those items are upgrades and which are not, and where they all fit in the grand scheme of things. Well if you don’t feel like you’re up to the task, here is a little handy trick that will make things a lot easier. I assign every multiclass item a value of 200. Then I look at each item and figure out how many classes can use it, and add 40 points for each class. So let’s take a +healing cloak for instance. Three classes can use that: Druids, Priests, and Paladins/Shamans. The base item value is 200, +120 for 3 classes, for a total of 320. I do that for every item and get a large group of items that all cost between 240 and 320. Then I add 200 points if it’s a weapon, and 40 more points if it is a drop off the final boss in the dungeon. Basically the itemization comes out like this:

Junk and single class loot: 240
Weird leather items and healing offhands: 280
Most cloth items, and +healing items: 320
Final boss loot excluding weapons: 360
Bad weapons: 440
Good weapons: 560
Final boss weapons: 600

This is a good start, but the work is not over yet. Now, I actually look at the stats of the items, and if there is an item that is the best item in the game at the current time, I bump up the price. So items like Onslaught Girdle, even though only 2 classes can use it, end up being 320 because it was the best item in the game. (If only I knew it was still the best item in the game, I’d have made it 600.) If there is an item that is fairly useless compared to other items in the dungeon, I reduce the value by 40 or 80 points. The last step is to submit the point values for peer review. Usually you let the officers and class leaders look over the values, since they know their classes best. They can give you feedback on items that are too expensive, not expensive enough, or don’t fit the broad strategy of 200 +40/class that can wear it. After everyone gives their approval, you have item values to work with.

What if DKP is not for my guild?

Suicide Kings

Suicide Kings is an interesting system and a good alternative for those who do not want the complexity of a DKP system. In Suicide Kings, every player on the raid randomly rolls a number. That number represents your standing among all players in the raid. Here’s an example:

Warrior A rolls 992.
Warrior B rolls 661.
Warrior C rolls 3.

Warrior A has priority on the next drop. If an item drops and he wants it, he tells the raid leader and he goes to the bottom of the list. The new list would look like:

Warrior B with 992.
Warrior C with 661.
Warrior A with 3.

There is no re-rolling, each player moves up one position in the chart. If a replacement raider needs to come in, then that person rolls and is inserted into the loot order wherever his roll places him:

Warrior B with 992.
Warrior C with 661.
Priest D rolls 341.
Warrior A with 3.

Let’s say the Breastplate of Might drops, but Warrior B is saving for something else. He may pass the item down the line and retain his position. If Warrior C also passes (also saving for something else), then Warrior A can take the item for free. He is already in the last position, and cannot go further down. No loot will be disenchanted in this system.

If a player has been to a previous raid but cannot attend this one, then their spot is frozen in place. For instance, Warrior C is on vacation and not in the raid:

Warrior B with 992.
Warrior C with 661.
Priest D with 341.
Warrior A with 3.

If Warrior B takes an item, then he goes to the bottom of the list. However, Warrior C does not move up because he is not there:

Priest D with 992.
Warrior C with 661.
Warrior A with 341.
Warrior B with 3.

Players are awarded for attendance by skipping past players that are not there, but players are not penalized too harshly for missing a raid. Furthermore, players can hoard their points for a single item if they wish, because the person at the bottom of the list is guaranteed to take the item if it is free. I think the merits of this system are apparent, and this is a reasonable alternative to DKP.

Loot Council

Loot council brings immediate laughter from most people, because they don’t believe that people can be unbiased. Although extremely rare, loot council can work. Items are distributed by a group of players that decide who gets what. For instance, in Molten Core, the main tank should get all the fire resist pieces so the guild as a whole can progress quickly through the instance and take down Ragnaros. As a player in a guild that uses loot council, you tell the loot council what items you want, and they decide who gets items and in what order.

Advantages of Loot Council:
Easy – Nothing to keep track of, items are dealt with on a case by case basis.
Progression Based – Loot is distributed by humans, so the guild will generally progress as fast as possible.
No Wasted Loot – The loot is always distributed to someone because there is no penalty for taking an item.

Disadvantages of Loot Council:
Collusion – Players can work together behind the scenes to get more items for themselves.
Biased – The loot system is only as good as the council that distributes it.
Accountability – At any time, someone can claim favoritism or bias in the loot council and there is no way to disprove the accusation.

Although this can be a good system, it is generally not. MMOs are games that revolve around people. Sometimes those people are objective and thoughtful about the wants and needs of other people. Sometimes those people are selfish and close-minded, or lack the ability to see a situation from another human’s perspective. Unless everyone in the guild is very like-minded and trusting, loot council will fail. However, I’d like to give one example of where loot council has and continues to succeed.

There was a guild on my server that used loot council from the conception of the guild. Each player talked within his class and they as a group decided the distribution of items within their class. For instance, Warrior X would get the first suit of fire resist, because he was going to be the main tank for the fire bosses. Warrior Y was going to get the second suit, Warrior Z the 3rd and on. In exchange, Warrior Z was going to get the first two-handed axe of monster slaying, and Warrior X was going to get the first boots of fire giant kicking. Everyone agreed within their class how the loot was going to be distributed amongst themselves.

The loot council consisted of 5 of the guild officers. The only job they had was to settle disputes when classes could not agree, and decide which multi-class items went to which classes first. This encouraged them to work it out between themselves, because if they had to resort to talking to the council, it meant they lost the ability to bargain with each other. Repeated arguments to the loot countil usually meant dismissal from the guild for one or both players due to personality conflicts. The loot council also decided whether a sword went to Rogues or Warriors first, and those classes had to accept the decision. The system is only as good as the people in it, and it can work but that is rare.


Finding a DKP system that works for your guild may be the single most important decision that can be made, especially in a game like World of Warcraft where items are the only significant method of advancing a character. Choosing wisely will keep your raid strong and healthy months and years down the line, while poor decisions early will cause nothing but drama and greed. Just remember that it is never too late to correct your mistakes. Do not let your guild hemorrhage slowly, if you are losing raiders figure out why and do what is necessary to rectify the situation.

About the Author

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to Top ↑