I'm the main DKP person for our guild and have been for about the past year. Pre-TBC, my guild used a modified silent bidding system. It was basically a normal bid system with a twist... You got X points for every boss kill/attempt, loot was announced and bids are whispered to the loot master.. the highest bid wins. However, if two bids are within 50 of each other, the two people /roll 50 and their roll is added to their bid, making their total. This gave a small chance for people who bid slightly under someone to still have a chance, while still allowing people to "lock" the item by simply bidding a lot of DKP. There was also the clause that someone under 50% attendance would not be able to outbid someone with <85&.. There was a minimum bid (usually 20 or 30), and that was it.

This system worked fine for us from MC all the way through Naxx with minimal loot drama. Granted, we had our issues with it like any DKP system and had to make a few changes.. but overall it worked rather smoothly and most people were happy with it. Going into TBC, we had no idea how the new Tiers would work, and how progression would be layed out, and we talked about changing our DKP system. I kicked around the idea of going to a fixed price, zero sum system.. but a lot of people were against it for some reason.. I think mostly because we didn't know how to figure out the right costs for items being that we didn't even know what drops in TBC. And also it was a foreign concept to most and change is scary or something.

Well TBC came and we still hadn't decided on a system so our GM just laid one down that I guess he had been working on. But didn't discuss it with any the officers or anyone first, just sort of laid down the rules and we started using it. Here's how it works:

Points are awarded for Boss Kills. 10 for farm status 20 for first time.

Item price is decided by taking a percentage of the average top 5 DKP, for example:
Armor pieces are 25%
Weapons are 50%

So let's say a weapon drops.. the top 5 people's DKP are averaged out, and then the weapon costs 50% of that price. People bid, and the person who bids with the highest DKP wins.

In theory, it seems relatively simple and straightforward.... but in practice it is an inefficient nightmare. We've been using it for about 2 months now and every raid I remember why I don't like it. DKP and Loot distribution tonight in Gruul took 10 minutes for Maulgar and 15 for Gruul loot. What should be a simple process taking 1 or 2 minutes turns into an 20 minute long arduous task of adding and dividing, manually entering this, figuring out that. Then going through the DKP list and figuring out who had the highest out of who bid.

What I'm also seeing is that a lot of the costs for items are completely arbitrary and change a lot from week to week. A trinket will drop that may be a slight upgrade for someone, but no one will bid because it's not worth the 75 DKP it costs this week. So I have what I feel is people holding out on small upgrades because they are afraid of blowing all or most of their DKP in case something they really want drops.

There is also the issue of the cost of all items coming from the top 5 DKP average. So lets say 3 or 4 people that come to every single raid don't need anymore gear from those places, or just dont want to roll on sidegrades.. but keep showing up and keep gaining DKP. This raises the average for everyone, week after week. So you have a few people with 300+ DKP, and most everyone else with 150ish.. the people with less DKP are usually the ones rolling and we have had quite a few cases of people going negative already.

Now I've argued with my officers about how much I hate this system and how I don't think it will work at all in the long run, but they have been telling me I am overreacting and we need to give it more of a chance and that it will be self-correcting for inflation, and that as we progress further it will start to balance out more. But I just don't see it happening, at all. I see it causing lots of loot drama and being a general pain in my ass, as I'm the main DKP person.

We only have Kara, Kazzak, and Gruul on farm, and are -just- starting to work on Mag and Hydross. So we're only really using DKP right now for Gruul's Lair, and it's a nightmare for me. I can't even imagine using it for multiple bosses and crap every night.


Now for the $100,000 question - am I just overreacting and being irrational... or is our system doomed to failure? I'm not too big on math so I thought some of you number junkies might be able to give me (my guildies..) a logical reason why it's not going to work.

One guy told me to not worry about inflation because it would take care of itself since the people with the highest DKP always win so eventually the people in the top 5 are going to drop down thus lowering the average.. but that's not the trend I'm seeing in practice.


And if this system is as FUBAR as I think it is... can someone suggest a good one we can use in its place? I'm really interested in some form of a zero-sum, fixed-price system.. but I don't really understand the mechanics behind it enough to convince everyone that we should use it.

Any help is greatly appreciated.

I think you've stumbled upon the major issue with this system:

Item price needs to roughly correlate with desire for the item

And in this system, price is linked to something completely different: it's roughly based on the points of the most active raiders. Until the system attempts to tie desire for an item with price for an item, you're going to have issues.

Here try this link to a dkp site created by one of my old guild mates, it has received a lot of interest over the past few months. May help you to find a new system by trying different ones out.

www.gsdkp.com

We currently use it and works very well.

It sounds like you have 2 big problems. The first is that, as Tedv mentioned, that item price does not correlate with item value. The second problem is that you have tremendous amounts of inflation going on in the system. If this keeps up, what will happen is those few people with 300-400 will always get first pick on items by the fact they can simply brute force their way to the item(Assuming no one else in the 300-400 group also wants this item)

The inflation will probably make your system a mess in the long run, and probably should be stemmed by adding in DKP caps, or a DKP tax, or the like. In theory, it might work out that the top people will drop down for paying exorbitant amounts for the first item, but the amount they drop down by will always be low enough so they maintain their lead(And this assumes no collusion in bidding-if there are only a handful of people at the top and of different classes, then it could very well be that there is never any competition between them).

There used to be a very well written post on the wow guild relations forums about different dkp systems and how they function, I'll see if I can dig it up(Even though it is on the official forums, it is still a very good read and worth the time)

As a general guideline, as long as your system meets these three requirements, things will be (relatively) drama free:

• Item price roughly correlates to demand
• The system has built-in anti-inflation measures
• Its easy to determine and bookkeep an item's cost

The first criteria reduces sharding, the second makes sure that the same people don't get all the loot, and the third makes sure you don't lose valuable raiding time from item distribution. There are a LOT of systems that meet these three, but it seems like your system doesn't meet any. You might be able to argue for anti-inflation in that item prices increase as the top 5 people gain more points. If there's ever a time to scrap a loot distribution system and try something new, it's now.

Putting a system in place without discussing it with your officers (especially the guy who has to maintain the system) == bad form. That's your first problem, and I would focus on that if I were you. You are not overreacting with having to be stuck with a system that suffers from problems the previous posters have highlighted.

Your real problem is the way that your guildmaster introduced the system. It is now "his baby", and he will be very reluctant to admit that it is pure garbage. If you want to convince him about that, you cannot start a theoretical discussion, you will need hard evidence, like people leaving the guild or good items being sharded.

Try to start talking about the time it takes to perform the loot distribution. You gm will be hard pressed to pretend that 15 minutes is an acceptable time for that. So the price of an item needs to be determined before a raid starts.

You can gradually move the system to something better by fixing the problems when they become apparent. Try not to solve the problems you don't have yet, there are far too many of them in that crappy system.

Out of interest, are the officers/GM making up a good part of the top 10 or so DKP holders? If so then you may find they're directly benefiting and it'll be hard to convince them how bad it is for the rest of the guys if all they see is "I get the loot I want when I want it and I pass if I don't". Not necessarily saying your GM and officers are overly greedy but it might be something to think about in the way you approach getting the system changed.

We recently decided to switch to a wishlist system because of hoarding issues. In general, when using dkp, there are 3 possible things that can occur in general.

Note also that items being perfectly valued(which is very hard) will help ease the problems.

#1: Everyone takes items as they come, nothing shards, and the system functions fairly well.

#2: People hoard for big ticket items, the system ends up being inefficient as much gear gets sharded. This can often be an indication that item values are off.

#3: If your system is not zero sum, inflation can kick in and effectively devalue your entire system into a /roll.

Basically, if your guild is willing, wishlist allows you to stop sharding decent stuff while still letting everyone prioritize the best stuff. It has flaws, but, so does everything.

Your system seems particularly vulnerable to wild fluctuations in item values though, which in general is a very bad thing unless your system is a bidding system. I would recommend fixing item values to make your system more stable.

This system is absolutely awesome, but I couldn't use it because it doesn't have decimal points (how exactly do you use zero sum with no decimals!?). I know the guy administering it was hoping to add that, but as of a few months ago he hadn't yet.

2 things.
1. Way too complex, there's way too much work involved in implementation and upkeep.
2. The system doesn't scale well. 6 or 7 months down the road with people joining/leaving the raiding scene, and more content opened, while other content becomes obselete, you're looking at a terrific nightmare managing the system in the future.

I would suggest converting the current system to something less complex that will be easier to manage over time.

Well thanks everyone. This pretty much confirms my suspicions that I'm not just crazy and that we WILL run into some major problems down the line, even though we're having some now. It's already a nightmare to maintain, I can't imagine when we start downing multiple bosses a night.

And yes, generally our officers are all in the Top 10.. as most officers make every raid.

Then we have the issue of people in the Top 5, not logging on... thus their DKP staying the same.

This system is a nightmare. Can anyone think of an easy way we can convert to something easier? - http://www.awakeningwow.net/modules/...members.php?s=

How would you implement a system like this? I have no idea where to begin... and can't seem to find much info on setting up such a system

In my opinion, zero sum systems only work for progression guilds as they lack incentive for learning attempts. Yes you can hack that in but ultimately those hacks are just bandaids.

I don't know if you are a progression guild, but from the fact that Gruul is the hardest boss you have killed I think the answer is no. So some of the systems suggested here are probably not suitable. That said I agree that your current system is suboptimal.

Here's what we did:

We liked the idea of EPGP. It's basically a zero sum-esque differential system. You award Effort Points for any contributions that you want (for us we award attendance mainly but the door is open to award people for bringing consumables or farming mats the guild needs). Unlike traditional DKP where you then spend those points on gear, you have a second pool of points for the gear you earn called Gear points. You decide who wins an item by dividing EP/GP and whoever is highest wins. Inflation is handled by decaying each pool by a certain amount each raid. Hence recent efforts count in your favor more heavily than past ones and recent loot counts against you more heavily than past loot. The decay means no adjustments to balance the DKP. (Full write up of EPGP is found here: http://code.google.com/p/epgp/wiki/WhatIsEPGP_v2)

I know this sounds a lil complicated but there are ways to handle it. The creator of EPGP makes an excellent mod to handle it all for you however we found that it stores your EPGP numbers in the ingame officer notes. The pros to this are that the info is always available to people in game. The cons are that you can't store raid history in there (big one IMO) and if you have a bunch of clowns in your guild like we do who kick people out of the guild in jest.. it resets their DKP.

Our solution was to modify and EQDKP, CT_Raidtracker and a script we found from EQDKP forums called RTImport. Attendance and loot is managed thru CT_Raidtracker in game and at the end of the raid we just export an XML string to EQDKP which manages the hourly attendance awards, the loot awards and the decay function.

Meh, the setup was a bit of a pain but in the end we've got a system that can be managed with very little user interaction that follows the principle (if not the implementation) of EPGP.

Loot systems really have two parts: How you acquire points and how you spend points. The two don't have much to do with each other. "Zero Sum" is just a method that (mostly) deals with the inflation problem. Most zero sum systems use priority based fixed pricing for spending, but there are zero sum auction systems as well, or at least an equivalent. (Everyone starts at X points instead of 0, and if you spend Y points on an item, everyone in the raid gains Y/Raid_size points. The average points of each person in the system is still X.)

The standard way to give incentive for learning attempts is to award extra points, and that's the same in any system. It's not a bandaid. But when you award 10 points to 25 people in a zero sum system, you have to remove a total of 250 points from everyone not there. I don't see what makes this a hack.

Zero sum is indeed a means of acquiring points. You get the points others lose. But mixing points gained from items and points gained from other sources dilutes the basic principle of most zero sum systems. If all you use it for is an inflation check then it is probably fine, but then it is not really a system where you use zero sum to gain points. Then you basically use a fixed point system that uses zero sum as normalizer. As I said, very possible but not really in the spirit of most zero sum systems I have seen.

We use a semi-silent bid system. Points are accrued for attendance and kills (1 per, so you don't accumulate a lot quickly), and each item is bid on. Bids are sent to the ML and the current high bid + class of the bidder is announced. Bidding continues until there's a winner, and you can't bid more than you have. Anything that isn't bid on is sharded. You can't take a second of something over someone that doesn't have one, which leads to loot saturation relatively quickly, and tends to distribute loot well. The end result is that people can pay as much as they'd like for an item, and higher-demand items go for more, while low-demand or highly saturated items go for not-a-lot. Inflation really hasn't been a problem due to saturation.

We were using a bastardized Suicide Kings system before, which ended up being really horrid once we started getting raider turnover, and decided to just move to a straight-up bidding war system.

We use separate categories for class armor/everything else, to prevent class imbalances from causing problems, and it works out nicely.

That would definitely go a long way. An export to EQDKP wouldn't hurt either or something more dynamic than the static html tables at the least. I like having the numbers in game but when evaluating an applicant's attendance, the detailed history from EQDKP helps a lot. Storing DKP in the officer notes was the initial turn off. One guy gets kicked out each night.. its nearly a pre-raid tradition.

I didn't think you'd be interested in too exhaustive a set of changes but I'd be happy to post some thoughts on your Google code pages about it.

The final plan is to build an online service to export history and enable embedding of the web tables in your site. But that's a long way from what it is now. I just need to cover most of the needs of current users and take it from there. Things move slow though, since being a GM, an addon author and trying to have a life doesn't leave much time for addon writing :-)

Please write ideas up in the google forum. It will at least start some discussion and give me better feedback on people's needs.

We used a zero-sum system (pre-TBC) where each week the total points spent on items was handed out on a single day, based on how many hours of raids you'd attended. It made every Tuesday DKP "Payday", kept everything zero-sum, and made learning hours worth just as much as farm hours (though we eventually started counting MC as less, and BWL was capped at 3 hours total for the week, in order to make the learning stuff actually have MORE incentive).

It worked REALLY well and if we hadn't gone more casual post-TBC, we'd still be using it (we're using modded-/randoms now...)

Wow, pretty much every incorrect stereotype of dkp systems is summed up in the above posts. No incentive in ZS? Check - Our system is more fair? Check... Not reading? yep!

A new DKP system - Balance for Hybrids?

We use the one described by heel of vodka. Not the one in my original post.